Intelligent Analytics for

­Industry 4.0 Applications

The advancements in intelligent decision-making techniques have elevated the
­efficiency of manufacturing industries and led to the start of the Industry 4.0 era.
Industry 4.0 is revolutionizing the way companies manufacture, improve, and dis-
tribute their products. Manufacturers are integrating new technologies, including the
Internet of Things (IoT), cloud computing and analytics, and artificial intelligence
and machine learning, into their production facilities throughout their operations. In
the past few years, intelligent analytics has emerged as a solution that examines both
historical and real-time data to uncover performance insights. Because the amount of
data that needs analysis is growing daily, advanced technologies are necessary to col-
lect, arrange, and analyze incoming data. This approach enables businesses to detect
valuable connections and trends and make decisions that boost overall performance.
In Industry 4.0, intelligent analytics has a broader scope in terms of descriptive, pre-
dictive, and prescriptive subdomains. To this end, the book will aim to review and
highlight the challenges faced by intelligent analytics in Industry 4.0 and present the
recent developments done to address those challenges.
 Intelligent Analytics for
­Industry 4.0 Applications

Edited By
Avinash Chandra Pandey, Abhishek Verma,
­Vijaypal Singh Rathor, Munesh Singh, and
Ashutosh Kumar Singh
First edition published 2023
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742

and by CRC Press

4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2023 selection and editorial matter, Avinash Chandra Pandey, Abhishek Verma, Vijaypal Singh
Rathor, Munesh Singh, and Ashutosh Kumar Singh; individual chapters, the contributors

CRC Press is an imprint of Taylor & Francis Group, LLC

Reasonable efforts have been made to publish reliable data and information, but the author and
­publisher cannot assume responsibility for the validity of all materials or the consequences of their
use. The authors and publishers have attempted to trace the copyright holders of all material repro-
duced in this publication and apologize to copyright holders if permission to publish in this form has
not been obtained. If any copyright material has not been acknowledged please write and let us know
so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known
or hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.
com or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA
01923, 978-750-8400. For works that are not available on CCC please contact m
­ pkbookspermissions@
tandf.co.uk

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are
used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Names: Pandey, Avinash Chandra, editor. | Verma, Abhishek (Professor of
computer science), editor. | Rathor, Vijaypal Singh, editor. |
Singh, Munesh, editor. | Singh, Ashutosh Kumar, editor.
Title: Intelligent analytics for industry 4.0 applications / edited by:
Avinash Chandra Pandey, Abhishek Verma, Vijaypal Singh Rathor, Munesh
Singh and Ashutosh Kumar Singh.
Description: First edition. | Boca Raton : CRC Press, 2023. |
Includes bibliographical references and index.
Identifiers: LCCN 2022055814 (print) | LCCN 2022055815 (ebook) |
ISBN 9781032342412 (hardback) | ISBN 9781032342429 (paperback) |
ISBN 9781003321149 (ebook)
Subjects: LCSH: Industry 4.0. | Decision making—Data processing.
Classification: LCC T59.6 .I56 2023 (print) | LCC T59.6 (ebook) |
DDC 658.4/03028563—dc23/eng/20230302
LC record available at https://lccn.loc.gov/2022055814
LC ebook record available at https://lccn.loc.gov/2022055815

ISBN: 978-1-032-34241-2 (hbk)

ISBN: 978-1-032-34242-9 (pbk)
ISBN: 978-1-003-32114-9 (ebk)

DOI: 10.1201/9781003321149

Typeset in Times
by codeMantra
Contents
Acknowledgments......................................................................................................ix
Editors........................................................................................................................xi
Contributors............................................................................................................ xiii

Chapter 1 Analytics Approach for Intelligent Cyber-Physical System

Integration in Industrial Internet of Things (Industry 4.0)...................1
Lukman Adewale Ajao and Simon T. Apeh

Chapter 2 Digital Twins A State of the Art from Industry 4.0

Perspective........................................................................................... 23
Ganesh A. Deiva and P. Kalpana

Chapter 3 Human-Centered Approach to Intelligent

Analytics in Industry 4.0..................................................................... 37
Varad Vishwarupe, Milind Pande, Vishal Pawar, Prachi Joshi,
Aseem Deshmukh, Shweta Mhaisalkar, Shrey Maheshwari,
Neha Mandora, and Nicole Mathias

Chapter 4 Advance in Robotics Industry 4.0....................................................... 55

Muchharla Suresh, Rajesh Kumar Dash,
Deepika Padhy, Swastid Dash, Subhransu Kumar Das,
and K. Nageswara Rao Achary

Chapter 5 A Cloud-Based Real-Time Healthcare Monitoring System for

CVD Patients....................................................................................... 65
Anjali Kumari, Animesh Kumar, Kanishka Vijayvargiya,
Piyush Goyal, Ujjwal Tripathi, and Avinash Chandra Pandey

Chapter 6 Assessment of Fuzzy Logic Assessed Recommender System

A Critical Critique............................................................................... 73
S. Gopal Krishna Patro, Brojo Kishore Mishra, and
Sanjaya Kumar Panda

Chapter 7 Intelligent Analytics in Big Data and Cloud: Big Data;

Analytics; Cloud.................................................................................. 85
K. Aditya Shastry and B. A. Manjunatha

v
vi Contents

Chapter 8 Various Audio Classification Models for Automatic Speaker

Verification System in Industry 4.0................................................... 113
Sanil Joshi, Mohit Dua, and Shelza Dua

Chapter 9 Trending IoT Platforms on Middleware Layer.................................. 131

Neha Katiyar, Priti Kumari, Surabhi Sakhshi, and Jyoti Srivastava

Chapter 10 Healthcare IoT: A Factual and Feasible Application

of Industrial IoT................................................................................. 149
Subasish Mohapatra, Amlan Sahoo, Subhadarshini Mohanty,
and Munesh Singh

Chapter 11 IoT-Based Spacecraft Anti-Collision HUD Design Formulation..... 165

Bibhorr

Chapter 12 Coverage of LoRaWAN in Vijayawada: A Practical Approach........ 175

P. Saleem Akram and M. Lakshmana

Chapter 13 Intelligent Health Care Industry for Disease Detection.................... 185

Ishaan Deep, Payaswini Singh, Aditi Trishna, Sanskar,
Biswa Ranjan Senapati, and Rakesh Ranjan Swain

Chapter 14 Challenges with Industry 4.0 Security.............................................. 195

Mohit Sajwan and Simranjit Singh

Chapter 15 Dodging Security Attacks and Data Leakage Prevention for

Cloud and IoT Environments............................................................209
Ishu Gupta, Ankit Tiwari, Priya Agarwal, Sloni Mittal, and
Ashutosh Kumar Singh

Chapter 16 Role of Blockchain in Industry 4.0.................................................... 233

Keshav Kaushik and Bhavana Kaushik

Chapter 17 Blockchain and Bitcoin Security in Industry 4.0.............................. 241

Yadunath Pathak and Praveen Pawar
Contents vii

Chapter 18 Technology in Industry 4.0................................................................ 253

Rashmy Moray and Amar Patnaik

Chapter 19 Intelligent Analytics in Cyber-Physical Systems.............................. 271

Manjushree Nayak, Priyanka P. Pratihari, Sanjana Mahapatra,
and Shyam Sundar Pradhan

Chapter 20 An Overlook on Security Challenges in Industry 4.0....................... 281

R. Ramya, S. Sharmila Devi, and Y. Adline Jancy

Index....................................................................................................................... 293
Acknowledgments
We express our heartfelt gratitude to CRC Press (Taylor & Francis Group) and the
editorial team for their guidance and support during completion of this book. We are
sincerely grateful to reviewers for their suggestions and illuminating views for each
book chapter presented here in Intelligent Analytics for Industry 4.0 Applications.

ix
Editors
Dr. Avinash Chandra Pandey (Member, IEEE) is
currently an Assistant Professor in the Discipline of
Computer Science & Engineering at PDPM Indian
Institute of Information Technology Design and
Manufacturing, Jabalpur. He has more than 8 years
of teaching and research experience. He has guided
many M.Tech. dissertations and B.Tech. projects.
He has published more than 20 journal and confer-
ence papers in the area of text mining, NLP, and
soft computing. His research areas include data
analytics, NLP, social network analysis, IoT, cyber-
physical systems, and soft computing.

Dr. Abhishek Verma i s an Assistant Professor in the

Department of Computer Science & Engineering at
IIITDM Jabalpur, India (an institution of national
importance). He obtained a Ph.D. degree (2020) in
the Internet of Things security from the National
Institute of Technology Kurukshetra, Haryana,
India. He has more than 7 years of experience in
research and teaching. He has published several
research articles in international SCI/SCIE/Scopus
journals and conferences of high repute. He is an
editorial board member of Research Reports on
Computer Science (RRCS) and an active review
board member of various reputed journals, including IEEE, Springer, Wiley, and
Elsevier. His current areas of interest include information security, intrusion detec-
tion, and the Internet of Things.

Dr. Vijaypal Singh Rathor received his M.Tech.

from Maulana Azad National Institute of Technology,
Bhopal, and Ph.D. from ABV-Indian Institute of
Information Technology and Management Gwalior,
India in 2014 and 2020, respectively. At present, he is
working as an Assistant Professor in CSE Discipline
at PDPM Indian Institute of Information Technology,
Design and Manufacturing (IITDM), Jabalpur, India.
Prior to joining IIITDM, he worked as an Assistant
Professor at Thapar Institute of Engineering and
Technology, Patiala, India. He has also worked as an
Assistant Professor in CSE Department at Bennett
University, Greater Noida, India. Dr. Rathor has

xi
xii Editors

also received a grant as a Principal Investigator for the Project entitled “HT-Pred: A
complete defensive machine learning tool for Hardware Trojan Detection” from the
Data Security Council of India (DSCI). His research interests include hardware secu-
rity, trustworthy circuit design, hardware Trojan, machine learning, IoT, and cloud
computing.

Dr. Munesh Singh  received a Ph.D. degree in

Computer Science and Engineering from the
National Institute of Technology (NIT), Rourkela,
India, in 2017. He has industrial experience of one
year with the Honeywell Technology Solution Lab,
Bangalore, India. He also has academic experience
with VIT University from 2017 to 2018 and with
IIITDM Kancheepuram from 2018 to 2021. He is
currently an Assistant Professor in PDPM Indian
Institute of Information Technology, Design and
Manufacturing (IIITDM), Jabalpur, India. He has
published 14 research articles in IEEE, Springer,
and Elsevier journals. He works as a reviewer for Wiley, IEEE, and Springer. His
research interests include wireless sensor networks, cloud computing, software-
defined networks, industrial IoT, embedded system, and robotics. He is a professional
member of IAENG and IEEE.

Dr. Ashutosh Kumar Singh is a Professor and

Head in the Department of Computer Applications,
National Institute of Technology Kurukshetra,
India. He has research and teaching experience in
various universities of India, the UK, and Malaysia.
He received his Ph.D. in Electronics Engineering
from the Indian Institute of Technology, BHU, India
and Post Doc from the Department of Computer
Science, University of Bristol, UK. He is also
charted engineer from the UK. His research area
includes verification, synthesis, design and testing
of digital circuits, data science, cloud computing,
machine learning, security, and big data. He has published more than 330 research
papers in different journals, conferences, and news magazines.
Contributors
K. Nageswara Rao Achary Swastid Dash
Department of Electronics and NIST
Communication Engineering Odisha, India
NIST
Odisha, India Ishaan Deep
Department of Computer Science and
Priya Agarwal Engineering
Amdocs Development Center India LLP Siksha ‘O’ Anusandhan University
Haryana, India Odisha, India

Lukman Adewale Ajao A. Deiva Ganesh

Department of Computer Engineering Indian Institute of Information
Federal University of Technology Technology
Minna Kancheepuram, India
Minna, Nigeria
Aseem Deshmukh
P. Saleem Akram MIT Pune
Department of ECE Pune, India
Koneru Lakshmaiah Education
Foundation S. Sharmila Devi
Guntur, Andhra Pradesh, India Department of Electronics and
Communication Engineering
Simon T. Apeh Sri Ramakrishna Engineering College
Department of Computer Engineering Coimbatore, India
University of Benin
Benin City, Nigeria Mohit Dua
Department of Computer Engineering
Bibhorr National Institute of Technology
IUBH International University Kurukshetra, Haryana
Bad Honnef, Germany
Shelza Dua
Subhransu Kumar Das Department of Electronics and
Department of Electronics and Communication Engineering
Communication Engineering National Institute of Technology
NIST Kurukshetra, Haryana
Odisha, India
Piyush Goyal
Rajesh Kumar Dash Discipline of Computer Science &
Department of Electronics and Engineering
Communication Engineering PDPM IIITDM
NIST Madhya Pradesh, India
Odisha, India

xiii
xiv Contributors

Ishu Gupta M. Lakshmana Kumar

National Sun Yat-Sen University Department of ECE
Kaohsiung, Taiwan Koneru Lakshmaiah Education
Foundation
Y. Adline Jancy Guntur, Andhra Pradesh, India
Department of Electronics and
Communication Engineering Anjali Kumari
Sri Ramakrishna Engineering College Discipline of Computer Science &
Coimbatore, India Engineering
PDPM IIITDM
Prachi Joshi Madhya Pradesh, India
VIIT Pune
Pune, India Priti Kumari
Noida Institute of Engineering &
Sanil Joshi Technology
Department of Computer Engineering Greater Noida, India
National Institute of Technology
Kurukshetra, Haryana Sanjana Mahapatra
Department of Computer Science and
P. Kalpana Engineering
Indian Institute of Information NIST
Technology, Design and Odisha, India
Manufacturing
Kancheepuram, India Shrey Maheshwari
MIT Pune
Neha Katiyar Pune, India
Noida Institute of Engineering &
Technology Neha Mandora
Greater Noida, India MIT Pune
Pune, India
Bhavana Kaushik
University of Petroleum & Energy B. A. Manjunatha
Studies Nitte Meenakshi Institute of Technology
Uttarakhand, India Bengaluru, India

Keshav Kaushik Nicole Mathias

University of Petroleum & Energy Georgetown University
Studies, Uttarakhand, India Washington, DC

Animesh Kumar Shweta Mhaisalkar

Discipline of Computer Science & MIT Pune
Engineering Pune, India
PDPM IIITDM
Madhya Pradesh, India
Contributors xv

Brojo Kishore Mishra Milind Pande

GIET University MIT WPU Pune
Odisha, India India

Sloni Mittal Avinash Chandra Pandey

Hewlett-Packard (HP) Discipline of Computer Science &
Haryana, India Engineering
PDPM IIITDM
Rashmy Moray Madhya Pradesh, India
Symbiosis Institute of Management
Studies (SIMS) Yadunath Pathak
Symbiosis International (Deemed) Visvesvaraya National Institute of
University Technology
Pune, India Nagpur, India

Subhadarshini Mohanty Amar Patnaik

Department of Computer Science & Symbiosis Institute of Management
Engineering Studies (SIMS)
Odisha University of Technology AND Symbiosis International (Deemed)
Research University
Bhubaneswar, India Pune, India

Subasish Mohapatra S. Gopal Krishna Patro

Department of Computer Science & GIET University
Engineering Gunupur, India
Odisha University of Technology AND and
Research Department of CSE
Bhubaneswar, India Koneru Lakshmaiah Education
Foundation
Manjushree Nayak Vaddeswaram, Andhra Pradesh, India
Department of Computer Science and
Engineering Praveen Pawar
NIST Visvesvaraya National Institute of
Odisha, India Technology
Nagpur, India
Deepika Padhy
Department of Electronics and Vishal Pawar
Communication Engineering MIT WPU Pune
NIST India
Odisha, India
Shyam Sundar Pradhan
Sanjaya Kumar Panda Department of Computer Science and
NIT Warangal Engineering
Hanamkonda, Telangana, India NIST
Odisha, India
xvi Contributors

Priyanka P. Pratihari Ashutosh Kumar Singh

Department of Computer Science and National Institute of Technology
Engineering Kurukshetra, India
NIST
Odisha, India Munesh Singh
Department of Computer Science &
R. Ramya Engineering
Department of Electronics and PDPM IIITDM
Communication Engineering Jabalpur, India
Sri Ramakrishna Engineering College
Coimbatore, India Payaswini Singh
Department of Computer Science and
Amlan Sahoo Engineering
Department of Computer Science & Siksha ‘O’ Anusandhan University
Engineering Odisha, India
Odisha University of Technology AND
Research Simranjit Singh
Bhubaneswar, India Bennett University Greater Noida
Greater Noida, India
Mohit Sajwan
Bennett University Greater Noida Jyoti Srivastava
Greater Noida, India Madan Mohan Malviya University of
Technology
Surabhi Sakhshi Gorakhpur, India
Indian Institute of Management
Rohtak, India Muchharla Suresh
Department of Electronics and
Sanskar Communication Engineering
Department of Computer Science and NIST
Engineering Odisha, India
Siksha ‘O’ Anusandhan University
Odisha, India Rakesh Ranjan Swain
Department of Computer Science and
Biswa Ranjan Senapati Engineering
Department of Computer Science and Siksha ‘O’ Anusandhan University
Engineering Odisha, India
Siksha ‘O’ Anusandhan University
Odisha, India Ankit Tiwari
TechMatrix IT Consulting Pvt. Ltd.
K. Aditya Shastry Uttar Pradesh, India
Nitte Meenakshi Institute of Technology
Bengaluru, India
Contributors xvii

Ujjwal Tripathi Kanishka Vijayvargiya

Discipline of Computer Science & Discipline of Computer Science &
Engineering Engineering
PDPM IIITDM PDPM IIITDM
Madhya Pradesh, India Madhya Pradesh, India

Aditi Trishna Varad Vishwarupe

Department of Computer Science and Amazon, University of Oxford, MIT
Engineering Pune
Siksha ‘O’ Anusandhan University Pune, Maharashtra
Odisha, India
 1 Analytics Approach for
Intelligent Cyber-Physical
System Integration in
Industrial Internet of
Things (Industry 4.0)
Lukman Adewale Ajao
Federal University of Technology
University of Benin

Simon T. Apeh
University of Benin

CONTENTS
1.1 Introduction.......................................................................................................2
1.1.1 Cyber-Physical System (CPS)................................................................3
1.1.2 Industrial Internet of Things (IIoT)/Industry 4.0..................................3
1.1.2.1 Advantages of Industry 4.0 Technology.................................4
1.1.2.2 Challenges of Industry 4.0......................................................4
1.1.3 Research Motivation..............................................................................5
1.2 Intelligent Cyber-Physical System Design Consideration................................. 6
1.2.1 Fault Tolerance and Reliability..............................................................6
1.2.2 Scalability µ (S).....................................................................................7
1.2.3 Hardware System Design......................................................................7
1.2.4 Network Topology.................................................................................7
1.2.5 Data Model Design................................................................................8
1.2.6 Operating Environment.........................................................................8
1.2.7 Communication and Transmission Technology.................................... 9
1.2.8 Energy Consumption........................................................................... 10
1.2.9 Routing Protocols................................................................................ 11
1.2.10 Security Challenges............................................................................. 12
1.3 The Industrial Internet of Things Architecture Design and
Industry 4.0...................................................................................................... 13
1.4 The Communication Model of ICPS in Industry 4.0...................................... 15
1.5 Simulation Results of Cyber-Physical Communication Model in IIoT
Environment.................................................................................................... 16
DOI: 10.1201/9781003321149-1 1
2 Intelligent Analytics for Industry 4.0 Applications

1.6 Conclusion....................................................................................................... 19
References.................................................................................................................20

1.1 INTRODUCTION
Internet of Things (IoT) is a worldwide network facility that is capable of connecting
billions of devices remotely for interaction and information exchanges over Internet
protocol version 6 (IPv6) for low wireless personal area network (6LoWPAN) [1].
The IoT architecture is flourished with resource-constrained devices that may include
embedded sensors, wireless connectivity, communication protocols, and information
technology (IT) [2]. The communication of these devices depends on the restraint
protocols used for the data packet forwarding and routing information through the
6LoWPAN adaptation layer [3]. These protocols include IPv6, 6LoWPAN, and rout-
ing protocol for low power and lossy network (RPL). The 6LoWPAN adaptation layer
is utilized to overcome the challenges of end-to-end connectivity, and interoperabil-
ity between both physical and virtual things identities through IEEE 802.15.4 in the
network [4]. IoT is a promising emergent technology that provides conveniences,
comforts, and quality of life for domestic users, industries, research institutes, and
many others through the procuring of smart technology devices in Industry 4.0 [5]. It
also helps to create more opportunities in the physical world, better economic prac-
tices, and computational ecosystems. While the interaction of things with a human is
possible through the IoT communication layers which can be classified into physical/
data link, network, application layer, and cloud computing layer [6].
The physical and data link layer utilized Bluetooth low energy (BLE), ZigBee
(IEEE 802.11.15), Wi-Fi (IEEE 802.11b/n/g), and many others for communica-
tion and routing of big data. The IPv6, 6LoWPAN, RPL, CARPL, or CORPL was
adopted over the network adaptation layer for seamless end-to-end connection, com-
munication between the physical environment and logical layers [7]. The routing
protocol for low power and lossy network (RPL) was adopted at the lower link net-
work layer, route optimization, and energy management. Thus, the channel-aware
protocol (CARP) was used for underwater acoustic communication or in a hazard-
ous environment. The cognitive routing protocol for low power and lossy network
(CORPL) was used for cognitive radio or terrestrial communication [8]. However, the
­application layer functions on the existing constrained application protocols (CoAP)
and representational state transfer (REST) for the application program interface
(API). This cap is popularly used to conform with constrained devices and for the
seamless i­nteraction between the application layer and cloud computing in the IoT
networks [9].
The term smart technology in the IoT refers to the physical and logical connectiv-
ity adopted for the system conditioning to provide some functions such as remote
monitoring, surveillance, and tracking using various sensing technologies (sensors),
actuators, robotics, and artificial intelligence [10,11]. The roles of the smart system
in the emergent IIoT are varied like passenger trackers, car geolocation monitoring
using GPS [12], door access control [13], and security measures using RFID cards
[14]. So, smart technology is for green information that connects different indus-
trial technology (IT) products with things related over the network through wireless
Analytics Approach for Intelligent Cyber-Physical System Integration 3

Cyber Components

Embedded Microchip
Unit for Computaon
Cyber-Physical System
Computaonal Network Human-Machine
Interface (HMI)

Embedded
Actuator Sensor
System

Physical Components
Computer System
Environment

FIGURE 1.1  The cyber-physical system architecture.

devices [15]. Some of these industrial products are cyber-physical systems (CPSs),
sensors, artificial intelligence (AI), robotics, and so on.

1.1.1  Cyber-Physical System (CPS)

The CPS is incorporation of different facilities such as digital computing, network-
ing, physical objects, and human interfacing systems. This CPS can widely be used
for the interrelation of physical components through the human–machine inter-
face (HMI) and computational networks to achieve real-time monitoring, big data
acquisition, automation, surveillance, and better ecosystem linkages [16]. The CPS
implementation comprises sensors, actuators, embedded wireless facilities, and
­
­intelligence microchips to achieve various levels of remote-control decisions and far
distance monitoring. The illustration in Figure 1.1 shows the CPS design architecture.
This efficient design of a CPS has been considered for its suitability in the present
IoT and the future emergent IIoT that will accommodate Industry 4.0 technologies.

1.1.2 Industrial Internet of Things (IIoT)/Industry 4.0

The IIoT is an enhanced global communication technology that was built to augment
the functionalities of IoT operation through the interconnection of sophisticated sen-
sors, robotics, CPS, industrial applications, business strategies, and manufacturing
products [17, 18]. This is a technology that built a strong merger for adopting AI,
cloud computing, real-time data processing, smart factory, and resource planning
enterprises over the constraint networks [19]. While Industry 4.0 is the recent fourth
industrial revolution technology, its emphases are on the integration of industry man-
agement, manufacturing companies, and their products in the IIoT networks [20,
21]. The Industry 4.0 revolution platform accommodates the integration of business
processes into the model for resource-rich utilization, technological savvy stakehold-
ers, and supply chain management [22, 23]. Others include digitizing information,
4 Intelligent Analytics for Industry 4.0 Applications

Smart Factory Resource

Real-Time Planning
Data Enterprise
Processing

Internet of
Cyber- Things (IoT)
Physical
system (CPS) Industry 4.0
/IIoT
Technology Digitization
Cloud
Computing

Big data
Machine
Artificial
Learning
Intelligence
(ML )
(AI)

FIGURE 1.2  The components merger in Industry 4.0.

automation, big data, CPS, machine learning for enterprise resource planning, and
timely analytics in customer satisfaction. The illustration of components merger in
Industry 4.0 is presented in Figure 1.2.
The aim of this emerging fourth industrial revolution (Industry 4.0) was not only
to focus on the interconnectivity of smart systems but also to emphasize holistically
manufacturing approaches that will provide linkages between physical and digital
technology. It will also provide better collaboration between the business vendors,
partners, products, and people belonging to Industry 4.0. Meanwhile, interoperability
among the smart products is crucial to boost instant data productivity and enhance
processes for fast growth.

1.1.2.1  Advantages of Industry 4.0 Technology

The technology of Industry 4.0 is compared with the IIoT, and it delivers a wide range
of benefits through the life cycle of CPS products and supply chain management.
It includes sharing of real-time information, views of relevant business processes,
productions through engineering and design, sales and inventory, field services, and
customers [24, 25]. Some other benefits are presented in Figure 1.3.

1.1.2.2  Challenges of Industry 4.0

The incorporation of emergent smart manufacturing or CPS in the IIoT faces
some challenges. Industry 4.0 technology focuses are toward the interconnectiv-
ity and interoperability of smart IoT products with resource enterprises. The focus
also includes better manufacturing companies or ecosystems that will enhance
Analytics Approach for Intelligent Cyber-Physical System Integration 5

It helps in positioned It improves the collaborations

It makes the companies
and retain of hi-tech among various units of hi-tech
more competitive and
or experience young workers by leverage on real-time
timely-to the markets in
workers data and adopt intelligence
their productions
approach in the decision making

It helps to manage and

It enable the use of
predictive and
Industry 4.0 optimize manufacturing
processes and supply
prescriptive analytics Benefits chain
in problem solving

It helps to gain visibility into

business operations with in- It helps to create a It helps to boost the
built reporting, and retains consistent experience or business profitability by
customers by delivering workshop training for making smarter decisions on
quality products on time customer everytime the businesses

FIGURE 1.3  Illustration of Industry 4.0 benefits.

Big Data Poor Inter-

Cyber
Storage
Threats relaon Team
(Hackers) Player

Industry 4.0 Inadequate

Smart Product
Technology of Human
discriminaon
Challenges Resources

Smart
Poor IT Poor Supply Manufacturing/
Infrastructure Chain Discrepancy
Management

FIGURE 1.4  Industry 4.0 realization challenges.

automation, and real-time big data acquired through the use of machine learning,
robotics, and AI with a vision of business strategy that involved people, partners,
and products [26]. Figure 1.4 illustrates some of the Industry 4.0 challenges over the
emergent IIoT networks.

1.1.3 Research Motivation
The IIoT platform is an explosive technology that enhances the evolution and avail-
ability of smart systems in recent times. But the cooperation of this CPS component
6 Intelligent Analytics for Industry 4.0 Applications

brings a lot of concerns during the operation and design phase, such as security
vulnerability, power consumption, interoperability, communication delay, quality of
services (QoS) deterioration, and many others. However, the researchers have not
discussed these issues holistically but rather focus on one pertinent issue like security
attacks in the CPS based on IIoT, and optimization of power prediction. These evolv-
ing challenges will continue to disturb the existence of the IIoT network, and its future
focuses on the emerging Industry 4.0 revolution. This research work is organized by
introducing the Internet of Things, smart technology, CPS, IIoT, and Industry 4.0 in
Section 1.1. It also includes research motivation and organization of research. Section
1.2 presents a CPS design consideration. IIoT architecture design and Industry 4.0
are discussed in Section 1.3. Section 1.4 presents the CPS communication model in it
using Markov Chain Model (MCM). Section 1.5 shows the ­simulation results of the
communication model and performance analysis. Section 1.6 concludes the research
with recommendations for future research.

1.2 INTELLIGENT CYBER-PHYSICAL SYSTEM

DESIGN CONSIDERATION
The design consideration for this smart embedded wireless system or intelligent
cyber-physical system (ICPS) in the fourth generation of an industrial revolution
depends on hardware design, software coding, and control theory with a holistic
approach of computation to suit physical environmental constraints. For instance, a
typical mote in a wireless sensor network (WSN) consists of sensor nodes, a micro-
controller unit (MCU), and a wireless transmission device such as a radio frequency
(RF) circuit. This mote is mostly powered using a lithium battery or other means
of energy and is very hard to replace. Then, the power consumed during sensing
and packet transmission is a critical stage in that CPS design, which can shorten the
longevity of a sensor node. Others are hardware components that are involved in
the designed architecture. Particularly, the Internet of Things (IoT) gateway called a
radio or router consumes the maximum amount of power during packet transmission
and receiving [27]. So, it is very significant to be enlightened on these analytic fac-
tors that can affect the efficient design and performance of the ICPS in the emergent
Industry 4.0. The challenges of ICPS design include fault tolerance and reliability,
scalability, hardware constraint, network topology design, operating environment,
transmission media, routing algorithm techniques, energy consumption, and security.

1.2.1 Fault Tolerance and Reliability

Fault tolerance is the ability to sustain system functionalities without any interrup-
tion, or failure that occurs in any part of the system. This failure occurs as a result
of environmental interferences, physical damages, or energy source depletion, which
in norms and circumstances should not affect the network. Reliability R(t) can be
described as the probability of a system that will continue to survive until it becomes
unreliable (failure) F(t) as expressed in equation (1.1). The unreliability of the system
and the probability that an embedded system will not fail within the interval (0, t) are
expressed as in equations (1.2) and (1.3):
Analytics Approach for Intelligent Cyber-Physical System Integration 7

R ( t ) = P ( X > t ) = 1 − F ( t ) (1.1)

F ( t ) = 1 − R ( t ) (1.2)

R j (t ) = e − λ jt (1.3)

where X is the random variation of the component lifetime, P is the probability, λj is

the failure rate of node j, and t is the interval period (time).

1.2.2 Scalability μ (S)
Scalability is the ability to extend or increase the size of a network node without
collapsing the network structure and services. It is a process of adding more sensor
nodes to a network and still maintaining its efficiency even when several numbers of
nodes were added to it. Therefore, a scalable network is a network model that grows
with increasing in-network loads as many of these Industry 4.0 products are welcome
and integrated into the platform, such as robotic systems, machine learning, AI, big
data, techno-preneureship, and many others. The scalability of a network system can
be calculated as expressed in equation (1.4):

µ( S ) = MπT 2 ⁄E (1.4)

where M is the number of scattered sensor nodes in the network environment (E), T
is a transmission radio range, and E is the network environment.

1.2.3 Hardware System Design

Hardware system design refers to the parallel integration or the interconnection of
the components, where each component performs its specific function as proved in
the analytical models of designs. The hardware component implementation is con-
figured with coding techniques (computational models), which leads to constraints in
processor speeds, memory capacity for data storage, power utilization, and hardware
failure rates. The ICPS design in the revolution of it includes hardware component
integration as a full-functional device (FFD) such as microchips, memory, proces-
sors, and power as depicted in the block diagram of Figure 1.5. This system design
may also include additional units like a global positioning system (GPS) for location
detection and industrial applications.

1.2.4 Network Topology
Network topology is a model or structural linkage designed for the controllability of
both internal and external sensor nodes, which are interconnected to capture or share
empirical properties in a network. Therefore, these topological properties in relating
information and profile control between source and sink nodes are constraints when
integrated within an embedded wireless system. This constraint is a result of the
complexity of the system (density of nodes) and interaction between pairs of nodes
8 Intelligent Analytics for Industry 4.0 Applications

Memory
(Storage Unit)

Field Wireless Analog-Digital Processor Actuator

Digital-Analog
(Microcontrolle
Sensor Converter Module r Unit) Module Unit
Controller Unit

FIGURE 1.5  Block diagram of wireless-based hardware system design.

connected (unweighted edge components) over the networks. Therefore, a ­suitable

topology is important to be adopted for efficient communication in the network due
to the various number of sensor nodes involved in the network for surveillance, track-
ing, and monitoring. The most common network topologies used are star, tree, mesh,
and hybrid. Each of these topologies has challenges and better performances in the
network as analyzed in Table 1.1, based on their communication range, power con-
sumption, and synchronization period.

1.2.5 Data Model Design

The data model refers to the application requirements for the control of low latency,
reading, and writing data with a predefined pointer into a data block [28]. This block
of data model required different data types with the predefined schema to store data
or a list of values sorted in a table. It includes primitives, collections (lists, maps, and
sets), primary key, keyspace (a collection of related tables), and so on. As a result of
this illustration in Figure 1.6, efficient data model design for CPS is a critical aspect
of smart technology design for better functions in the Industry 4.0 revolution.

1.2.6 Operating Environment
The wireless sensor node is expected to be effectively designed when they are
densely deployed directly in contact with the phenomenon or closer to the object for
efficient operations and adaptation to the environmental condition. As in robotic sys-
tem design for a specific application using AI techniques or machine learning for the

TABLE 1.1
Network Topology Analysis
Topology Range Power Period
Mesh Long Higher Not at all
Star Short Low Not at all
Tree Long Low Yes
Hybrid Long Very low Depending on the sensor pattern
Analytics Approach for Intelligent Cyber-Physical System Integration 9

Service

Virtual
Entity Knowledge

City
Health (geographical)

AccessRights
User

FIGURE 1.6  IIoT data model design networks.

big data acquisition and analysis. In addition, other applications such as underwater
acoustic monitoring, surveillance on battlefields, harsh conditions like the nozzle of
an aircraft engine, and so on also require effective design of wireless sensor node.
The formation of wireless connections and network services in erratic environ-
mental conditions, such as atmosphere/weather, rain, and icing conditions, requires
­improvement to meet the standard of Industry 4.0 products.

1.2.7  Communication and Transmission Technology

Signal transmission is very important and crucial to ICPS communication by adopt-
ing standardized industrial wireless technologies such as RF and industrial scien-
tific and medical (ISM) bands. This will enhance the efficient communication and
interoperability among several things connected over the network of Industry 4.0.
In wireless communication systems, the frequency used for transmission affects the
amount of data and the speed at which the data can be transmitted. The strength or
power level of the transmission signal determines the distance over which the data
can be sent and received without errors or loss of signal. In general, the principle that
governs this wireless transmission dictates the lower channel frequency can carry
less data more slowly over a long distance. The distance between the centers of the
cell that uses the same cluster frequency shows the cluster size C and cluster cell. The
frequency reuse distance (frd) is used to normalize the size of each cell in hexagonal
shape as expressed in equations (1.5) and (1.6):

frd = sqrt{3C} (1.5)

C = i 2 + ij 2 + j 2 (1.6)

where C = 1, 3, 4, 7, 9, … as imaginable cluster size or single-cell size, and i and j are
integers that determine a relative location of coexisting channel cells. Therefore, the
equation can be rewritten as follows:

frd = sqrt{3(i 2 + ij 2 + j 2 )} (1.7)

10 Intelligent Analytics for Industry 4.0 Applications

The effective power transmission between the antenna transmitters to the receiver
(Pr, Pt) through the antenna area and their directivity (Dt, Dr) is expressed as in
equations (1.8–1.10):

Pr ⁄Pt = (ar at ⁄d 2λ 2 )2 (1.8)

Pr ⁄Pt = DtDr (λ⁄ 4πd )2 (1.9)

Pr = Pt + Dt + Dr + 20 log10 (λ⁄ 4πd )2 (1.10)

where Ar and At are the effective radio antenna areas for receiving and transmitting
the signal, d is the distance between radio antennas, and λ is an RF antenna.

1.2.8 Energy Consumption
The battery life of ICPS can be significantly extended if sensor nodes are configured
to coordinate or dictate when the entire circuit should be ON or OFF or by using a
meta-heuristic algorithm. A sensor node in the ICPS can be configured to determine
when the entire circuit should begin monitoring an event, by activating the circuit or
else should be put into deep sleep mode for energy savings. The lifetime of WSN is
highly dependent on the power available at each node in the network called active
power (ActiveP). Without providing any reverse energy, this power is draining off
during operation. It means energy constraint in an embedded wireless system [29,
30]. The average power (AvP) consumption of a sensor node and the sensor node duty
cycle (Dnode), which is the fraction of time (t) when the node is active and the period
is T, can be expressed as in equations (1.11–1.13):

AvP = Active P ∗ Dnode (1.11)

DN = Activet /T (1.12)

T = Activet + Sleept (1.13)

However, one of the most challenges of this ICPS component is a power constraint
because each node or component in the architecture will be loaded and expended
energy of the information flow. The energy expended by each node of ICPS in the
network will be dependent on the functions (loads) which can be affected by location
(topology) and material properties (operating characteristics). The power consump-
tion of the node and the entire network can be estimated. Let us consider a single
ICPS Ψ𝑝𝑠 = (𝜙𝑛, 𝜉𝑠) with components (nodes) integrated and determine the load 𝜁, but
explicit service classifications are disregarded, and faulty physical node (component)
is not considered. The total load of the ICPS per unit time is given as equation (1.14),
and the structural load of ICPS node j up-to-the-minute time step n is expressed as in
equation (1.15). However, the constraint of this system occurs when there is an excess
flow of information transmission that causes a node buffer overflow and packet loss
[31] as given in equation (1.16):
Analytics Approach for Intelligent Cyber-Physical System Integration 11

ζΨ ps = ∑ j∈∧ p − normal
α .K njδ (1.14)

ζΨ ps .d njδ
ζ njc = (1.15)
∑ j ∧ p − normal
d njδ

ζ njc ≤ (1 + σ c ).ζ 0c j (1.16)

where Ψ𝑝𝑠 is the CPS, 𝜙𝑛 is predictable of nodes, 𝜉𝑠 is the established node edge,
Λ𝑝−𝑛𝑜𝑟𝑚𝑎𝑙 is the standard process in the physical network environment, 𝜅𝛿 is the sen-
sor node degree, and j is the number of nodes in the network up-to-the-minute time
step n, while 𝜅 and 𝛿 are arbitrary variables in the control of the CPS. 𝜎𝑐 is the node
tolerance coefficient and 𝜁𝑐 is the load that is added to the sensor node j for the system
faulty identification.

1.2.9 Routing Protocols
A wireless sensor packet routing is relying on the network layers of the IIoT stack
model, while a routing algorithm is adopted to discover the best routes for packet
transmission and to ensure the safe delivery of data packets to the destinations in
the networks. The transport control protocol/Internet protocol (TCP/IP) is mostly
responsible to guarantee packet delivery and establishing a reliable communication
connection between two hosts in the network. It also repairs and maintains the routes
disruption when radio links (or hops) along established routes are broken. This prob-
lem can ensue due to sensor node relocation or node failure, server RF interference,
and congestion.
Also, the routing processes in the network are very challenging due to several
characteristics that distinguish them from contemporary communication and wire-
less ad-hoc networks as in Table 1.2. These include the building of global addressing
and routing algorithms for classical IP-based protocols. The application of sensor
nodes in the network is to sense data from multiple regions (sources) and route to
a particular sink or gateway (destination), which consumes energy and bandwidth
utilization. Therefore, the nodes that are tightly deployed in a constrained network
environment will have limited or constrained resources like power consumption,
processing capacity, and data storage [32]. The HEED and LEACH energy-aware
routing protocols that function as a multihop routing network algorithm using trans-
mission power of intercluster communication to select cluster head (CH) among the
sensor nodes in a network are expressed as in equation (1.17) [33]:

nprob = Cprob ∗ ε ri ⁄ε T (1.17)

where ƕprob is the probability of selecting a sensor node as a CH, ∁prob is the maxi-
mum number of clusters probability in the network, ε ri is the residual energy of an ith
node, and εT is the total energy of the network. However, routing is a critical concern
in the distributed sensor network when considering the homogeneity in the way pack-
ets are routed between source and destination across varying network topologies.
12 Intelligent Analytics for Industry 4.0 Applications

TABLE 1.2
Routing Algorithm and Its Functions
Protocols BSN Mobility Functions
LEACH 1 Stable BS The distributed cluster node is formed to extend
the node’s lifetime.
EWC 1 Stable BS Distributed cluster is formed to guarantee data
delivery and life improvement.
SPIN 1 Movable Exchange metadata to reduce the number of
messages and lifetime.
REAR >1 Restricted Lifetime extended and data delivery insured.
Energy aware >1 Restricted Lifetime extended.
Direct diffusion >1 Restricted It establishes efficient n-way communication paths
for fault tolerance.
Information driven 1 Restricted It saves more energy using optimization of direct
diffusion techniques.
Gradient 1 Restricted It delivers data through the minimal number of
hops.
PEGASIS 1 Stable base station It is lifetime with bandwidth optimization.
Energy-aware based 1 Not at all It is lifetime and operates in a real time.
cluster head
Self-organized 1 Movable It improves fault tolerance.
Minimum energy 1 Not at all It is lifetime and self-reconfiguration.
communication
Geographic adaptive >1 Restricted It increases the network lifetime as the number of
fidelity nodes increases.

But, the security mechanism for this activity (routing process) is reflected as more
problematic and interesting compared to others.

1.2.10 Security Challenges
The significant function of a secure mechanism in the IIoT network requires the
urgent attention of the experts due to billions of things and smart manufacturing
components that are expected to be flourishing in the universal network shortly. This
IIoT provides many conveniences and benefits to humans in the wide areas of smart
technologies. But the security challenge and resource constraint devices will make it
unsafe, and it may become an untrustworthy network [34]. The integration of ICPS
into the IIoT model is susceptible to different attacks due to weak security counter-
measures at each layer of the network, and it has become a serious challenge to the
existence of networks. Any provision to counter these security challenges with a
robust cryptographic method or advanced encryption algorithm may result in net-
work overhead. Other challenges include energy inefficiency, latency (delay), and
packet loss with throughput degradation because of their computation power, which
is not suitable for such constraint devices. So, any nodes that are compromised by
Analytics Approach for Intelligent Cyber-Physical System Integration 13

File inclusion attacks Directory traversal

HTTP splitting 0% SQL injection attacks attacks
0% Cross-site request
3% 3%
forgery
DoS Attack 16% 5%
Cross-site scripting
7%

Code execution 14%

Priviledge escalation
6%

Memory corruption
9%
Overflow
13%
Restriction Bypass
EXTENT OF ATTACKS (%) 11%
Information hijacking 13%

FIGURE 1.7  Analysis of securities attack measurement in the IIoT environment.

malicious actors, hackers, or social networks intruder can affect the performance of
the ICPSs in the network. The consequence of these malicious nodes in the network
is the misleading routing of sensor node traffic, generating false alarm reports, mes-
sage modification, false route identification, masquerading, and breaching the trust,
privacy, and confidentiality of the network information [35].
This significant security flaw in IIoT networks threatens the existence of ICPS
functionality, and it requires the urgent attention of experts due to billions of things
and machines that are expected to be flourishing in the Industry 4.0 network shortly.
Also, it is noted that smart devices are increasing in the market with several threats
that are replicated to destroy the existence of industrial revolution worldwide net-
works. These include bonnets, data spoofing, eavesdropping, ransomware, social
engineering, denial of service (DoS), and distributed denial of service (DDoS). The
extent of the cyber threat effect on smart technology devices is analyzed as shown
in Figure 1.7. This IIoT network threat is established in a distributed network during
the routing protocol initialization through an identified route from one ICPS node to
another until the destination is discovered.

1.3 THE INDUSTRIAL INTERNET OF THINGS

ARCHITECTURE DESIGN AND INDUSTRY 4.0
The IIoT design and Industry 4.0 technology include the integration of different
­facilities and services (such as embedded system (ES), network technology (NT), and
IT) for sensing, and control of physical objects by linking them to cyberspace network
[36]. Thus, these three technologies are very crucial to the functionality of Industry
4.0 architecture. Other expected technologies are robotics, AI, machine learning, big
data, digital computing, smart manufacturing, supply chain, business strategies, and
 14

Supply Chain Management Android Phone

Cloud
Web App
Server
Industry 4.0
Remote Users

Boarder Router 5G/6G IEEE802.15.4/IPv6

6LoWPAN/RPL
Cyber-physical
Ecosystem

6LoWPAN
Adaptation
Smart TV

Door Washing Printer

Security Machne Refrigerator Light switches
Smart Manufacturing Components

FIGURE 1.8  Intelligent cyber-physical system implementation in IIoT.

Intelligent Analytics for Industry 4.0 Applications
Analytics Approach for Intelligent Cyber-Physical System Integration 15

industrial products [37, 38]. The implementation of Industry 4.0 architecture in the
IIoT network illustrated in Figure 1.8 is based on the following technologies:

i. Embedded system can be a subset of IIoTs that provide service networks

such as a software package and firmware for industrial products. This
industrial product may include automobiles, electronics, industrial con-
trol systems, biometric systems, medicare devices, digital signal devices,
and robotics. The embedded system architecture may include a microchip,
gateway, sensor/mote, GSM module, and others to render functions like the
QoS, end-to-end communication, and interoperability.
ii. Network technology in this context supports end-to-end connectivity in the
IIoT network using wired technology or wireless protocols. The role of this
technology allows a seamless distribution of resources and application pro-
grams over the Internet. The network service is used for the building of
infrastructure that will relay information to the cloud database through an
information system, Internet, voice calls, and GSM/SMS.
iii. IT is a subsection of information and communication technology that
­supported the computing application, processing, storage, retrieving, and
handling of information over cyberspace. The IIoT information service
includes mobile web applications, programming, and a cloud database
system.

1.4  THE COMMUNICATION MODEL OF ICPS IN INDUSTRY 4.0

The communication model of ICPS depends on the protocol handler which utilized
TCP/IP for end-to-end communication and encapsulation of data from the local area
network to the remote interfaces. The communication protocol of low-level firmware
implementation is targeted to function over a wide area of network connectivity due
to the gateway connection between the cloud database and remote server using a
wireless access point (WAP) and wireless gateway support node (WGSN). In the
process, the dynamic session renegotiations (DSRs), forced session renegotiations
(FSRs) on the Internet gateway, and content management system server (CMSS) are
to significantly improve the sensor nodes’ connectivity over the Internet and reduce
packet loss rates among the sensor nodes.
The DSR functions in two ways: exchange packet transmission and verifying the
status of both gateway and CMS server channels (either in uplink or downlink sta-
tus). The DSR also verifies the appearance of socket failure or deadlock of a packet
sent between the clients and servers. If the data packets are unable to deliver or
received from a TCP/IP connection, the FSR procedure begins to check available
packets or data received from the gateway unit for a period of elapsed time. This
process will cause CMSS to wait for a new renegotiation or reconnection between
the client, server, and closed TCP/IP socket. But, if network service or packet is lost
during the transmission and communication setup between gateway and CMSS, the
TCP/IP will attempt a reconnection until connection establishment is s­uccessful.
The involvement of these different application layer protocols and programming lan-
guage design is used for communication between IoTs gateway and cloud database
16 Intelligent Analytics for Industry 4.0 Applications

0..* User 0..*

Subscribes/
Invoke Digital
Object Human
Interact
with
Active Passive Contains
0.* Digital Digital 0.*
Service Object Object 0..*
0.* Augmented
XOR
Entity 1.*
Contains
1 Contains
0.* 1.*
Exposes
0.1
0..* 0.1
Virtual
1. 0..*
Entity 1.*Contains Physical
0.* Attached 0.*
0.* 0.* to
0.* Entity
0.* 0.*
0.1 0.*
Resources Associated 0.*
with Device
0.*
Hosts 1

Network Connected
Resources Device Actuator Tag Sensor
Resources

FIGURE 1.9  The UML illustration of the cyber-physical system communication model in
Industry 4.0.

using java native language (JNL), Micro-C, MySQL, and many protocols like CoAPs,
MQQT, and HTTP. The UML of the CPS communication model in Industry 4.0 is
given in Figure 1.9.

1.5 SIMULATION RESULTS OF CYBER-PHYSICAL

COMMUNICATION MODEL IN IIOT ENVIRONMENT
The communication model of the CPS in the IoT architecture was developed and
simulated in Cooja-Contiki software. The CoAP web service with RESTful HTTPS
was adopted as an interoperable application-level protocol. These protocols focus on
end-to-end communication, and interoperability achievement among things (nodes)
connected in a network. Cooja-Contiki is an Internet of Things–supported simula-
tion tool supported with C-language for application program development through a
java application interface (JNI). It supports both large and small networks of sensor
motes to be simulated and implemented with objective functions and system settings
like generation of packets rate, media access control (MAC) protocols, throughput,
packet loss, and network topology. This CPS network design consists of four sen-
sor motes that run CoAP servers as resource-constrained protocols, and they are
sporadically queried by the network client using CoAP protocol. The GETs request
command is sent to the network server in the simulation setting to acquire the
parameter from the border router and help to connect the network with the Internet
Analytics Approach for Intelligent Cyber-Physical System Integration 17

using the Contiki-tunslip utility. The communication of these wireless sensor nodes
­simulation results shows the accuracy and efficiency of the packet transmitted from
the field (physical layer) to the cloud database system. The results of packet genera-
tion (packet/sec) and packet loss against throughput during packet transmission are
­presented in Tables 1.3 and 1.4, respectively. Figures 1.10 and 1.11 depict graphs of

TABLE 1.3
Simulation Results of Packet Rate Generation Against Throughput
Network Throughput (kb/s)
Sensor Sensor Sensor Sensor
PGR (packet/s) Node1 Node2 Node3 Node4
1 0.915 0.8757 0.7254 0.6505
5 1.2714 0.9932 0.7792 0.6566
10 1.9454 1.3552 1.1732 1.8463
15 2.122 1.4545 1.1921 1.2571
20 2.2748 1.9837 1.3534 1.5462

TABLE 1.4
Simulation Results of Packet Rate Generation Against Packet Loss
Packet Loss (%)
Sensor Sensor Sensor Sensor
PGR (packet/s) Node1 Node2 Node3 Node4
1 1.20 2.70 4.55 5.50
5 3.90 5.71 8.55 11.61
10 5.25 7.02 12.25 21.77
15 6.42 7.05 13.45 14.55
20 7.51 9.21 14.6 17.05

Node1 Node2 Node3 Node4

2.5
NETWORK THROUGHPUT (KB/S)

1.5

0.5

0
1 2 3 4 5
PACKET GENERATION RATE (PER SEC)

FIGURE 1.10  Network throughput against packet generates rate.

18 Intelligent Analytics for Industry 4.0 Applications

Node1 Node1 Node2

PACKET GENERATION RATE (PER SEC) Node1 Node3 Node1 Node4

60
50
40
30
20
10
0
1 2 3 4 5
PACKET LOSS (%)

FIGURE 1.11  Packet generates rate against packet loss.

FIGURE 1.12  Average packet delay in point to point.

Analytics Approach for Intelligent Cyber-Physical System Integration 19

FIGURE 1.13  Average period/packet transmits (bits/s).

packet generation and packet loss against throughput, respectively. Figures 1.12 and
1.13 show the average packet delay and the average period of packet transmission
during end-to-end communication nodes in the simulation environment.

1.6 CONCLUSION
This research analyzed the selected operational factors that affect the design of the
CPS and constraints of the IIoT model. The research survey exposed that energy
consumption, fault tolerance, reliability, and scalability affect the optimal perfor-
mance of the CPS-based IIoT. It also gives a clear and cohesive description of the
proposed design model together with the pertinent systemic components of the CPS.
The design requirements and technical procedures for integrating embedded devices,
sensors, actuators, and wireless connectivity for the enhancement of the IoT architec-
tural model are also elucidated in this work. The efficient design of a CPS depends
on hardware design, software coding, and control theory with the holistic approach
of computation to suit physical environmental constraints. The simulation of ICPS
20 Intelligent Analytics for Industry 4.0 Applications

communication performance was measured using packet generation rate, packet loss
rate, throughput, and packet delay. The future works will be focused on the devel-
opment of a lightweight secure framework for the CPS in Industry 4.0. Also, a big
data storage vulnerability to cyber attacks demanded an optimized approach for the
detection and mitigation of attacks in the IIoT environments.

REFERENCES
1. Ajao, L. A., Agajo, J., Kolo, J. G., Maliki, D., & Adegboye, M. A. (2017). Wireless
sensor networks based-internet of thing for agro-climatic parameters monitoring and
real-time data acquisition. Journal of Asian Scientific Research, 7(6), 240–252.
2. Ai, Y., Peng, M., & Zhang, K. (2018). Edge computing technologies for Internet of
Things: a primer. Digital Communications and Networks, 4(2), 77–86.
3. Verma, A., & Ranga, V. (2020). Security of RPL based 6LoWPAN Networks in the
Internet of Things: A review. IEEE Sensors Journal, 20(11), 5666–5690.
4. Pasikhani, A. M., Clark, J. A., Gope, P., & Alshahrani, A. (2021). Intrusion detec-
tion systems in RPL-based 6LoWPAN: A systematic literature review. IEEE Sensors
Journal, 21(11), 12940–12968.
5. Manavalan, E., & Jayakrishna, K. (2019). A review of Internet of Things (IoT) embed-
ded sustainable supply chain for industry 4.0 requirements. Computers & Industrial
Engineering, 127, 925–953.
6. Kumar, N. M., & Mallick, P. K. (2018). The Internet of Things: Insights into the building
blocks, component interactions, and architecture layers. Procedia Computer Science,
132, 109–117.
7. Cirani, S., Ferrari, G., Picone, M., & Veltri, L. (2018). Internet of Things: Architectures,
Protocols and Standards. John Wiley & Sons: Hoboken, NJ.
8. Yang, Z., Ping, S., Sun, H., & Aghvami, A. H. (2016). CRB-RPL: A receiver-based
routing protocol for communications in cognitive radio enabled smart grid. IEEE
Transactions on Vehicular Technology, 66(7), 5985–5994.
9. Tariq, M. A., Khan, M., Raza Khan, M. T., & Kim, D. (2020). Enhancements and chal-
lenges in coap—A survey. Sensors, 20(21), 6391.
10. Darby, S. J. (2018). Smart technology in the home: time for more clarity. Building
Research & Information, 46(1), 140–147.
11. Brougham, D., & Haar, J. (2018). Smart technology, artificial intelligence, robotics,
and algorithms (STARA): Employees’ perceptions of our future workplace. Journal of
Management & Organization, 24(2), 239–257.
12. Aliyu, S., Yusuf, A., Abdullahi, U., & Hafiz, M. (2017). Development of a low-cost
GSM-bluetooth home automation system. International Journal of Artificial Intelligent
and Application, 8(8), 41–50.
13. Jimoh, O. D., Ajao, L. A., Adeleke, O. O., & Kolo, S. S. (2020). A vehicle tracking
system using greedy forwarding algorithms for public transportation in urban arterial.
IEEE Access, 8, 191706–191725.
14. Cheng, J., Chen, W., Tao, F., & Lin, C. L. (2018). Industrial IoT in 5G environ-
ment towards smart manufacturing. Journal of Industrial Information Integration, 10,
10–19.
15. Goddard, N. D. R., Kemp, R. M. J., & Lane, R. (1997). An overview of smart t­ echnology.
Packaging Technology and Science: An International Journal, 10(3), 129–143.
16. Humayed, A., Lin, J., Li, F., & Luo, B. (2017). Cyber-physical systems security–A sur-
vey. IEEE Internet of Things Journal, 4(6), 1802–1831.
17. Boyes, H., Hallaq, B., Cunningham, J., & Watson, T. (2018). The industrial internet of
things (IIoT): An analysis framework. Computers in Industry, 101, 1–12.
Analytics Approach for Intelligent Cyber-Physical System Integration 21

18. Malik, P. K., Sharma, R., Singh, R., Gehlot, A., Satapathy, S. C., Alnumay, W. S., … &
Nayak, J. (2021). Industrial Internet of Things and its applications in industry 4.0: State
of the art. Computer Communications, 166, 125–139.
19. Bai, C., Dallasega, P., Orzes, G., & Sarkis, J. (2020). Industry 4.0 technologies assess-
ment: A sustainability perspective. International Journal of Production Economics,
229, 107776.
20. Serpanos, D., & Wolf, M. (2018). Industrial Internet of Things. In Internet-of-Things
(IoT) Systems (pp. 37–54). Springer: Cham.
21. Ghobakhloo, M. (2020). Industry 4.0, digitization, and opportunities for sustainability.
Journal of Cleaner Production, 252, 119869.
22. Dalenogare, L. S., Benitez, G. B., Ayala, N. F., & Frank, A. G. (2018). The expected
contribution of Industry 4.0 technologies for industrial performance. International
Journal of Production Economics, 204, 383–394.
23. Vaidya, S., Ambad, P., & Bhosle, S. (2018). Industry 4.0–A glimpse. Procedia
Manufacturing, 20, 233–238.
24. Xu, L. D., & Duan, L. (2019). Big data for cyber physical systems in industry 4.0: A
survey. Enterprise Information Systems, 13(2), 148–169.
25. Enrique, D. V., Druczkoski, J. C. M., Lima, T. M., & Charrua-Santos, F. (2021).
Advantages and difficulties of implementing Industry 4.0 technologies for labor flex-
ibility. Procedia Computer Science, 181, 347–352.
26. Castelo-Branco, I., Cruz-Jesus, F., & Oliveira, T. (2019). Assessing Industry 4.0 readiness
in manufacturing: Evidence for the European Union. Computers in Industry, 107, 22–32.
27. Lozano, C. V., & Vijayan, K. K. (2020). Literature review on cyber physical systems
design. Procedia Manufacturing, 45, 295–300.
28. Stavroulaki, V., Kelaidonis, D., Petsas, K., Moustakos, A., Vlacheas, P., Demestichas,
P., & Hashimoto, K. (2016). D2. 1 Foundations of Semantic Data Models and Tools, IoT
and Big Data Integration in Multi-Cloud Environments, University of Surrey, 1–115.
29. Ajao, L. A. Agajo, J., Mua’zu, M. B., & Schueller, J. (2020). A scheduling-based algo-
rithm for low energy consumption in smart agriculture precision monitoring system.
Agricultural Engineering International: CIGR Journal, 22(3), 103–117.
30. Ajao, L. A., Agajo, J., Gana, K. J., Ogbole, I. C., & Ataimo, E. E. (2017). Development
of a low power consumption smart embedded wireless sensor network for the ubiq-
uitous environmental monitoring using ZigBee module. ATBU Journal of Science,
Technology and Education, 5(1), 94–108.
31. Qu, Z., Xie, Q., Liu, Y., Li, Y., Wang, L., Xu, P., Zhou, Y., Sun, J., Xu, K., & Cui, M.
(2020). Power cyber-physical system risk area prediction using dependent Markov chain
and improved grey wolf optimization. IEEE Access, 8, 82844–82854.
32. Manzoor, B., Javaid, N., Rehman, O., Akbar, M., Nadeem, Q., Iqbal, A., & Ishfaq, M.
(2013). Q- LEACH: A new routing protocol for WSNs. Procedia Computer Science, 19,
926–931.
33. Touati, Y., Ali-Chérif, A., & Daachi, B. (2017). Routing Information for Energy
Management in WSNs. In Touati, Y., Ali-Cherif, A., & Daachi, B. (Eds), Energy
Management in Wireless Sensor Networks (pp.  23–51). Elsevier: Amsterdam, The
Netherlands.
34. Nigam, G. K., & Dabas, C. (2021). ESO-LEACH: PSO based energy efficient cluster-
ing in LEACH. Journal of King Saud University-Computer and Information Sciences,
33(8), 947–954.
35. Yu, X., & Guo, H. (2019, August). A survey on IIoT security. In 2019 IEEE VTS Asia
Pacific Wireless Communications Symposium (APWCS) (pp. 1–5). IEEE.
36. Jhanjhi, N. Z., Humayun, M., & Almuayqil, S. N. (2021). Cyber security and privacy
issues in industrial Internet of Things. Computer Systems Science and Engineering,
37(3), 361–380.
22 Intelligent Analytics for Industry 4.0 Applications

37. Ngabo, D., Wang, D., Iwendi, C., Anajemba, J. H., Ajao, L. A., & Biamba, C. (2021).
Blockchain-based security mechanism for the medical data at fog computing architec-
ture of internet of things. Electronics, 10(17), 2110.
38. Frank, A. G., Dalenogare, L. S., & Ayala, N. F. (2019). Industry 4.0 technologies:
Implementation patterns in manufacturing companies. International Journal of
Production Economics, 210, 15–26.
 2 Digital Twins A State of
the Art from Industry 4.0
Perspective
Ganesh A. Deiva and P. Kalpana
Indian Institute of Information Technology

CONTENTS
2.1 Introduction..................................................................................................... 23
2.2 Recent Studies on Digital Twin.......................................................................24
2.3 Components of Digital Twin........................................................................... 27
2.4 Applications of Digital Twin........................................................................... 29
2.4.1 Smart Manufacturing.......................................................................... 29
2.4.2 Supply Chain Management.................................................................. 29
2.4.3 Sustainability....................................................................................... 30
2.4.4 Intelligent Risk Management............................................................... 30
2.5 Challenges and Implementation Issues............................................................ 31
2.6 Future Research Avenues................................................................................ 32
2.7 Conclusion....................................................................................................... 33
References................................................................................................................. 33

2.1 INTRODUCTION
In the present data-centric environment, organizations are focusing on the transfor-
mation toward digitalization. It is leading to numerous modifications in production
processes, enhancing communication and cooperation. This kind of transformation
for creating responsive, flexible, and visible networks is known as digital transfor-
mation. This is the basis of the present industrial revolution that brings a paradigm
shift to manufacturing. For instance, Industry 4.0 builds on the foundation of the
Internet of Things (IoT) to enable monitoring the real-time systems, develop cyber-
physical frameworks, and offer the integration between the real and virtual world
[1]. It is more about intelligent, real-time communication, and self-adapting that
integrates all areas of manufacturing systems such as engineering, manufacturing,
logistics, and sales [2]. It aims to facilitate more efficient, higher quality, and faster
production. Given this, digital twin (DT) is a virtual model of a real-time system
based on a simulation, which optimizes the system for enhanced efficiency. This DT
theory has evolved from a conceptual model for product life cycle management to
an important tool for decision-making in this digital environment [3]. These models
offer opportunities for researchers and industry experts to achieve smart, sustainable
DOI: 10.1201/9781003321149-2 23
24 Intelligent Analytics for Industry 4.0 Applications

manufacturing systems, efficient supply chain management (SCM), and intelligent

mechanisms to predict, assess, and manage sudden disruptions. The global pandemic
COVID-19 has driven many organizations toward automated production processes
using IoT, faster decision-making against disruption, and accurate prediction in pro-
duction planning. The virtual design and optimization of intelligent systems are the
significant pillars for the successful transformation to Industry 4.0. Some of the sig-
nificant inabilities of the conventional industry practices are as follows [4]:

• The optimization of the product design outcome (gap between design and
manufacturing).
• The improvement of the production processes concerning the dynamic
environment.
• The prediction of unprecedented events across the various echelons of the
supply chain (SC).
• Real-time monitoring of the quality, performance, transportation, and the
information sharing.

These observations highlight the necessity to transform the internal structure in the
organizations to improve product quality, enhance process monitoring, and mitigate
uncertainty. In industries, DTs can be introduced at any stage in the production pro-
cess. They are capable of predicting and optimizing the product and product lines in
addition to the preproduction planning and design. Nowadays, they are exploited in
the modeling of the advanced SC that can assess strategies supporting SC optimiza-
tion. Recent studies highlighted the significance of developing digital SC models
that represent the actual SC digitally. The potential of DTs and the recently emerging
technologies demand further investigations to understand the challenges, applica-
tions, and under-explored avenues in Industry 4.0.
The residual part of this chapter provides the structure as follows: we present the
bibliometric analysis of the extant literature in the next section. Then, we illustrate
the significant components of DT architecture. Some notable applications are dis-
cussed in Section 2.4. The last section concludes the study.

2.2  RECENT STUDIES ON DIGITAL TWIN

This section illustrates the bibliometric analysis of recent studies on DT technolo-
gies. Bibliometric analysis is a publication statistical research method. Researchers
have exploited this method frequently to evaluate scientific publications systemati-
cally and identify the research gaps in a specific area.
The authors selected ‘Science direct’ as the data source. ‘Digital twin’ and
‘Industry 4.0’ were chosen as the keywords. We restricted the articles that are pub-
lished only in the year 2022 to capture the recent advancements concerning Industry
4.0. A total of 395 articles have been identified and exploited for this analysis. The
journals like Computers in Industry, Computers and Industrial Engineering, Journal
of Manufacturing Systems, Advanced Engineering Informatics, and International
Journal of Production Economics are some of the significant contributors. The
results were reexplored using the R tool and Biblioshiny software.
Digital Twins 25

Figure 2.1 illustrates the treemap chart according to the keywords of the recent
articles. This chart contains the frequency of the essential terms and keywords that
have been mentioned in the selected articles, such as Industry 4.0, DT, smart manu-
facturing, AI, sustainability, digitalization, predictive maintenance, IoT, blockchain,
Machine Learning (ML), simulation, and augmented reality. All major significant
terms related to DT components and applications have been covered in the selected
articles. The co-occurrence of the keywords network is shown in Figure 2.2. It can
be understood from the diagram that the majority of the DT and Industry 4.0 applica-
tions have been linked with ‘smart manufacturing’, ‘sustainability’, ‘predictive main-
tenance’, and ‘circular economy’. Based on the selected articles, clusters form in the
network of keywords. It indicates that the study has focused on investigating DT and
Industry 4.0. Similarly, the major components involved in DT technologies such as
IoT, AI, ML, blockchain, cyber-physical systems (CPSs), simulation, and big data
analytics have been highlighted in the network diagram.
Among the collection, the articles that are discussing concepts like digitalization,
sustainability, DT, and Industry 4.0 have been published more in number. The size
of the circle represents the density of the publications with respect to the key terms
(Figure 2.3).

FIGURE 2.1  Keywords treemap.

FIGURE 2.2  Co-occurrence network.

26 Intelligent Analytics for Industry 4.0 Applications

FIGURE 2.3  Keywords density map.

FIGURE 2.4  Conceptual structure map.

Digital Twins 27

The key terms that are associated with the selected studies are categorized into
several clusters to comprehend the significant avenues that influence DT in Industry
4.0. The conceptual structure map shown in Figure 2.4 is developed by multiple cor-
respondence analysis (MCA) using Biblioshiny software. The chart illustrates the
distribution of the key terms according to their position on the map. The terms with
the same dimension seem nearer on the map. Based on these, the authors consoli-
date the studies into two notable clusters such as DT applications and visualization
technologies.
As a result of the bibliometric analysis, the components and applications involved
in DT architecture are discussed in the following sections:

2.3  COMPONENTS OF DIGITAL TWIN

This section presents the notable Industry 4.0 technology components in the develop-
ment of DT architecture (Figure 2.5).

• IoT: The IoT for interconnected systems drives the significant interest in
adopting DT frameworks in organizations. Conventional manufacturing
systems are incapable of getting/accessing crucial data and dealing with
the large volume of data. Presently, the advancements in IoT allow indus-
tries to access, organize, and assess complex unstructured data technically
and economically. It drives DT in numerous industries by providing mass
production efficiency, mass customization, and mass personalization [5].
The extant studies defined the industrial Internet of things (IIoT) as a smart
infrastructure that integrates physical and virtual environments to fur-
nish advanced operations. For example, Zhao et al. developed a DT-based
architecture for tracking real-time information using IoT [6]. Most manu-
facturing industries have automated their processes using these IoT-based

FIGURE 2.5  The components of digital twin framework.

28 Intelligent Analytics for Industry 4.0 Applications

advancements, generating digital information. Recently, IoTwins have been

explored to implement cost-effective DTs with the support of cloud, ML,
edge, and IoT in Industry 4.0 settings [7].
• Artificial intelligence (AI): Advanced data analytics, one of the p­ rominent
AI techniques, has enhanced the quality of information extraction from
unstructured data [8]. Various AI/ML algorithms support DTs to identify
future uncertain events and make faster decision-making. Integrating AI
with manufacturing technology helps industry practitioners to develop
data-driven decision-making systems. Cavalcante et al. investigated the
incorporation of ML and simulation to develop DTs to aid in the develop-
ment of mitigation strategies related to supply disruption [9]. Today’s DT
models are primarily confined to understanding and rudimentary reasoning
of decisions for designing the smart manufacturing system. Integrating the
advanced AI algorithms into the DT models in the future could result in
achieving the potential to conceive and develop new design solutions auto-
matically. Given this, AI is integrated with IoT and CPS for illustrating the
virtual representation of DT in real time [10].
• Cloud computing: The productivity and efficiency can be enhanced by
enclosing manufacturing resources in an open environment like cloud
computing. It is applicable in various domains for developing business
architectures, and it is popular among researchers and industry experts.
The exploitation of DTs is a data-intensive process, which involves high
­computing and storage requirements. Therefore, the necessity of advanced
­systems for real-time information monitoring and the integration of cloud
into ­conventional manufacturing systems are significant challenges in
implementing DT systems [11]. Advanced data analytics plays a crucial role
in assessing the transfer of data to cyberspace.
• CPS: CPS is using the IoT to integrate disguised knowledge of production
systems. These systems have the potential to predict operational decisions
in smart manufacturing without the need for physical setup. However, the
important challenges in implementing CPS are real-time synchronization
and dynamic decision-making. Numerous CPS-based frameworks have
been addressed in the literature. Web-based frameworks for increasing
productivity, distributed production frameworks to adapt to the dynamic
environment, and stochastic production systems to attain real-time control
in making decisions are some of the significant applications. Therefore, the
existing studies highlight that DTs are the technical fundamental for devel-
oping CPS in Industry 4.0 [12].
• Blockchain: DT is a centralized architecture that lacks traceability, audit,
and trusted data province. Blockchain offers decentralized and collabora-
tive manufacturing processes. Therefore, incorporating IIoT with block-
chain can support a decentralized manufacturing network. The studies
addressed that the physical verification and mapping of manufacturer rela-
tionships can be overcome by exploiting blockchain and smart contracts
[13]. Also, they have been adopted to secure data sharing in fog-based IoT
systems [14]. These advancements are expected to track, and monitor the
Digital Twins 29

transactions, improve data security, and share efficiency in DTs. In Industry

4.0, ­numerous research has been done on blockchain-enabled smart manu-
facturing applications. However, practical concerns like scalability and
flexibility of systems have hampered the adoption of blockchain in Industry
4.0, and need to be addressed further [15].
• Augmented reality: The impact of DTs is quite powerful when the out-
puts are delivered in 3D. This requires the transformation of 2D systems
like computers and screens to high-end 3D display technologies such as
augmented reality. Virtual reality in the recent DT frameworks provides
productive integration between the available digital information and the
physical system. However, the lack of knowledge about power and train-
ing, and complex system incorporation are some of the notable barriers
to proposing augmented reality-based DT frameworks [16]. Scholars are
attempting to enhance design efficiency by incorporating DT systems with
augmented reality/virtual reality technologies.

2.4  APPLICATIONS OF DIGITAL TWIN

The above discussion reveals the essential components of a DT framework. This
­section discusses the notable applications of DT technologies.

2.4.1 Smart Manufacturing
In this data-centric world, the manufacturing process is transforming toward data-
centric and perception-enabled smart manufacturing. This involves the creation and
exploitation of data that addresses the integration of information technology and data
analytics. DT helps to achieve on-demand manufacturing services such as smart fac-
tory layout, scheduling, smart manufacturing, and human–robot collaboration by
acting as an enabler or direct assistance [17]. The literature describes how a DT can
be exploited to achieve the collaboration of humans and robots at a higher level [18].
Because of DTs, it is now possible to analyze human thinking and behavior in real
time using theoretical aspects [19]. Zhuang et al. presented a four-layer smart man-
ufacturing architecture for complex assembly lines [20]. It includes managing the
synchronization of real-time data, building an assembly floor DT, big data-enabled
predictive analysis, and exploiting DTs to assess services. It has been discovered
that adopting a DT in production processes can help save energy and resource [21].
Despite its impressive advances, the current DT model is still inadequate to satisfy
the demands of the accurate interconnection of the virtual model/data and physical
systems [22]. The development of feasible standards and proven architecture is a
critical step toward the success of DT-smart manufacturing.

2.4.2 Supply Chain Management

In SCM, new paradigms, concepts, and frameworks are being accelerated by digi-
tal technologies. The growth of digital SCs and intelligent process management
is aided by IoT, CPSs, and smart, connected devices. The introduction of DTs for
30 Intelligent Analytics for Industry 4.0 Applications

automating enterprises in the direction of SC integration is essential; however, it is

lacking. In addition, the lack of real-time insights and flexible designing/planning
systems makes DT adaptation problematic. According to recent studies designing
and developing a digital SC twin that digitally replicates the organization’s actual SC
is essential [23]. To achieve advanced SC modeling based on prescriptive analytics
that operates in real time and can decide the proper strategies supporting SC optimi-
zation, a digital SC twin is exploited. Since it relies on AI and simulation approaches
proposed by individual developers (vendors), it is impossible to copy or generate
them later using different developers. The DT’s promise in SCs as well as current
advances in new technologies like additive manufacturing, BDA, and IoT, neces-
sitates a systematic review of the extant literature to comprehend DT components,
applications, and challenges. As previously discussed, a DT depicts the present state
of SCs, by incorporating real-time transportation, inventory, demand, and capacity
information. For instance, if a strike occurs at an international logistics hub, a risk
monitoring tool can detect the disruption and send it to the simulation model as a dis-
ruptive event for further assessment. The simulation part in the DT can aid in predict-
ing and quantifying the risk propagation. Furthermore, it enables rapid testing and
adoption of recovery strategies and contingency measures. The studies addressed
that the investigations regarding the adoption of DT in production logistics are still
in a nascent stage [24].

2.4.3 Sustainability
The steady escalation of environmental problems, including carbon emissions and
nuclear pollution, has necessitated a transition in industries from traditional broad
economic growth to sustainable development in recent years. To assist product
design, manufacturing, efficient service, and sustainability enhancement, DTs incor-
porate physical, virtual, and connected product data. Given this, a simulation-based
DT framework contributes to sustainable development and circularity by assessing
the resource efficiency and recovery of high-quality secondary resources [25]. In
general, the role of DTs is important in delivering a data-rich runtime for achieving
sustainability. Also, concerning the evaluation of sustainable ideas, DTs based on
data fusion concepts can aid in reducing ambiguity and uncertainty. With the support
of ML algorithms, sustainability for a better life cycle can be addressed. For instance,
concerning smart city infrastructure, we can explore the noise levels in industries and
residential areas [26]. Similarly, the integration of ML and IoT has potential benefits
in building environmental monitoring systems [27]. The extant literature highlights
that the DTs are becoming the critical component of a product, liabilities, and indus-
trial operations in order to accomplish long-term goals.

2.4.4 Intelligent Risk Management

Predictive maintenance is a preventative strategy for predicting unprecedented
events well in advance of incompetence to avoid them. Because of its potential to
respond to uncertainty developing in real-time scenarios, DT appears to be more
beneficial than conventional approaches that had prior empirical data. Ivanov et al.
Digital Twins 31

proposed a decision-support system that incorporates data analytics, simulation, and

optimization [28]. The optimization allows proactive, resilient SC optimization, and
the simulation provides the assessment of SC dynamic behavior in the case of antici-
pated disruptions. Furthermore, this enables reactive, predictive modeling of disrup-
tion consequences on SC performance and of recovery strategies, which are then
optimized using an analytical model in a prescriptive way. On the other hand, the
system’s data analytics component is intended to identify disruptions in real time uti-
lizing process feedback data [29]. Given this, Deiva Ganesh and Kalpana developed
a conceptual DT-based framework to provide a holistic intelligent risk management
mechanism [30]. In general, SCs often are highly coordinated yet they are stressed
and lead to failures when an unprecedented event occurs. DTs can help simulate
these sudden rises and drops in supply, in order to meet the production demands and
make the whole process seamless. To achieve this, Big data analytics (BDA) and AI
provide an entirely new potential benefit to data-driven risk management. Due to its
dynamic tracking, this technique shows promise in minimizing unscheduled mainte-
nance and lowering excessive expenses. Predictive maintenance using DT extends its
application to aerospace industries, and healthcare processes also [31].

2.5  CHALLENGES AND IMPLEMENTATION ISSUES

This section highlights the significant challenges and issues related to DT imple-
mentation. However, as the interest in DT research is increasing rapidly, the studies
addressed the challenges found in terms of infrastructure, data availability, techno-
logical and security issues, expectations, trust, and modeling [32], as follows:

• Similar to AI, blockchain, and IoT, the next primary challenge is infrastruc-
ture development. Since data analytics, virtual reality tools, and IoT facilitate
the successful framework of a DT, it requires infrastructure that permits the
success of the aforementioned technologies. The absence of well-connected
smart networks, information sharing, and monitoring devices hampered the
implementation of these frameworks in most industries.
• The current level of data standards and structures is the next important
­challenge. DT requires quality data with constant, and continuous data
availability. Therefore, inspecting, handling, and storing such a huge vol-
ume of data need efficient data management technologies. Inconsistent and
noisy data may lead to the failure or poor performance of DTs in attaining
their objectives. Given this, an extensive analysis is needed for the right data
collection and its exploitation for achieving an effective outcome.
• In continuation, the data processing requirements will increase due to the
exponential growth of information growth. Hence, to provide a greater level
of processing speed, the data analytics algorithms like ML and deep learn-
ing need high-end processors. The lack of advanced technology adoption
can hinder the progress of introducing DTs in organizations.
• Concerning Industry 4.0, privacy and security issues are the important
threats with DT adoption. Because of the large volume of data, they are
vulnerable to cyber and hacker attack issues. The annual risk report (2021)
32 Intelligent Analytics for Industry 4.0 Applications

reported that the global pandemic increased the trust issues related to cyber
risks like data breaches to 73%. Therefore, emerging technologies like data
analytics and IoT for DT implementation should follow the standard, secu-
rity policies, updating practices, and ethical guidelines to overcome the
trust issues.
• Further investigation on DT technology is required to overcome the trust
issues from both industries and users’ perceptions. Verifying and validat-
ing the performance is crucial for reducing trust-related issues. Hence, this
enabling technology will provide valuable insights with more understand-
ing into the security and ethical-related steps.
• Despite the infrastructure, data processing, and privacy challenges, some
notable specific issues relating to the design and modeling of DT frame-
works also should be addressed. The current practices are lacking in stan-
dardized modeling and domain-specific modeling approaches. Following
standardized approaches in modeling can improve user understanding
while providing information flow across the different stages and implemen-
tation of a DT.

2.6  FUTURE RESEARCH AVENUES

An extensive future investigation is required to ensure the users and organizations
understand the ultimate benefits of a DT. In addition, the significant challenges that
are addressed need to be considered to attain the full potential of DT in various
domains. Some of the promising avenues for further research are as follows:

i. Though the literature addressed integrated and manufacturing DTs, they

are not realized in industries. There is a need for research in modeling DT
that exploits cyber-physical fusion.
ii. DT technology requires further research in fault diagnosis and predictive
maintenance for industrial processes.
iii. DTs for monitoring the health and well-being of humans through simulating
the impact of lifestyle changes.
iv. Prognostics and health management is one of the most promising areas with
DT through the integration of advanced data analytics.
v. Concerning data collection, processing, and analysis for DTs, methodolo-
gies should be explored to enhance accuracy. Data fusion, remote surgery,
and modeling for healthcare applications facilitate more opportunities for
future research.
vi. DTs for smart city development and traffic management systems are another
advanced open research area to be addressed. In such cases, exploring stan-
dardized modeling approaches is essential.
vii. In addition to manufacturing excellence, the scope of DTs for proactive
decision-making in SCM is increasing significantly to ensure SC operation
continuity.
viii. Deriving intelligent and holistic risk management mechanisms that com-
bine data analytics, simulation, and optimization is an emerging area to
minimize the impact of deep disruptions like COVID-19.
Digital Twins 33

ix. Developing high-fidelity dynamics models through AI-equipped DT to sup-

port complex robotic systems is under-explored in the extant literature.
x. Transitioning to high-end technologies such as augmented reality and vir-
tual reality can be an exciting avenue to enhance the interaction between the
physical and virtual environment.

2.7 CONCLUSION
In Industry 4.0, smart manufacturing has become an emerging direction for the global
manufacturing industry. This chapter surveys how the DTs are integrated into the
applications such as smart manufacturing, SCM, sustainable designs, and intelligent
risk management mechanisms. In addition, critical components such as IoT, CPSs,
virtual reality, data analytics, AI, blockchains, and cloud computing for ­supporting
the DT are analyzed. With ready access to them, most sectors have begun to exploit
DTs to manage their key assets. In addition, industry practitioners can use DTs to
anticipate operational issues, enhance product quality, and minimize downtime.
Currently, DT applications are primarily directed toward achieving manufacturing
excellence. Expanding the reach of DTs to cover all the things and SC practitioners
will enable them to make more proactive decisions. With increased awareness of the
economic benefits of adopting this technology in the next few years, it is expected to
serve as the backbone in the Industry 4.0 era in accelerating industrial transforma-
tion. Future work could include an assessment of DT technologies in the sectors of
renewable energy, smart cities and mobility, and healthcare.

REFERENCES
1. Kamble, S. S., Gunasekaran, A., Parekh, H., Mani, V., Belhadi, A., & Sharma, R.
(2022). Digital twin for sustainable manufacturing supply chains: Current trends, future
perspectives, and an implementation framework. Technological Forecasting and Social
Change, 176, 121448.
2. Zambrano, V., Mueller-Roemer, J., Sandberg, M., Talasila, P., Zanin, D., Larsen, P. G.,
… Stork, A. (2022). Industrial digitalization in the industry 4.0 era: Classification, reuse
and authoring of digital models on Digital Twin platforms. Array, 14, 100176.
3. Cinar, Z. M., Nuhu, A. A., Zeeshan, Q., & Korhan, O. (2019, September). Digital twins
for Industry 4.0: A review. In Global Joint Conference on Industrial Engineering and
Its Application Areas (pp. 193–203). Springer, Cham.
4. Jiang, Y., Yin, S., Li, K., Luo, H., & Kaynak, O. (2021). Industrial applications of digital
twins. Philosophical Transactions of the Royal Society A, 379(2207), 20200360.
5. Stavropoulos, P., & Mourtzis, D. (2022). Digital twins in industry 4.0. In Design and
Operation of Production Networks for Mass Personalization in the Era of Cloud
Technology (pp. 277–316). Elsevier, Amsterdam, The Netherlands.
6. Zhao, Z., Shen, L., Yang, C., Wu, W., Zhang, M., & Huang, G. Q. (2021). IoT and
digital twin enabled smart tracking for safety management. Computers & Operations
Research, 128, 105183.
7. Costantini, A., Di Modica, G., Ahouangonou, J. C., Duma, D. C., Martelli, B., Galletti,
M., … Cesini, D. (2022). IoTwins: Toward implementation of distributed digital twins
in Industry 4.0 settings. Computers, 11(5), 67.
8. Li, X., Cao, J., Liu, Z., & Luo, X. (2020). Sustainable business model based on digital
twin platform network: The inspiration from haier’s case study in China. Sustainability,
12(3), 936.
34 Intelligent Analytics for Industry 4.0 Applications

9. Cavalcante, I. M., Frazzon, E. M., Forcellini, F. A., & Ivanov, D. (2019). A supervised
machine learning approach to data-driven simulation of resilient supplier selection in
digital manufacturing. International Journal of Information Management, 49, 86–97.
10. Radanliev, P., De Roure, D., Nicolescu, R., Huth, M., & Santos, O. (2022). Digital twins:
Artificial intelligence and the IoT cyber-physical systems in industry 4.0. International
Journal of Intelligent Robotics and Applications, 6(1), 171–185.
11. Lu, Y., & Xu, X. (2019). Cloud-based manufacturing equipment and big data analyt-
ics to enable on-demand manufacturing services. Robotics and Computer-Integrated
Manufacturing, 57, 92–102.
12. Liu, C., Jiang, P., & Jiang, W. (2020). Web-based digital twin modeling and remote
control of cyber-physical production systems. Robotics and Computer-Integrated
Manufacturing, 64, 101956.
13. Reddy, K. R. K., Gunasekaran, A., Kalpana, P., Sreedharan, V. R., & Kumar, S. A.
(2021). Developing a blockchain framework for the automotive supply chain: A system-
atic review. Computers & Industrial Engineering, 157, 107334.
14. Mohapatra, D., Bhoi, S. K., Jena, K. K., Nayak, S. R., & Singh, A. (2022). A block-
chain security scheme to support fog-based internet of things. Microprocessors and
Microsystems, 89, 104455.
15. Leng, J., Wang, D., Shen, W., Li, X., Liu, Q., & Chen, X. (2021). Digital twins-based
smart manufacturing system design in Industry 4.0: A review. Journal of Manufacturing
Systems, 60, 119–137.
16. Kamble, S. S., Gunasekaran, A., & Gawankar, S. A. (2018). Sustainable Industry 4.0
framework: A systematic literature review identifying the current trends and future
perspectives. Process Safety and Environmental Protection, 117, 408–425.
17. Cimino, C., Negri, E., & Fumagalli, L. (2019). Review of digital twin applications in
manufacturing. Computers in Industry, 113, 103130.
18. Malik, A. A., Masood, T., & Bilberg, A. (2020). Virtual reality in manufacturing:
Immersive and collaborative artificial-reality in design of human-robot workspace.
International Journal of Computer Integrated Manufacturing, 33(1), 22–37.
19. Ardanza, A., Moreno, A., Segura, Á., de la Cruz, M., & Aguinaga, D. (2019). Sustainable
and flexible industrial human machine interfaces to support adaptable applications
in the Industry 4.0 paradigm. International Journal of Production Research, 57(12),
4045–4059.
20. Zhuang, C., Liu, J., & Xiong, H. (2018). Digital twin-based smart production man-
agement and control framework for the complex product assembly shop-floor. The
International Journal of Advanced Manufacturing Technology, 96(1), 1149–1163.
21. Kannan, K., & Arunachalam, N. (2019). A digital twin for grinding wheel: An infor-
mation sharing platform for sustainable grinding process. Journal of Manufacturing
Science and Engineering, 141(2), 1–14.
22. Cheng, J., Zhang, H., Tao, F., & Juang, C. F. (2020). DT-II: Digital twin enhanced
Industrial Internet reference framework towards smart manufacturing. Robotics and
Computer-Integrated Manufacturing, 62, 101881.
23. Simchenko, N. A., Tsohla, S. Y., & Chuvatkin, P. P. (2020). Effects of supply chain digi-
tal twins in the development of digital industry. International Journal of Supply Chain
Management, 9(3), 799–805.
24. Kaiblinger, A., & Woschank, M. (2022). State of the art and future directions of
­digital twins for production logistics: A systematic literature review. Applied Sciences,
12(2), 669.
25. Bartie, N. J., Cobos-Becerra, Y. L., Fröhling, M., Schlatmann, R., & Reuter, M. A.
(2021). The resources, exergetic and environmental footprint of the silicon photovol-
taic circular economy: Assessment and opportunities. Resources, Conservation &
Recycling, 169, 105516.
Digital Twins 35

26. Bhoi, S. K., Mallick, C., Mohanty, C. R., & Nayak, R. S. (2022). Analysis of noise
pollution during Dussehra festival in Bhubaneswar smart city in India: A study using
machine intelligence models. Applied Computational Intelligence and Soft Computing,
2022, Article ID 6095265.
27. Bhoi, S. K., Panda, S. K., Jena, K. K., Sahoo, K. S., Jhanjhi, N. Z., Masud, M., &
Aljahdali, S. (2022). IoT-EMS: An Internet of Things based environment monitor-
ing system in volunteer computing environment. Intelligent Automation and Soft
Computing, 32(3), 1493–1507.
28. Ivanov, D., Dolgui, A., Das, A., & Sokolov, B. (2019). Digital supply chain twins:
Managing the ripple effect, resilience, and disruption risks by data-driven optimiza-
tion,  simulation, and visibility. In Handbook of Ripple Effects in the Supply Chain
(pp. 309–332). Springer, Cham.
29. Ivanov, D., & Dolgui, A. (2021). A digital supply chain twin for managing the disrup-
tion risks and resilience in the era of Industry 4.0. Production Planning & Control,
32(9), 775.
30. Deiva Ganesh, A., & Kalpana, P. (2022). Future of artificial intelligence and its influ-
ence on supply chain risk management–A systematic review. Computers & Industrial
Engineering, 2022, 108206.
31. Liu, J., Zhao, P., Zhou, H., Liu, X., & Feng, F. (2019). Digital twin-driven machin-
ing process evaluation method. Computer Integrated Manufacturing Systems, 25,
1600–1610.
32. Fuller, A., Fan, Z., Day, C., & Barlow, C. (2020). Digital Twin: Enabling technologies,
challenges and open research. IEEE Access, 8, 108952–108971.
 3 Human-Centered
Approach to Intelligent
Analytics in Industry 4.0
Varad Vishwarupe
Amazon, University of Oxford, MIT Pune

Milind Pande and Vishal Pawar

MIT WPU Pune

Prachi Joshi
VIIT Pune

Aseem Deshmukh, Shweta Mhaisalkar,

Shrey Maheshwari, and Neha Mandora
MIT Pune

Nicole Mathias
Georgetown University

CONTENTS
3.1 Introduction...................................................................................................... 37
3.2  Related Work.................................................................................................... 38
3.3  Research Questions.......................................................................................... 41
3.4  Research Solutions...........................................................................................44
3.5  Discussion and Recommendations..................................................................46
3.6  Conclusion and Future Work........................................................................... 50
References................................................................................................................. 51

3.1 INTRODUCTION
Transformation of industrial machines, brought about by the collaboration of
­automation technology, machine learning, and communication abilities for the inde-
pendent running of an industry, is the concrete objective of the Industry 4.0 frame-
work. Industry 4.0 plays an important role in the manufacturing industry by ensuring
security, flexibility, customization, time efficiency, and a dynamic environment as
well as increased productivity and quality [1,2]. Through connecting smart devices
and machinery, employing self-learning solutions, and enhancing self-direction

DOI: 10.1201/9781003321149-3 37
38 Intelligent Analytics for Industry 4.0 Applications

capabilities, it is envisioned that the communication cost is reduced while flexibility

for manufacturing, mass customization capabilities, production speed, and quality
are increased [3–5].
Though the number of benefits of Industry 4.0 is tempting, Industry 4.0 also
reveals several challenges or in a more critical way obstacles to improving productiv-
ity at the workplace [6]. One such challenge is the work division between manpower
and machines, which allows us to explore Industry 4.0 catered human–computer
interaction (HCI) to reap the benefits of automated and smart machines, without
compromising on the strength of the current manpower. Work division would enable
a role change of humans to shift from low-level operations—which can be dangerous,
dirty, difficult, and dull tasks—to high expertise and safe tasks [7–10]. Moreover,
human intelligence and intervention remain a key role because of the safety, security,
social aspects, and uncertainties posed by such autonomous systems [11–14]. Even
with the dangerous roles out of the way of industry workers with the fourth industrial
revolution, a good design and HCI in place would be required to make humans adept
in the newly defined roles, so as to operate complex machines with negligible or no
error. Thus, this research work aims to review the current vision of HCI and UX
Design of Industry 4.0 and recommends practices best suited for the fourth industrial
revolution.

3.2  RELATED WORK

With the increasing stress on human satisfaction, ease and focus on better experi-
ences, there has been an emerging trend in research work related to human-centered
design and development of HCI from an Industry 4.0 perspective. Extracts from
such work are mentioned in this section, for obtaining a better picture of the overall
context. A good user experience (UX) and intuitive HCI is the cornerstone for the
smooth operation of any industry, and with Industry 4.0 mechanisms, these become
crucial. A system needs to be designed to generate positive UX for increasing user
association and encouragement [15–19] presented that bad user interface design can
lower user motivation to use a system. Beard-Gunter [20] worked with HCI design
in industries to develop and optimize engagement metrics compared to games. To
ensure the effectiveness of a system, a proper balance is required to be maintained
between usability and system functionality [21]. Considering design elements in
developing the industrial automation system can facilitate the meaningful interac-
tions between the system and the user [22,23].
Recent advances in artificial intelligence (AI) and HCI have seen remarkable
­contributions that have shaped the world. From being a virtual concept in scientific
fiction (sci-fi) a few years ago, AI is literally changing the way we act and interpret
the world around us. HCI and AI can be coined as two sides of the same coin accord-
ing to us and can be termed as entailing the same story in different ways or entailing
different parts of the same story, but ultimately working in unison for a common
goal. Ben Shneiderman, one of the pioneers in the field of HCI, has emphasized the
use of AI in collaboration with HCI researchers to further develop the field of human-
inspired AI [3,24].
Human-Centered Approach to Intelligent Analytics in Industry 4.0 39

There has also been a huge upheaval in the way HCI is used in the m ­ anufacturing
industry to optimize for quality and not cost. The work of Uzor et al. [25] is particu-
larly intriguing in this regard wherein they have tweaked a very important facet of
Bayesian optimization, the activation function, and a particularly important statisti-
cal machine learning method to improve user interface design using crowdsourcing.
The use of virtual reality (VR) and mixed reality (MR) as a driving force for the
domain of HCI has been on the rise too [5,6]. Using intuitive user interfaces (IUIs),
VR and MR gadgets, and head-mounted displays has also made it possible to inter-
act with the physical medium in a way that is more immersive. Simulation of touch
displays for people with motor disabilities, gesture typing, 3D-based VR models, and
gaze-based motion tracking gadgets are also some of the advancements that have
helped in fusing the two realms of AI and HCI, by developing smart and interactive
systems [7–11].
The use of AI-based HCI systems and HCI-oriented AI systems has also been
observed, especially in the medicine sector. AI-based diagnosis of difficult-to-­
identify diseases and in areas where human intervention seems to have stuck at a
dead end enables clinicians to find out ways which are improving healthcare such as
Sio and Hoven [14], Weichhart et al. [15] works on gender classification in task func-
tional brain networks. Use of AI–HCI conjunct systems in improving language mod-
els for natural languages also depicts the pervasive nature of the said field. Isen [16]
Islam et al. [17] have done some remarkable work in this regard. Inferring web page
relevance, using HCI models in certain Internet of Things (IoT) tasks, generating
personalized recommendations for certain users in the browser, and also using few
facets of HCI in the development of smart set-top box TV recommendation systems
have enabled the development of smart AI-based systems [12,18–22,26,27].
While there has been a substantial work at the crossroads of AI and HCI, it is still
not extensive enough to be able to use HCI and AI in scenarios wherein it becomes
difficult to have experts from both fields working together such as cyber-physical
systems [24,28–30]. Thus, it is important that the stakeholders from respective fields
are shown what, when, why, and how they can contribute, what the major roles of
each discipline and its experts are and how can AI be used as a catalyst for develop-
ing HCI systems and vice versa [26,27]. In the context of this paper, we try to gather
insights from the previous works in the aforementioned and present our research that
shall help develop this exciting field of knowledge.
When it comes to HCI in Industry 4.0, we need to understand the prior work in
the context of a plethora of subfields, viz. human–machine interaction (HMI), vir-
tual, augmented, and MR applications in the context of cyber-physical systems, and
surveys which have been conducted keeping HCI and HMI under the purview on
Industry 4.0 as a research topic [31,32]. A large number of surveys covering peculiar
aspects of HMI or human factors in Industry 4.0 has been identified in the endeavor
of collecting relevant literature for this paper. The vast majority of these studies are
specifically oriented toward either VR or augmented reality (AR) applications, or
both, within Industry 4.0 operations and, thus, cover only a subset of this paper’s
scope [33,34]. Büttner et al. [35] conduct a survey on AR and VR applications in
Industry 4.0 manufacturing activities, more precisely on the available platform
40 Intelligent Analytics for Industry 4.0 Applications

technologies and application areas, creating a small-scale design space for such MR
applications in manufacturing. This design space differentiates among four general
application scenarios and four types of MR technology platforms available for appli-
cation, respectively.
The application scenarios are manufacturing, logistics, maintenance, or t­raining
while the available platforms comprise mobile devices (AR), projection (AR), and
head-mounted displays (HMDs) (AR or VR). Dini and Dalle Mura [36] present sur-
veys on AR applications, however not restricted to Industry 4.0-related application
scenarios. They investigated general commercial AR applications including indus-
trial scenarios also, among others, civil engineering. The specific AR application sce-
narios which they examine based on related scientific literature are maintenance and
repairing, inspection and diagnostics, training, safety, and machine setup. Nahavandi
[8], in turn, examine scientific research on AR applications for assembly purposes
from a time span of 26 years starting as early as 1990 and concentrating mainly on
the period from 2005 until 2015. Thus, they extend the scope to many years before
the advent of Industry 4.0-related initiatives and ideas. The major application pur-
poses of AR in assembly tasks that they investigate are assembly guidance, assembly
design and planning, and assembly training [37,38].
Choi et al. [39] provide a literature review on AR applications with an even less
specific focus on industrial deployment by examining the state of the art at the time
in research on collaboration in AR considering a wide range of possible application
fields, the industrial sector being only one of those. As a result, they identify remain-
ing research challenges relating to collaboration in AR which are the identification of
suitable application scenarios and interaction paradigms as well as an enhancement
of the perceived presence and situational awareness of remote users.
Palmarini et al. [40], in turn, conduct a structured literature review on different
software development platforms and types of data visualization and hardware avail-
able for AR applications in various maintenance scenarios The aim and purpose of
their study is to derive a generic guideline facilitating a firm’s selection process of
the appropriate type and design of AR application, tailored to the specific type of
maintenance activity at hand which the firm is planning to enhance utilizing AR
technology. Lastly, Choi et al. [39] provide surveys on VR technology in an industrial
environment, the latter group of authors concentrating on a potential combination
of VR technology with discrete event simulation for scenario testing in Industry 4.0
activities. Hermann et al. [41] on the other hand, present a survey on VR applications
in manufacturing, concentrating on potential contributions of VR deployment in the
development process for new products and deriving a mapping of different types
of VR technology toward the different steps of the product development process.
Therefore, Dini and Mura [36] consider the applicability of various VR technologies
for the phases of concept development, system-level design, design of details, test-
ing and refinement, and launch of production. Besides those surveys on AR and VR
applications, a literature review by Hecklau et al. [42] exists on the major challenges
as well as skills and competencies needed for future employees under an Industry
4.0 scenario. The authors utilize the insights from the literature analysis to struc-
ture the required skills according to different categories, based on which a com-
petence model is created analyzing employees’ levels of skills and competencies,
Human-Centered Approach to Intelligent Analytics in Industry 4.0 41

which will be particularly important in an Industry 4.0 working environment. The

main c­ ategories for the competence model are technical, methodological, social, and
personal competencies.
Uzor et al. [25], Palmarini et al. [40], Zahoor et al. [43,44] and Vishwarupe et al.
[45,46] also present a very lucid case study of using Amazon Mechanical Turk and
crowdsourcing platforms for facilitating HCI studies in the context of Industry 4.0
applications, wherein the importance of having a synergy between the user interface
for crowdsourcing based studies and industrial requirements is enunciated. While we
have tried to cover the maximum bases for our work related to HCI and HMI in the
context of Industry 4.0, it is important to highlight that this list of related work is not
all-encompassing. Instead, it is rather supposed to provide an outlook of other exist-
ing scientific literature regarding the topic of HCI and Industry 4.0 and thereby reaf-
firm that, to the best of the author’s knowledge, a study matching of this magnitude
ceases to exist. We have strived to provide a very detailed bird’s eye view of the state
of the art in HCI and Industry 4.0 issues, which intersect at the crossroads of both
disciplines. This aspect facilitates the justification of this key research and signifies
the necessary importance as well as the relevance of this research paper. Though
UX and human-centered design have been explored in different contexts, there is a
dearth of studies and research on its consideration, use, and impact on an Industry 4.0
setup. This chapter focuses on the exploration of human-centered design principles
and customized HCI to tend to the complexities involved in Industry 4.0 and the role
of humans in such an ecosystem.

3.3  RESEARCH QUESTIONS

At the outset, we need to base our study on the fundamental understanding of what
HCI is and how it is relevant under the purview of Industry 4.0. In its entirety, HCI is
the study of the interaction between humans and computers, particularly as it pertains
to the design of technology. HCI is at the crossroads of user-centered design, user
interface (UI), and UX to create intuitive products and technologies. Researchers
who specialize in HCI think about conceptualizing and implementing systems that
satisfy human users. HCI also helps to make interfaces that increase productivity,
enhance UX, and reduce risks in safety-critical systems. This is especially relevant
in heavy industrial applications of Industry 4.0 such as manufacturing and cyber-
physical systems. Therefore, HCI is on the rise for developing intelligent interac-
tive systems. While defining the research questions for this paper, it is important to
identify the key principles of interaction. The gold standard for this is the Norman’s
model of interaction. Norman’s interaction model is a noteworthy and pioneering
model for HCI-based studies. It proposes that a user first establishes a goal and then
performs actions using the system to achieve that goal. Thereafter, the system pro-
vides the result of the user actions on the UI. A particular user then minutely assesses
this result and sees if their objective has been achieved or not. If not, a new goal is
established, and the cycle is repeated. This model of interaction explained is divided
into eight primary stages (Figure 3.1):
This model helps us to understand where things go haywire in our designs. Under
the context of Industry 4.0, there are issues when it comes to the sections of machine
42 Intelligent Analytics for Industry 4.0 Applications

FIGURE 3.1  Norman model of interaction.

execution and problem evaluation by the cyber-physical systems. Machines perceive

the world in binaries, as opposed to sensory inputs that are so readily perceived by
humans. There are also issues with the interpretation of perception which further
causes issues with the evaluation of those interpretations. There is a measurable dif-
ference between user actions and those that a particular system can perform. A potent
and engaging interface allows a user to perform an action without system limitations.
This is especially important for Industry 4.0 which is nothing but an interconnected
web of system of subsystems consisting of IoT and cyber-physical systems. The sec-
ond consideration wherein the importance of having a proper interaction mechanism
between humans and machines for cyber-physical systems is the difference between
the presentation of an output and the user’s expectations. An effective interface can
be easily evaluated by a user, but it fails to evaluate when only machines are involved
in the process. As a preliminary step, few research questions were chalked out to
define the basis of this research.
The focus of these research questions is to primarily study the current visions of
an Industry 4.0 ecosystem and especially explore this industrial revolution from the
perspective of humans. Answering the formulated research questions would provide
an overarching view of the human-centered approach of building the next indus-
trial revolution. For answering the research questions, domain-specific keywords
from each research question were used to explore the current research work through
Google Scholar. Google Scholar was preferred as the source of all research con-
tent, as it gave an expansive search result and encompassed research work published
from a wide range of publication organizations, thus covering a broad spectrum
of resources. Research material from a diverse set of publication houses—IEEE,
Research Gate, Elsevier, Science Direct, and MDPI—were studied and analyzed.
Providing better user support whenever there is a tangible difference expected output
Human-Centered Approach to Intelligent Analytics in Industry 4.0 43

and the output generated by the user can be a vital factor in deciding the end goal
of an efficient cyber-physical system. Thus, we pose the following important three
aspects and five questions and try to answer them with our study on the principles of
HCI under the context of intelligent analytics for Industry 4.0:

• The UI and UX design paradigm:

What facets of the UI–UX components of the HCI be envisioned from an
Industry 4.0 perspective?
• The interaction paradigm:
How should the interaction mechanism between humans and machines
be shaped under the context of Industry 4.0?
• The manpower–machinery paradigm:
How will manpower be trained to interact with, supervise, and maintain
complex machinery of Industry 4.0?
1. Why is there a need to involve humans in the process of developing this
Industry 4.0–HCI confluence?
Naturally, it is a human tendency to question things and have a rea-
soning that is both acceptable as well as viable when it comes to per-
ceiving the world around us. The question of why to involve humans in
the Industry 4.0–HCI confluence is vital because autonomous intelli-
gent systems can falter at times, where only human intervention works.
2. Which stakeholders need to be present in the development of the
Industry 4.0–HCI confluence?
Traditionally, AI has been looked upon as a branch of mathemat-
ics and computer science, which tries to mimic the reasoning abilities
of the human brain. However, with the pervasive nature of AI and its
encroachment on our day-to-day life, it is essential to broaden the scope
of involving different stakeholders in this major confluence.
3. When to trust the AI in Industry 4.0 and when to trust humans, in the
decision-making at the Industry 4.0–HCI confluence?
Gradually, with the advent of large amount of data that is available,
issues pertaining to trust, ethics, privacy, and security of user’s data
have been on the rise. It is important to be able to distinguish when to
trust the AI and when not to, and rely on human judgment instead. This
research question delves into the ethical side of AI and HCI.
4. Where can the users work in tandem with the AI/where they cannot,
pertaining to the confluence?
Eventually, the choice of where to use AI-enabled systems and where
to rely on human acumen is an important part of deploying AI and HCI
systems. The roles and responsibilities of both parties should be aptly
defined and coherently distinguished. This research question tries to
draw a lucid distinction between these two.
5. How much, is too much when it comes to the Industry 4.0–HCI
confluence?
Incidentally, as they say, everything is good only in modera-
tion. So, in this section we try to answer an important prerogative on
44 Intelligent Analytics for Industry 4.0 Applications

what should be the extent of the Industry 4.0–HCI confluence and

what ­regularization should be enforced at their crossroads, so that the
Industry 4.0–HCI ecosystem works in tandem with humans and does
not supersede them.

3.4  RESEARCH SOLUTIONS

Complex machinery of the fourth industrial revolution can be quite overwhelming
to work with, and with massive chunks of information with respect to each machine,
and the interconnection of machines, obtaining the right information at the right time
can be quite difficult. Information management and control over such information
then become critical to avoid chaos. A carefully designed and well-placed HCI would
help to steer clear of such a situation. The meticulous design and information display
of an HCI system can be achieved by incorporating design process elements such as
qualitative and quantitative analysis, and information architecture. We further stress
that this inclusion of AR and VR, would come with a requirement for training to
be given to employees working with Industry 4.0 machines, in which case the user
interface should be intuitive, and the UX should be a good one. At the same time,
employing technologies like AR/VR comes with its very own list of disadvantages.
AR/VR headsets are difficult to wear for a long duration of time and would cause
an uncomfortable experience. We recommend for implementing the AR/VR cor-
rectly into the Industry 4.0 ecosystem, users need to be considered from the start
in the design process of these interactions, to understand them and unearth their
concerns and pain points through qualitative analysis such as interviews. These con-
cerns should be further validated with a quantitative study such as surveys to realize
a hypothesis. This will enable to gauge the value of introducing AR and VR at the
initial level, and will also help in making modifications along the way, according to
the best-fit scenario with respect to industry personnel.
To answer the aforementioned questions on what recommendations or solutions
can be conceived to have a seamless UI–UX component and interaction mechanism
and how the manpower–machinery paradigm should work in the HCI, which is spe-
cifically developed for and from an Industry 4.0 perspective, we define the following
six components of the HCI centered on the Industry 4.0 realm:
As is evident from Figure 3.2, we envision that the UI–UX that is designed keep-
ing industrial recommendations in mind should involve a subset of six core com-
ponents. AR–VR-based HMDs are the way to go ahead when it comes to industrial
supervision roles. It would lessen the need to physically visit the site for inspec-
tion and avoid unnecessary hazards. The second component should be the use of
Explainable AI models which not only predict and classify but also explain the
rationale behind arriving at a specific decision. This is particularly important in the
context of Industry 4.0 since there is a significant level of reliance on machines to
perform seamlessly day in and out. Thus, knowing why a particular machine took
a particular decision at a particular time helps in reasoning and accounts for outlier
detections, when the mechanism or machine fails.
Third component of this HCI focuses on the need for a user-feedback net-
work which is pivotal to the smooth functioning of a proper HCI mechanism in
Human-Centered Approach to Intelligent Analytics in Industry 4.0 45

Gamificaon

Synergy in
Head Mounted
Systems of Sub-
Displays
Systems

HCI in Industry 4.0

Human-Machine
Explainable AI
Failure
Dashboards
Intervenon

User Feedback
Network

FIGURE 3.2  Components of HCI centered on Industry 4.0.

cyber-physical systems. Users should receive both instantaneous and cumulative

feedback for examining fault tolerance and industry-specific bottlenecks in the
manufacturing processes, which are often the first deterrents on the assembly line,
wherein the entire process comes to a standstill.
Fourth consideration includes the ability to have a combination of HMI when there
is a critical failure. Often, there are problems that are only identifiable by human
intervention such as system breakdown, actual physical issues with the machinery,
leaks, and overflows. During such catastrophic situations, it is important to have a
synergy between how much human vs machine intervention is needed for the failure
to be resolved in the least possible time, with minimal resources at hand.
The final component of a successful HCI system tailored for the Industry 4.0
framework should include the amalgamation of the system of subsystems which is
also known as IoT. In the IoT framework, device, network, and application are the
three most important components, which make up the internals in a manufacturing
facility. These systems should communicate with each other on a network that is
46 Intelligent Analytics for Industry 4.0 Applications

separated from the main hub, to accommodate the passage and relay of c­ ommands
and messages in a brisk and seamless manner. Whenever a certain subsystem
­malfunctions, there should be immediate relaying of the same to the interconnected
nodes with a modality to quickly intimate the nearest operator, wherein the other
systems should come to a stop and avoid dire consequences such as fire and loss of
property. Having these six components in an industrial HCI system would enable
that there are low downtimes for machinery and would establish a cohesive synergy
between humans and machines, through an actionable intuitive mechanism.

3.5  DISCUSSION AND RECOMMENDATIONS

Current modalities and practices of training can become quite mundane and over-
whelming for Industry 4.0-specific training. With high-technology procedures to
complete tasks on smart machines, the traditional training method will overburden
the manpower, and would not be effective in retaining operating procedures. With
AR and VR technologies to be used in Industry 4.0 ecosystem, the training provided
will be highly engaging and playful, and due to the visual aspects of these technolo-
gies, the training imparted would be retained by industry professionals. It could iden-
tify the impact of learning with AR in comparison to traditional vocational training
approaches.
Similarly, gamification is another modality that could be used in Industry 4.0-spe-
cific HCI to train and assist industrial manpower. Gamification is the method of
introducing game-like elements into a nongame environment, to incentivize users to
do an otherwise mundane task. However, gamification, if not done right, can lead to
an even more unpleasant UX than to start with. Therefore, the context of use should
be properly defined, and on that front, the concerned manpower should be evaluated
in terms of their interest level in game elements. Moreover, stakeholders must be
involved in taking an appropriate decision on the type and extent of gamification
to be involved in the industry, to benefit the industry as well as personnel mutually
(Figure 3.3).
In our humble opinion, an “ideal” Industry 4.0–HCI system should feature a bal-
anced scale of humans and machines working together, with a positive collaboration
between them. To achieve such a synergy, humans should not feel excluded from the
interconnected smart machines of the industry. Apart from a rational work division
between the two entities, for humans to be inclusive in Industry 4.0, high-technology
modalities should be used to foster an engaging and exciting experience of co-work-
ing between humans and machines. As discussed previously, the intent of using these
high-technology modalities, whether they are used to provide information across
the industry, or as an assistive technology to complete tasks successfully, needs to
be defined clearly and a design structure based on the intent is mandatory to be
constructed to gain a first-hand view of the space and function of such technological
implementation. At the same time, stakeholders and users must be involved in the
design process of Industry 4.0–HCI, to take care of the business requirements and
the perspective of users, respectively. For an intelligent Industry 4.0 framework to
effectively function with the collaboration of humans and machines, a human-cen-
tered HCI design process must be established. This would require a customization
Human-Centered Approach to Intelligent Analytics in Industry 4.0 47

FIGURE 3.3  Industry 4.0–HCI confluence framework for fostering intelligent analytics.

from the traditional HCI design process to fit the characteristics and operational
mechanism of the Industry 4.0. Our recommendation is to personalize the process
of gathering and understanding requirements, designing solutions based on require-
ments, testing, and evaluating, in the context of Industry 4.0. The below personal-
ization will cater to the efficient growth of Industry 4.0, all the while maintaining a
strong relationship and codependency between humans and machines. To do this we
suggest involving domain experts and stakeholders in the user requirement part of
the cycle. This will enable the stakeholder groups to minutely specify requirements
at the outset, thereby honing the process.
Thereafter, using Explainable Artificial Intelligence (XAI)-based modeling using
SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic
Explanations (LIME) model-agnostic frameworks shall help eliminate the black box
nature of AI-based solutions and help in bringing fairness and accountability to the
process. In the next step, we believe that it is important to not only evaluate the
solutions with initial user requirements but also perform usability testing involving
stakeholder requirements. This shall help in developing a synergy between users and
stakeholders, which shall benefit in the long run by removing the actual constraints
faced by users and alleviating them by involving the knowledge of stakeholders.
In the last leg of the Industry 4.0–HCI design process cycle, it is important to incor-
porate user feedback and not only explain them the mechanisms of how the system
works but also why it arrived at a certain decision. If there is a large gap between
the user-expected output, the stakeholder-predicted output, and the actual output,
it is important to reiterate the process by removing critical bottlenecks. This can
48 Intelligent Analytics for Industry 4.0 Applications

range from increasing the data set size by conducting more expansive user surveys
and user behavior studies, or it can involve changing the train–validate–test dataset
sample sizes and percentages, to counter for overfitting, especially when deep learn-
ing ­models are involved, which are inherently more black boxed in nature. Thus,
there needs to be a consensus between AI and HCI practitioners on when to draw
and derive from their respective fields and when to pause and introspect. The itera-
tive nature of both HCI design processes and AI–ML models make it easier to duly
bifurcate and then combine certain tasks with each another. Using the Industry 4.0–
HCI confluence framework along with the modifications in the HCI user-centered
design process, can help in making the merger of AI and HCI systems transparent,
seamless, and cohesive. Thus, after developing the above framework and recom-
mendations for tweaking the user-centered HCI design process, we believe that we
can answer the questions posed above. For the sake of reading convenience, we have
included the questions before the respective answers using bullet points. They are as
follows:

• Why is there a need to involve humans in the process of developing this

Industry 4.0–HCI confluence?
The involvement of users in developing the Industry 4.0–HCI confluence
is of utmost importance because without involving domain experts across a
wide variety of fields and without involving a randomized, a­ ll-encompassing
group of users, the solutions would be a black box in nature, like traditional
AI models and would severe the purpose of developing human-centered
intelligent interactive systems, which depict at least some form and type of
intelligence.
• Which stakeholders need to be present in the development of the Industry
4.0–HCI confluence?
The type and kind of stakeholders which shall be involved in the merger
of these two disciplines shall involve AI researchers, factory workers, data
scientists, social scientists, philosophers, mathematicians, and other domain
experts as per the application area of the respective domain. This involve-
ment is essential to the conducive growth of the field.
• When to trust the AI and when to trust humans, in the decision-making at
the Industry 4.0–HCI confluence?
While it is imperative that users feel empowered when it comes to deci-
sion-making, it is equally important for the AI systems to disempower them
in a way that they feel less in control of certain processes, which can pose
hazards to them. As a very crude and general guideline, if the AI can come
to a correct decision repeatedly, with excellent fault tolerance, precision–
recall values as well as confusion matrices, the users can learn to trust the
AI, especially if Explainable AI toolkits are used to justify and interpret the
outputs of the system. All in all, AI systems should be entrusted with tasks
and decisions that need huge and fast computation, and for others, humans
can be more in control.
• Where can the users work in tandem with the AI/where they cannot, per-
taining to the confluence?
Human-Centered Approach to Intelligent Analytics in Industry 4.0 49

Users can work in tandem with AI in the initial design and evaluation
phase of the Industry 4.0–HCI confluence and possibly cannot work in
certain situations where there is a considerable gap between the respective
operational capacities.
• How much, is too much, when it comes to the Industry 4.0–HCI confluence?
AI and HCI systems need to be regularized, made transparent, and
trusted with sensitive data that should be kept safe and away from potential
attacks. They should also work in tandem with, and for humans, by not
overpowering them (Figure 3.4).

In the five-step process represented above, the first and crucial step of designing
and setting up Industry 4.0 is to study and analyze the industrial manpower. This
is important to keep their experience and concerns in mind while designing intel-
ligent systems and their interfaces around them. The next step involves stating
the user requirements analyzed from the qualitative and quantitative study of the
concerned manpower and involving every stakeholder to ensure full acceptance of
user requirements. Although the above representation only lists stakeholder buy in

FIGURE 3.4  Intelligent analytical HCI design process for Industry 4.0.
50 Intelligent Analytics for Industry 4.0 Applications

the second stage, it is a practice that should be carried out at each stage. Once the
­requirements are zeroed upon, prototyping with respect to the complexity of Industry
4.0 ­machinery and ease of use for manpower should be done. From an Industry 4.0
perspective, it becomes important to make use of a good information architecture
and easy-to-use interfaces to operate complex machines. Usability testing on such
prototypes reveals inconsistencies and confusion with respect to the usage of inter-
faces, on the realization of which, iterations of the prototypes are constructed, and
the process is repeated until a good HCI interface is made.

3.6  CONCLUSION AND FUTURE WORK

The development of a HCI system specifically made for the Industry 4.0 environ-
ment could be made a reality focusing on the three paradigms discussed in this
chapter. It can be a cause of concern for currently working industrial manpower,
as it will automate industries extensively. While there are immediate ramifications
of this on the numbers of jobs that would be available in the production industry, it
does pave the way for developing better industrial practices which are focused on
creating a synergy between humans and machines. The first industrial revolution
started with the advent of the steam engine, second industrial revolution brought
electricity to the world, the third and ongoing revolution brought automation as a
boon for the industry and imminently, the fourth industrial revolution would usher
a new era in the world, where previously imaginable phenomena would take actual
shape. Such a revolution is inevitable, and industries will lose out on competition
if they steer clear from Industry 4.0 modifications. But, with a well-designed HCI
system in place, not only will the industry workers shift to better roles and respon-
sibilities, but also be able to learn new skills, and will experience an engaging job
with the help of well-designed assistive technologies to aid them in supervising and
maintenance of complex machinery. The aim of an HCI focused on Industry 4.0 is
to enable humans and machines to work in tandem with one another and not replace
each other, thereby making the Industry 4.0 ecosystem more efficient, transparent,
and productive. Future work in this realm would include more case studies that are
pertaining to the actual implementation of such cyber-physical systems, especially
in the production industry. This would include working with stakeholders right
from the shop floor in the industry to the designers who design prototypes for effi-
cient cyber-physical system integration with intuitive design ideas. It would also
include collaborating between experts from different disciplines such as machine
learning, design, cognitive science, and psychology, for developing a truly human-
inspired AI focused on intelligent Industry 4.0 applications. The goal would be
to synergize AI, HCI, and Industry 4.0 to attain tenable and tangible goals soon,
with a focus on improving labor wages, implementing efficient industrial practices,
improving work–life balance, and making immediate applications to every sphere
in the industry. Thus, a robust framework for implementing HCI systems in the
Industry 4.0 framework would open up unending opportunities at the confluence
of humans and machines, heralding the era of human-inspired AI research and
thereby, ultimately changing the world for the better, one cyber-physical system at
a time.
Human-Centered Approach to Intelligent Analytics in Industry 4.0 51

REFERENCES
1. C. Krupitzer et al. A Survey on Human Machine Interaction in Industry 4.0, Vol. 1,
No. 1, Article. Association for Computing Machinery (ACM), February 2020. https://
doi.org/10.1145/1122445.1122456
2. I. Islam, K. M. Munim, M. N. Islam, and M. M. Karim. “A proposed secure mobile
money transfer system for SME in Bangladesh: An industry 4.0 perspective,” in 2019
International Conference on Sustainable Technologies for Industry 4.0 (STI). IEEE,
2019.
3. J. Lee, B. Bagheri, and H.-A. Kao. “A cyber-physical systems architecture for industry
4.0-based manufacturing systems,” Manufacturing Letters, vol. 3, pp. 18–23, 2015.
4. H. Gruber. Innovation, Skills and Investment: A Digital Industrial Policy for Europe,
Economia e Politica Industriale. Springer, 2017, pp. 1–17.
5. C. J. Bartodziej. The Concept Industry 4+.0: An Empirical Analysis of Technologies
and Applications in Production Logistics. Springer, 2017.
6. J. A. Saucedo-Martínez, et al. “Industry 4.0 framework for management and opera-
tions: A review,” Journal of Ambient Intelligence and Humanized Computing, vol. 6,
pp. 1–13, 2017.
7. F. Ansari, S. Erol, and W. Sihn. Rethinking Human-Machine Learning in Industry 4.0:
How Does the Paradigm Shift Treat the Role of Human Learning? Science Direct,
2018.
8. S. Nahavandi. “Trusted autonomy between humans and robots: Toward human- on-
the-loop in robotics and autonomous systems,” IEEE Systems, Man, and Cybernetics
Magazine, vol. 3(1), pp. 10–17, 2017.
9. D. Bauer, S. Schumacher, A. Gust, J. Seidelmann, and T. Bauernhansl. “Characterization
of autonomous production by a stage model,” Procedia CIRP, vol. 81, pp. 192–197,
2019.
10. F. J. Campbell. Human factors: The impact on industry and the environment. In E. R.
Rhodes, and H. Naser (Eds.), Natural Resources Management and Biological Sciences
(pp. 1–14). IntechOpen, 2021.
11. T. Zhang, Q. Li, C. Zhang, H. Liang, P. Li, T. Wang, S. Li, Y. Zhu, and C. Wu. “Current
trends in the development of intelligent unmanned autonomous systems,” Frontiers of
Information Technology & Electronic Engineering, vol. 18, no. 1, pp. 68–85, 2017.
12. E. Fosch-Villaronga, C. Lutz, and A. Tamò-Larrieux. “Gathering expert opinions for
social robots’ ethical, legal, and societal concerns: Findings from four international
workshops,” International Journal of Social Robotics, vol. 12, no. 2, pp. 441–458,
2020.
13. M. Gil, M. Albert, J. Fons, and V. Pelechano. “Designing human-in-the-loop autono-
mous cyber-physical systems,” International Journal of Human-Computer Studies, vol.
130, pp. 21–39, 2019.
14. F. Santoni de Sio and J. van den Hoven. “Meaningful human control over autonomous
systems: A philosophical account,” Frontiers in Robotics and AI, vol. 5, pp. 15, 2018.
15. G. Weichhart, A. Ferscha, B. Mutlu, M. Brillinger, K. Diwold, S. Lindstaedt, T. Schreck,
and C. Mayr-Dorn. “Human/machine/roboter: Technologies for cognitive processes.”
Elektrotechnik Und Informationstechnik, vol. 136, no. 7, pp. 313–317, 2019.
16. A. M. Isen. “An influence of positive affect on decision making in complex situations:
Theoretical issues with practical implications,” Journal of Consumer Psychology,
vol. 11, no. 2, pp. 75–85, 2001.
17. M. N. Islam, S. J. Oishwee, S. Z. Mayem, A. N. Mokarrom, M. A. Razzak, and
A.  H.  Kabir, “Developing a multi-channel military application using interactive dia-
logue model (IDM),” in 2017 3rd International Conference on Electrical Information
and Communication Technology (EICT). IEEE, 2017, pp. 1–6.
52 Intelligent Analytics for Industry 4.0 Applications

18. T. Zaki, Z. Sultana, S. A. Rahman, and M. N. Islam, “Exploring and comparing the
performance of design methods used for information intensive websites,” MIST
International Journal of Science and Technology, vol. 8, pp. 49–60, 2020.
19. H. Thimbleby and I. H. Witten. “User modeling as machine identification: New design
methods for HCI.” In H. R. Hartson et al. (Eds.) Advances in Human-Computer
Interaction. Ablex (vol. 4, pp. 58–86), 1991.
20. A. Beard-Gunter, G. Ellis, C. Dando, and P. Found, Designing industrial user
­experiences for industry 4.0. 2018.
21. M. A. Razzak and M. N. Islam. “Exploring and evaluating the usability factors for
military application: A road map for HCI in military applications,” Human Factors and
Mechanical Engineering for Defense and Safety, vol. 4, no. 1, p. 4, 2020.
22. J. Nielsen. Usability Engineering. Elsevier, 1994.
23. S. Deterding, R. Khaled, L. E. Nacke, and D. Dixon. “Gamification: Toward a
Definition,” in CHI 2011 Gamification Workshop Proceedings, vol. 12. Vancouver, BC,
Canada, 2011.
24. B. Shneiderman, C. Plaisant, M.S. Cohen, S. Jacobs, N. Elmqvist, and N. Diakopoulos.
Designing the User Interface: Strategies for Effective Human-Computer Interaction.
Pearson, 2016.
25. S. Uzor, J.T. Jacques, J.J. Dudley, and P.O. Kristensson. “Investigating the accessibility
of crowdwork tasks on mechanical turk,” in Proceedings of the 2021 CHI Conference
on Human Factors in Computing 2021, https://doi.org/10.1145/3411764.3445291
26. Sebastian Büttner, Henrik Mucha, Markus Funk, Thomas Kosch, Mario Aehnelt,
Sebastian Robert, and Carsten Röcker. “The design space of augmented and virtual
reality applications for assistive environments in manufacturing: A visual approach,”
ResearchGate Conference Paper, June 2017.
27. Markus Funk, Albrecht Schmidt, Tilman Dingler, and Jennifer Cooper. “Stop Helping
Me - I’m Bored! Why Assembly Assistance needs to be Adaptive,” ResearchGate,
September 2015.
28. D. Gorecky, M. Schmitt, M. Loskyll, and D. Zuhlke, “Human-Machine- ¨ Interaction
in the Industry 4.0 era,” in 2014 12th IEEE International Conference on Industrial
Informatics (INDIN). 2014, pp. 289–294.
29. E. Aranburu, G. S. Mondragon, G. Lasa, J. K. Gerrikagoitia, I. S. Coop, and G. S.
Elgoibar. “Evaluating the human machine interface experience in industrial work-
places,” in Proceedings of the 32nd International BCS Human Computer Interaction
Conference. BCS Learning & Development Ltd., 2018, p. 93.
30. Peter Papcun, Erik Kajáti, and Jií Koziorek. Human Machine Interface in Concept of
Industry 4.0. IEEE, October 2018.
31. R. Tulloch. Reconceptualising gamification: Play and pedagogy. Online.Available:
http://www.digitalcultureandeducation.com/cms/wpcontent/uploads/2014/12/tulloch.
pdf.
32. T. Schmidt, I. Schmidt, and P. R. Schmidt. “Digitales Spielen und Lernen – A Perfect
Match?: Pädagogische Betrachtungen vom kindlichen Spiel zum digitalen Lernspiel,”
In K. Dadaczynski, S. Schiemann, and P. Paulus, (Eds.) Gesundheit spielend fördern:
Potenziale und Herausforderungen von digitalen Spieleanwendungen für die
Gesundheitsförderung und Prävention, 2016, pp. 18–49.
33. Jacqueline Schuldt, and Susanne Friedemann. “The challenges of gamification in the
age of Industry 4.0.” in IEEE Global Engineering Education Conference, April 2017.
34. T. Schmidt. MALL meets Gamification: Möglichkeiten und Grenzen neuer (digitaler)
Zugänge zum Fremdsprachenlernen. http://www.uni- potsdam.de/fileadmin01/projects/
tefl/
35. Sebastian Büttner, Henrik Mucha, Markus Funk, Thomas Kosch, Mario Aehnelt,
Sebastian Robert, and Carsten Röcker. “The design space of augmented and virtual
reality applications for assistive environments in manufacturing: A visual approach,” in
Human-Centered Approach to Intelligent Analytics in Industry 4.0 53

Proceedings of the 10th International Conference on Pervasive Technologies Related

to Assistive Environments - PETRA ’17. ACM Press, Island of Rhodes, Greece, 2017,
pp. 433–440. https://doi.org/10.1145/3056540.3076193
36. G. Dini and M. Dalle Mura. “Application of augmented reality techniques in through-
life engineering services,” Procedia CIRP, vol. 38, pp. 14–23, 2015. https://doi.
org/10.1016/J.PROCIR.2015.07.044
37. C. D. Fehling, A. Mueller, and M. Aehnelt. “Enhancing vocational training with
­augmented reality.” In: S. Lindstaedt, T. Ley, and H. Sack (Eds.) Proceedings of the
16th International Conference on Knowledge Technologies and Data-Driven Business.
ACM Press, 2016.
38. S. Deterding, D. Dixon, R. Khaled, and L. E. Nacke, Gamification: Toward a Definition.
CHI, 2011.
39. SangSu Choi, Kiwook Jung, and Sang Do Noh. “Virtual reality applications in
manufacturing industries: Past research, present findings, and future directions,”
­
Concurrent Engineering: Research and Applications, vol. 23, no. 1, pp. 40–63, 2015.
https://doi.org/10.1177/1063293X14568814
40. R. Palmarini, J. Ahmet Erkoyuncu, and R. Roy. An innovative process to select
­augmented reality (AR) technology for maintenance. Procedia CIRP, vol. 59, pp. 23–28,
2017. https://doi.org/10.1016/J.PROCIR.2016
41. M. Hermann, T. Pentek, and B. Otto. “Design principles for Industrie 4.0 scenarios,
2016,” in 49th Hawaii International Conference on System Sciences (HICSS). IEEE,
2016, pp. 3928–3937. https://doi.org/10.1109/HICSS.2016.488
42. F. Hecklau, M. Galeitzke, S. Flachs, and H. Kohl. Holistic approach for human
resource management in Industry 4.0. Procedia CIRP, vol. 54, pp. 1–6, 2016. https://
doi.org/10.1016/J.PROCIR.2016.05.102
43. S. Zahoor, M. Bedekar, and V. Vishwarupe. A framework to infer webpage rele-
vancy for a user. In S. Satapathy and S. Das (eds) Proceedings of First International
Conference on Information and Communication Technology for Intelligent Systems:
Smart Innovation, Systems and Technologies, vol 50. Springer, Cham, 2016. https://doi.
org/10.1007/978-3-319-30933-0_16
44. S. Zahoor, M. Bedekar, V. Mane, and V. Vishwarupe. Uniqueness in User Behavior While
Using the Web. In: S. Satapathy, Y. Bhatt, A. Joshi, and D. Mishra (Eds.) Proceedings of
the International Congress on Information and Communication Technology. Advances
in Intelligent Systems and Computing, vol. 438. Springer, Singapore, 2016. https://doi.
org/10.1007/978-981-10-0767-5_24
45. V. Vishwarupe, M. Bedekar, and S. Zahoor. “Zone specific weather monitoring system
using crowdsourcing and telecom infrastructure,” in 2015 International Conference on
Information Processing (ICIP), 2015, pp. 823–827, doi: 10.1109/INFOP.2015.7489495
46. V. Vishwarupe, M. Bedekar, M. Pande, and A. Hiwale. Intelligent Twitter spam
­detection: A hybrid approach. In: X. S. Yang, A. Nagar, and A. Joshi (Eds.) Smart
Trends in Systems, Security and Sustainability. Lecture Notes in Networks and Systems,
vol. 18. Springer, Singapore, 2018. https://doi.org/10.1007/978-981-10-6916-1_17
 4 Advance in Robotics
Industry 4.0
Muchharla Suresh, Rajesh Kumar Dash,
Deepika Padhy, Swastid Dash,
Subhransu Kumar Das, and
K. Nageswara Rao Achary
NIST

CONTENTS
4.1 Introduction...................................................................................................... 55
4.2  Overview of Industry 4.0................................................................................. 58
4.3  Opportunities of Industry 4.0........................................................................... 59
4.4  Challenges in Industrial Robotics 4.0..............................................................60
4.5  Properties of Robotics Industry 4.0................................................................. 61
4.6 Conclusion....................................................................................................... 62
References................................................................................................................. 63

4.1 INTRODUCTION
Due to the advent and prominence of Industry 4.0, robotic technology, which provided
an essential contribution to modern industry, has undergone significant development
in recent years. In the era of Industry 4.0 and the Industrial Internet of Things (IIoT),
it is anticipated that the next age of robotics and its related technologies will play a
major role in meeting the flexible needs of cooperative and smarter production [1].
Industrial robots are embedded with artificial intelligence (AI), which are capable
of replacing the human workers in the manufacturing industries. Human labor is
ineffective in factories for a variety of reasons, including physical limitations that
affect the production performance, manufacturing costs, etc. In the current competi-
tive business, firms have turned away from using human labor and toward industrial
robots in order to accomplish more accurate production in shorter amounts of time
without even any risks [2].
However, for effective and reliable manufacturing, human and robot collaboration
is necessary. Industrial robots are necessary for companies in the modern, tough
marketplace not just for safety reasons to prevent worker injuries while production as
well as for the most need faster, more precise manufacturing when taking financial
prosperity into account. Real-time result and effectiveness predictability in manufac-
turing do not aid in the unsupervised management and optimization of the subsystem
cost and duration. The phrase “Industry 4.0” refers to a new industrialization that is

DOI: 10.1201/9781003321149-4 55
56 Intelligent Analytics for Industry 4.0 Applications

built on smart technology and intends to integrate data science into the ­manufacturing
and production processes to create smart factories. The robots can also indepen-
dently spot a decline in product performance and apply optimization to fix it [3].
Industrial robots are created extremely effectively and cooperatively with people
and other robots through networking in the Industrial revolution 4.0, enabling them
to be self-adaptable and self-aware on new goods and production techniques. As
a result, smart factories with many of these robots will be provided in the future
of industry thanks to recent technologies utilizing IoT, such as for controlling and
wirelessly monitoring on robots, cloud platforms analyzing big data has enhanced
information insights. The robots may also detect product performance degradation
on their own and apply modifications to fix it [1].
About 373,000 industrial robots had been based on these factors in 2019, accord-
ing to the International Federation of Robotics (IRF). Globally, there were 2.7 million
industrial robots in use in factories in 2020. The effective use of industrial robots,
their dependability and affordability, and the effective adoption of the Industry 4.0
concept have sparked an increase in interest in robot enhancement and the investiga-
tion of latest advancements in a variety of fields, especially in typical and nontypi-
cal application domains. Using the term “Industrial robot” as a keyword, more than
4,500 scientific papers had been published in 2019, and 5,300 papers with a similar
interest and study focus had been published in 2020, according to Science Direct,
one of the largest research journals. Regardless of the legislative, socioeconomic, and
sociological forces influencing the market for new robots, scientific interest in this
topic is based on a continual increase in publications [4].
This scholarly research is primarily concerned with the industrial robot appli-
cations in 2018–2021 in industries, where there has historically been little support

FIGURE 4.1  Annual ratio of publications.

Advance in Robotics Industry 4.0 57

FIGURE 4.2  Markets of industrial robots.

for robotization. It also covers core topics including feature extraction, route plan-
ning, and optimizations, as well as human contact. Although the field of regular
­automation and general societal widespread usage of smart gadgets, semi uses of
58 Intelligent Analytics for Industry 4.0 Applications

robotics are still frequently viewed with significant suspicion. The most pervasive
misconception regarding robots is that they will take over human jobs, depriving
human employees without the need for a means of support [5].
Figure 4.1 indicates the yearly ratio of implemented robot publications in
­industries. Here, the number of installations of industrial robots is increasing year by
year. The idea was first introduced in 2012, and still, this process is going on with the
addition of new features and with new technologies.
However, the research presented in, which evaluated the huge backlash about
robots replacing humans in Japan’s fabric and electronics industries, disproved this
notion. Based on an analysis of the number of robots used and the actual cost of
implementation, it was found that the employment of robots increases productivity,
which benefits the much more vulnerable individuals in society, including women,
part-time or full-time workers, college grads, and elderly people. Every robotization
task is distinctive in some manner. In order to increase the functionality or enhance
the qualities of standard automatons, these tasks frequently call for the use of specific
tools, the development of an appropriate working environment, the addition of new
sensors or measurement systems, and the application of sophisticated control algo-
rithms. Industrial robots typically work in larger groups called robotic units or inde-
pendent production lines. As a consequence, especially seemingly simple tasks that
are robotized require a complicated response and a structured method. Figure 4.2
depicts the market of industrial robots.

4.2  OVERVIEW OF INDUSTRY 4.0

Due to the rise of Industry 4.0 and its widespread use, robotic innovations provided
an essential contribution to contemporary manufacturing and have undergone sig-
nificant growth in recent times. According to predictions, the next wave industrial
robots and the innovations that go along with it will play a bigger part in addressing
the changing demands of collaborated and interactive production within the frame-
work of Industry 4.0 [6]. The actual use of a robotic system is indeed a challenging
task that demands providing answers to several queries upon the feasibility of doing
so and the procedure altogether. Based on the industry of production, the situation
differs a little. In certain fields, where they have been used for at least 30 years,
robot implementation is just getting started. Creative innovations are comparatively
easier to implement in industrial settings where robotics has a lengthy history. Such
approaches are generally limited about using additional tools, computational meth-
ods, and process improvement systems for robotic operation. Although their wide
range of applications, many robotics advancements might effectively move from
one sector to the other. Furthermore, overcoming constraints in one area secures
improvements throughout robotics in plenty of other fields. Industrial robots’ ability
to complete common tasks is constrained by a variety of factors. Great identification
efficiency demands, absence of adequate methodologies, and functional requirements
are the key constraints. Except for the tools, these limitations have nothing to do with
the mechanical systems of the robot. Thus, the development of new machinery or
optimization techniques stimulates the majority of mainstream robotic ­applications.
Machine learning (ML), sensor technology, and data analysis are arguably the main
Advance in Robotics Industry 4.0 59

forces behind the expanded use of industrial robots. Despite the fact that robots pri-
marily intended to do repetitive tasks, technologies create robots extremely adapt-
able as well as enable them to come up with clever responsive alternatives. Internal
structures of robots have improved as a result of the development of robot control
systems. These improvements frequently involve the use of new computational mod-
els enabling the control of robots [7].

4.3  OPPORTUNITIES OF INDUSTRY 4.0

Zero downtime and maximum efficiency are the goals of Industry 4.0-enabled robot-
ics [8]. Robotics is an essential Industry 4.0 technology that offers a wide range of
capabilities in the industrial area. Automation systems have been improved because
of this technology. Industry 4.0, often known as the factory of the future, employs
sophisticated robots to boost efficiency by taking over manual jobs and completing
them quicker. This technology performs repeated tasks with greater accuracy and
at a reduced cost. One example is the assembling of flexible pieces. Because robots
can operate in hazardous areas, they eliminate risky employment for people. They
are capable of carrying huge items, hazardous material, and doing repeated jobs.
This has aided businesses in preventing several mishaps while also saving time and
money. Recent advancements in industry 4.0, the Internet of Things, and AI are
expected to have a significant influence on robotics. New maintenance tactics, more
autonomous robots, and new collaborative robotics technologies are made possible
by Industry 4.0 technology.
Industrial robotics may execute many functions at the same time. By replacing a
massive multicore processing capability with an existing programmable logic con-
troller, facility administrators might efficiently leverage industrial floor space while
shrinking the hardware footprint (Programmable Logic Controller, PLC). Major
food firms have integrated robots into their existing food production operations. With
vision technologies, cameras, and AI, the robots perform a variety of tasks in produc-
tion. Everything is being cut, measured, packed, and palletized by them. Industries
with an abundance of sensors will track machinery and manufacturing processes
in real time to forestall anomalous output and services. Machine vision robots are
able to execute sophisticated optical procedures with pinpoint accuracy. Microscopic
flaws or slight color variations can be recognized and repaired instantly to maintain
performance quality [4].
Automation is becoming far more trustworthy than human labor. For decades, the
business has used robots on floors. The robots are equipped with lasers and cameras,
allowing for high-precision welding. Because of the effectiveness of automation, manu-
facturers may eliminate overall waste in the manufacturing chain [5]. When robotic
systems may be improved or reallocated as the market model evolves to fulfill new
operations, replacement expenses can be reduced. Although they may appear to be a
hefty initial investment, industrial robots rapidly pay for themselves through decreased
labor costs and quicker production cycles. Long-term operating and service expenses
would be cheaper than if the same functions were performed by an employee [9–11].
The installation of a smart camera at the end of a robotic arm opens up a wide
range of applications since the arm will traverse over the investigated component to
60 Intelligent Analytics for Industry 4.0 Applications

monitor numerous characteristics. Computer vision uses a variety of innovations to

deliver realistic outcomes from picture gathering and interpretation for robot-based
surveillance and assistance. Vision system inspection, which is rapidly expanding, is
used to identify surface defects, color, and existence. The vehicle sector has been a
significant driver of industrial robotics, using the majority of the robots now are in
use. Body welding has been the most prominent robotic activity in vehicle manufac-
ture. At one point during this process, two metal pieces are fused together to generate
a combined fusion of both materials.
Robots that are programmed to perform continuously throughout the day. In
practically any condition to assist industrial organizations in increasing production,
throughput, and profitability. Despite the fact that automation disrupts the core of
employment, it has offered significant profit prospects in the industrial business.
One of the most frequent types of automation robots is the collaborative robot.
Collaborative robotics will become more versatile in manufacturing and perform
creative activities as technology evolves. Alternatively, recent sensor designs and
algorithms have evolved to guarantee that early robots are more suited for big quanti-
ties, minimum variations, and novel purposes. Robots are so rapidly being employed
in a wide range of industries, including manufacturing [6,7].
Robots are widely employed to apply sealing material cables or adhesives in a
variety of businesses. The material to be applied is liquid or paste in consistency. In
addition, robots are employed in the automobile sector to drill, install, identify, mod-
ify, and process improvements. By automated production, industrial robots assist
many manufacturing industries in increasing efficiency and improving performance.
Sorting, wrapping, and packaging are all significant jobs. Work that is labor intensive
and time-consuming may be done concurrently. It is unrealistic to expect mankind
to have unlimited energy reserves. Furthermore, people will make blunders while
completing projects [12,13].

4.4  CHALLENGES IN INDUSTRIAL ROBOTICS 4.0

Industrial robots are increasingly being connected to the cloud. This means they
can gather data about themselves to enhance productivity, service standards, and
security, all while lowering the overall cost. In the coming decades, industrial robot
­developers will encounter a slew of other new challenges. As these systems get
more complicated, the ability to guarantee reliability becomes extremely critical. Its
purpose is to transform existing industrial production plants into smart factories of
the future by incorporating various technologies such as the IoT, cloud computing,
and AI.
Industrial robots use a variety of memory methods, including storing the boot
code as expansion memory, associated with data memories. With the introduction of
Industry 4.0, there is a strong need to secure systems from cyber attacks.
Because of the possibility of injury, human–robot contact is now challeng-
ing. Manufacturers prior to Industry 4.0 dealt with the problem by not allow-
ing humans and robots to share workstations. This will change when superior AI
develops, allowing these people and a new generation of cooperative robots to
work together. Collaboration entails people and robots working closely together
Advance in Robotics Industry 4.0 61

while maintaining flexibility and productivity. People and robots will be able to
work more closely together while maintaining flexibility and productivity, says
Microsoft’s artificial intelligence boss, Satya Nadella in an interview with BBC
World News.
A new type of cobot is also forming. Chatbots are computer programs that are
designed to assist knowledge workers by gathering data from backend systems, such
as current inventory levels or shipment arrival times. They will work with a range of
interfaces, including the web, smartphone apps, and augmented reality glasses.

4.5  PROPERTIES OF ROBOTICS INDUSTRY 4.0

The industrial manufacturing developed speedily in the last few years. Ten years ago,
the German government began encouraging manufacturers to exploit new technolo-
gies and embrace digitalization. The Industry 4.0 initiative, as it becomes known,
appeared as the early concepts of the IoT started to gather momentum. Industrial
potential for IoT was timely, and the IIoT formed a symbiotic and synonymous rela-
tionship with Industry 4.0. The widespread use of industrial robots started decades
before. However, together the IIoT and Industry 4.0 enabled factory-wide machinery
connectivity, performance monitoring, and predictive asset maintenance on a scale
not previously possible. The abilities of industrial robots have advanced considerably,
from single-axis, single-function units to more complex multiaxis robots that have
interchangeable grippers, manipulators, and effectors. Today’s industrial robots are
able to withstand the harsh and physically demanding environments that are experi-
enced across the many different manufacturing processes. Their ability to continu-
ally perform repetitive and complex tasks precisely improves the overall operating
effectiveness of any manufacturing function significantly. As robotic systems and
electronics technology evolved, researchers started investigating how programmers
could train robots to work collaboratively alongside skilled human workers. The con-
cept of a collaborative robot, or “cobot”, was born and is gathering momentum [4]
with its applications as follows:

i. Technologies for Better Vision, Localization, and Supervision

Instead of operating in solitary surroundings, robotic systems may oper-
ate independently as well as effectively among people in regard to powerful
output detection as well as visual technologies.
ii. Making Use Out of Computer Vision through Robot Control Systems
The ability of the robotic controllers to be connected to a virtual reality
tool allows the operator to see the motives, locations, territories target area,
and whether as has been demonstrated throughout publications also in a
number among the most recent years of advertisement robots. According
to the above, these can be employed when channels of communication
are restricted, such as when wearing rubber gloves prevents the use of a
coaching necklace, as well as when boundaries are present. As a result,
fully functioning human sustainability as well as operability may be imple-
mented at the community level without needing to work in the automation
field [12,13]. In addition to enabling collaborative information exchange
62 Intelligent Analytics for Industry 4.0 Applications

(a) Cell (b) Coexistence (c) Synchronised (d) Co-operation (e) Synergy

FIGURE 4.3  Basic level of collaboration.

between humans and robots, this capability also prevents robots from doing
their duties autonomously.
iii. Technologies for Such IIoT and Wireless Transmission Potential
The most notable technology in today’s industrial generation is arguably
the IoT. IoT investment is anticipated to reach its peak in 2023 at $450 bil-
lion [14]. In 2017, ABB Robotics debuted its own “ability” technology. Such
technologies were created to achieve huge manufacturing with something
like a high level of real-time customization adaptability [15].
Further researches revealed how cloud computing and robots may work
together for commercial applications and how robots improved performance
that encourages the implementation of industry 4.0.
iv. Fundamental Robotization Techniques
In industrial robotics, there are five standard degrees of human–robot
interaction, including (i) no collaboration, (ii) coexistence, (iii) synchroni-
zation, (iv) cooperation, and (v) collaboration.

The basic level of collaboration is covered by the classical strategy as shown in Figure
4.3. It is based on the strategies that robots use to keep people out of their workspace,
such as creating sealed robotic compartments where human activity is forbidden; if
a person was to enter the robotic work area, the robot would need to be halted. The
above method denotes the employment of multiple security measures to identify and
prohibit human access to the robotic work area. The final four collaboration degrees
make up the modern strategy. This is mostly based on an alternative viewpoint that
contends that humans and robots may coexist and operate together in the same work-
place. Such strategies generate new requirements for the architecture, management,
and detection systems of robots. Machines and robots are often known as cobots,
which are designed to operate alongside the human workforce [16].

4.6 CONCLUSION
To help companies fulfill expanding customer demand and maintain their com-
petitiveness in the global market, industrial manufacturing is continually changing.
Advance in Robotics Industry 4.0 63

Currently, robotic devices are penetrating a number of industrial markets. Robotic

gadgets will soon be accessible to customers in a variety of formats as they become
more reasonably priced, now with the potential to have a variety of effects on our
lives. Applications of robotics in the industrial industry have increased protection,
quality, and sustainability for businesses. Robotics integration is advantageous to
the industrial sector with the arrival of Industry 4.0 for a number of reasons, includ-
ing dependability, accuracy, efficiency, and tolerance to harsh situations. It could be
applied to Industry 4.0 to enable more logical decision-making. Additionally, it can
be used in conjunction with business procedures to improve collaboration among
different data systems. It offers a production method that is more dependable and
efficient. Intelligent robots that work quickly and precisely have improved several
production processes in the sector. Due to the current demand, solutions that enable
frequent product modifications at a cheap cost are required. As a consequence, fac-
tory robots have become the best choice for automated assembly. Automation will
therefore provide the manufacturing sector with significant profit prospects in the
future years [17].

REFERENCES
1. L. S. Dalenogare, G. B. Benitez, N. F. Ayala, and A. G. Frank, “The expected contribu-
tion of Industry 4.0 technologies for industrial performance,” Int. J. Prod. Econ., vol.
204, pp. 383–394, 2018.
2. G. Adamson, L. Wang, and P. Moore, “Feature-based control and information frame-
work for adaptive and distributed manufacturing in cyber physical systems,” J. Manuf.
Syst., vol. 43, pp. 305–315, 2017.
3. R. Dekle, “Robots and industrial labor: Evidence from Japan,” J. Jpn. Int. Econ., vol.
58, p. 101108, 2020.
4. K. Aggarwal, S. K. Singh, M. Chopra, S. Kumar, and F. Colace, “Deep learning in
robotics for strengthening Industry 4.0.: Opportunities, challenges and future direc-
tions,” in N. Nedjah, A. A. Abd El-Latif, B. B. Gupta, and L. M. Mourelle (eds) Robotics
and AI for Cybersecurity and Critical Infrastructure in Smart Cities, Springer Cham,
2022, pp. 1–19.
5. E. Erős, M. Dahl, A. Hanna, P.-L. Götvall, P. Falkman, and K. Bengtsson, “Development
of an industry 4.0 demonstrator using sequence planner and ROS2,” in Anis Koubaa
(ed) Robot Operating System (ROS), Springer, 2021, pp. 3–29.
6. B. Kehoe, S. Patil, P. Abbeel, and K. Goldberg, “A survey of research on cloud robotics
and automation,” IEEE Trans. Autom. Sci. Eng., vol. 12, no. 2, pp. 398–409, 2015.
7. I. Zambon, M. Cecchini, G. Egidi, M. G. Saporito, and A. Colantoni, “Revolution 4.0:
Industry vs. agriculture in a future development for SMEs,” Processes, vol. 7, no. 1,
p. 36, 2019.
8. A. Olivares-Alarcos, S. Foix, and G. Alenya, “On inferring intentions in shared tasks
for industrial collaborative robots,” Electronics, vol. 8, no. 11, p. 1306, 2019.
9. A. M. Ghouri, V. Mani, Z. Jiao, V. G. Venkatesh, Y. Shi, and S. S. Kamble, “An
­empirical study of real-time information-receiving using industry 4.0 technologies in
downstream operations,” Technol. Forecast. Soc. Change, vol. 165, p. 120551, 2021.
10. T. Brogårdh, “Present and future robot control development—An industrial perspec-
tive,” Annu. Rev. Control, vol. 31, no. 1, pp. 69–79, 2007.
11. V. Villani, F. Pini, F. Leali, and C. Secchi, “Survey on human–robot collaboration in
industrial settings: Safety, intuitive interfaces and applications,” Mechatronics, vol. 55,
pp. 248–266, 2018.
64 Intelligent Analytics for Industry 4.0 Applications

12. F. D. Ferreira, J. Faria, A. Azevedo, and A. L. Marques, “Product lifecycle management

enabled by industry 4.0 technology,” Transdiscipl. Eng., vol. 6, no. 3, pp. 349–354,
2016.
13. A. T. Rizvi, A. Haleem, S. Bahl, and M. Javaid, “Artificial intelligence (AI) and its
applications in Indian manufacturing: A review,” in S. K. Acharya, D. P. Mishra
(eds) Current Advances in Mechanical Engineering. Lecture Notes in Mechanical
Engineering. Springer, Singapore, 2021. https://doi.org/10.1007/978-981-33-4795-3_76.
14. B. Khoshnevis, “Automated construction by contour crafting—Related robotics and
information technologies,” Autom. Constr., vol. 13, no. 1, pp. 5–19, 2004.
15. I. Malý, D. Sedláček, and P. Leitao, “Augmented reality experiments with industrial
robot in industry 4.0 environment,” in 2016 IEEE 14th International Conference on
Industrial Informatics (INDIN), 2016, pp. 176–181.
16. E. Ruffaldi, F. Brizzi, F. Tecchia, and S. Bacinelli, “Third point of view augmented
reality for robot intentions visualization,” in International Conference on Augmented
Reality, Virtual Reality and Computer Graphics, 2016, pp. 471–478.
17. B. Insider, “Business insider intelligence report.” 2018.
 5 A Cloud-Based Real-Time
Healthcare Monitoring
System for CVD Patients
Anjali Kumari, Animesh Kumar,
Kanishka Vijayvargiya, Piyush Goyal,
Ujjwal Tripathi, and Avinash Chandra Pandey
PDPM IIITDM

CONTENTS
5.1 Introduction..................................................................................................... 65
5.2 Related Works..................................................................................................66
5.3 Preliminary Work............................................................................................ 67
5.4 Methodology.................................................................................................... 67
5.4.1 Data Acquisition.................................................................................. 68
5.4.2 Data Preprocessing.............................................................................. 68
5.4.3 ML Models.......................................................................................... 70
5.5 Performance Analysis...................................................................................... 70
5.6 Application...................................................................................................... 71
5.7 Conclusion....................................................................................................... 71
References................................................................................................................. 71

5.1 INTRODUCTION
Quality healthcare is a need of this century. Chronic diseases such as heart dis-
ease, cancer, and diabetes have risen dramatically due to economic, environmental,
and social development and lifestyle changes. Quality in health care is essential and
refers to ensuring that the patient receives the proper care at the right time. Delays
in monitoring a patient might lead to adverse effects and can even be fatal some-
times. An inefficient healthcare system adds up to poor safety, poor care coordina-
tion, and inefficiencies, costing millions of lives. The healthcare system in India is
not well regulated. Rural areas suffer the most from unattended health checkups.
According to a report published by Novartis, Arogya Parivar: Health for the Poor,
about 72% of the elderly population live in villages [1]. This is often an alarming
number considering that the population of rural areas does not seem to be attended
to too well by health officials. Therefore, it is important to encourage reforms that
result in health systems that are more resilient, better centered on the needs of people,
and sustainable over time. Thus, the idea is to implement systems that make it easier

DOI: 10.1201/9781003321149-5 65
66 Intelligent Analytics for Industry 4.0 Applications

for doctors to understand the areas of abnormality and attend to the sick in their
homes. If the health condition passes a specific threshold, and with the assistance of
machine learning (ML) analysis, it becomes easier to investigate the world as a whole
and deploy any resources if necessary. We use artificial intelligence (AI) due to its
significant impact on recognizing and predicting patient health status based on physi-
ological indications. These readings can be acquired using a variety of biological
signal-measuring instruments. We introduced a brand-new cloud-based, real-time
health monitoring system in this chapter that makes use of Internet of Things (IoT)
and AI. The proposed system can analyze real-time biological data and predict the
health condition of the user. The proposed system also recommends possible treat-
ments for the users.
The rest of the chapter is arranged as follows: Section 5.2 discusses the state-of-
the-art approaches for real-time health monitoring, while in Section 5.3, machine
learning classifiers have been discussed. The proposed model has been discussed
in Section 5.4, and Section 5.5 assesses the efficiency of the proposed method. In
Section 5.5, the application areas of the proposed model have been discussed, and
Section 5.7 concludes the chapter.

5.2  RELATED WORKS

Many researchers have worked in a wide range of medical sectors in order to develop
a more effective health monitoring system that would help people in maintaining their
health. Almost all other countries have been dealing with this public health issue for
the past 10–15 years. There have been numerous attempts by various research groups
to develop effective regression and classification models using machine learning and
deep learning methodologies to predict health-related risks. The well-known classi-
fiers logistic regression, random forest, multiplayer perception, and Gaussian Naive
Bayes were used to create high voltage classification using electroencephalogram
(EEG) signals. Several methods have been presented using AI, IoT, big data, virtual
reality, 3D scanning, 3D printing, and cloud computing [2]. Bhanuteja et al. [3] dis-
cussed the various applications of Industry 4.0 technologies and also identified the
requirements of customized medical gear. Besides, they collected information for
healthcare systems to treat COVID-19 patients properly. Ahmad et al. [4] presented
CAMISA, an AI-powered system for remote monitoring of COVID-19 patients. The
proposed system creates sensor device in the form of a smart shirt and a nebulizer
equipped with several sensors for obtaining a patient’s physiological information.
The system measures a patient’s pulse, SpO2 level, temperature, and breathing rate.
If these metrics surpass their threshold values, notifications are sent to the hospital,
including the patient’s current location. In addition, the authors have used a neural
network model to estimate the likelihood of a user contracting a coronavirus. Palwe
et al. [5] proposed an enhanced CNN model to identify the suspicious person suf-
fering from COVID-19. Habib et al. [6] discussed many IoT-based approaches pre-
sented to handle various health monitoring system. Further, they also discussed the
usage of IoT method to detect COVID-19. Chern et al. [7] presented an ensemble
hard voting model for cardiovascular disease prediction. They also observed that the
advances in machine learning technologies have enabled early disease identification
and diagnosis
Cloud-Based Real-Time Healthcare Monitoring System 67

5.3  PRELIMINARY WORK

A supervised machine learning algorithm understands the function that transforms
an input into an output. It learns this function from analyzing the training data and
can then produce outputs for unknown inputs, which is explained as follows:

i. Decision Tree Classifier (DTC): DT is a paradigm for supervised machine

learning [7]. In the sense that each node (representing one property) links
into two or more subnodes, it can be compared to or thought of as a flow-
chart. A DTC predicts the value of a variable using the decision rules it
has learned or inferred from the training data, like how flowcharts support
decision-making processes by adhering to basic assertions.
ii. Random Forest Classifier (RFC): RFC is a machine learning model that is
used for both classification and regression and is an ensemble model [8]. To
obtain a greater degree of accuracy in prediction, RFCs merge numerous
decision trees.
iii. Logistic Regression (LR): a discrete output’s likelihood can be modeled
using the LR method given an input variable [9]. The most common logistic
regression models produce binary results, which might be true or false, yes
or no, and so on.
iv. Gaussian Naive Bayes (GNB) Classifier: GNB is a classification method
based on Bayes’ theorem and the assumption of predictor independence
[10]. A Naive Bayes classifier believes that the presence of one feature in a
class has no influence on the existence of subsequent features.
v. Ensemble Learning: a broader meta-approach to machine learning called
“ensemble learning” aims to improve predictive performance by aggregat-
ing predictions from many models [11]. Ensemble learning is primarily
used to enhance a model’s performance (classification, prediction, function
approximation, etc.) or lessen the possibility of making a mistaken choice of
a subpar model.
vi. Bagging: bagging, also known as bootstrap aggregation, is an ensemble
learning technique that is widely used to reduce variance in noisy datasets.
A single data point may be chosen more than once during bagging, which
chooses a random sample of data from a training set with replacement [12].
After the creation of numerous data samples, these weak models are indi-
vidually trained, and depending on the task—for example, classification or
regression—the average or majority of those predictions leads to a more
accurate estimate.

5.4 METHODOLOGY
This chapter introduces a novel real-time health monitoring system using IoT and
various ML approaches. The flow of the proposed approach has been presented in
Figure 5.1. The proposed IoT-based remote monitoring combines the capabilities of
IoT and ML to predict areas that are more susceptible to any rise in diseases. The
proposed ML model is first trained with ground truth and then it can be used for ana-
lyzing the real-time data. The real-time data are collected by biological sensors and
68 Intelligent Analytics for Industry 4.0 Applications

FIGURE 5.1  Proposed methodology.

hosted on the cloud to keep track of the records of patients, which in turn improves
the model. Moreover, a notification system has also been implemented, which will
trigger an alert in critical conditions to medical professionals and caretakers of the
patient.
Further, the overall architecture that shows the tools used for developing the com-
plete arrangement is depicted in Figure 5.2. The overall architecture is based on
healthcare IoT, which administers the physiological factors via the vital signs affect-
ing a patient’s health. All the sensors are interfaced with the Raspberry Pi, which is
a microprocessor. The developed hardware enables the system to collect and transfer
vital sign data to cloud storage. The vital signs to be measured with this design are
heart rate, blood pressure, oxygen saturation (SpO2), and body temperature. These
sensed parameters are sent to the edge computing device (Raspberry Pi). Besides, the
detailed description of each phase to monitor the health condition of elderly people
has been discussed in the following section.
First, the data acquisition phase is used to collect real-time data of the target user
using a biological sensor, and then collected data are preprocessed to make it ready
for the analysis. After that, ML models are used to detect any abnormality in real
time and possible treatments are also recommended.

5.4.1 Data Acquisition
The first step is to create a dataset that will hold the training data for the model. Here,
we have used a cardiovascular disease dataset from Kaggle, which includes param-
eters such as age, height, weight, gender, systolic blood pressure, diastolic blood pres-
sure, cholesterol, glucose, smoking, alcohol intake, and physical activity to predict
the presence or absence of cardiovascular disease.

5.4.2 Data Preprocessing
The collected data are cleaned and prepared for model training. If any parameters
contain null values or outliers, those values are dropped from the dataset. Highly
Cloud-Based Real-Time Healthcare Monitoring System 69

FIGURE 5.2  Overall architecture.

correlated features are used to create new features, like weight and height are used
to create BMI. All features are quantized into bins and further encoded using label
encoding techniques. As gender has an enormous impact on the presence of car-
diovascular disease, the dataset is divided into two, and then clustering analysis
(KModes clustering) is performed. The predicted clusters are added as features to
the new combined dataset on which feature selection is performed. The prepared
dataset is divided into three parts for training, validating, and testing.
70 Intelligent Analytics for Industry 4.0 Applications

5.4.3 ML Models
A modified ensemble learning bagging approach is used where a total of four differ-
ent models, namely, Gaussian Naive Bayes, decision tree, random forest, and logistic
regression, run in parallel to predict the class simultaneously. The most suitable class
is selected using a voting classifier. The model allows us to take advantage of all four
classifiers and prevents the system from overfitting.

5.5  PERFORMANCE ANALYSIS

To test the efficiency of the proposed RF-based model, precision, recall, and F1
scores of ML algorithms have been computed. Confusion matrix (CM) is created
to compute the considered performance metrics. The CM for the proposed RF
model has been depicted in Figure 5.3. It can be observed from the figure that the
proposed RF model accurately classifies more than 90% of patients which shows
its efficacy. Based on CM, precision, recall, and F-measure are also computed.
The details of the performance metrics have been presented in Table 5.1. It can
pertinent from the table that the proposed model returns better results on each
performance metric.

FIGURE 5.3  Confusion matrix.

Cloud-Based Real-Time Healthcare Monitoring System 71

TABLE 5.1
Performance Metrics
Measure Value
Sensitivity 0.9222
Specificity 0.93
Precision 0.83
Negative predictive value 0.0778
False positive rate 0.07
False discovery rate 0.1545
False negative rate 0.88
F1 score 0.8857
Matthews correlation coefficient 0.7638

5.6 APPLICATION
Early diagnosis of chronic diseases is one of the applications of health monitoring
utilizing machine learning, which can help physicians, nurses, patients, and other
caregivers make better decisions. Similar to how doctors use this system, regular
people can use it to evaluate whether they have serious health issues and, if so, seek
care at nearby hospitals. The suggested IoT-based intelligent health monitoring sys-
tem is beneficial, particularly for people living alone or elderly patients who require
regular monitoring.

5.7 CONCLUSION
This chapter proposes a method to prevent the disease with early intervention rather
than treatment after diagnosis. It is possible to anticipate diseases more accurately
using the proposed system. This proposed system will help medical professionals
decide more wisely, spot trends and developments, and boost the effectiveness of
studies and clinical trials. It enhances the way healthcare services are delivered,
reduces expenses, and handles patient data with extreme precision.

REFERENCES
1. Novartis, “Arogya Parivar-Improving access to healthcare.” https://www.novartis.com/
in-en/about/arogya-parivar (Accessed on May 5, 2022).
2. Li, J., Y. Meng, L. Ma, S. Du, H. Zhu, Q. Pei, and X. Shen. “A federated learning
based privacy-preserving smart healthcare system.” IEEE Transactions on Industrial
Informatics 18, 2021, 2022.
3. Bhanuteja, G., Kaustubh Anand Kandi, K. Srinidhi, R. Dikshitha, and Anil Kumar.
“CAMISA: An AI solution for COVID-19.” In 2021 International Conference on
Design Innovations for 3Cs Compute Communicate Control (ICDI3C), pp. 216–222.
IEEE, 2021.
72 Intelligent Analytics for Industry 4.0 Applications

4. Ahmad, Muhammad, Saima Sadiq, Ala Saleh Alluhaidan, Muhammad Umer, Saleem
Ullah, and Michele Nappi. “Industry 4.0 technologies and their applications in fight-
ing COVID-19 pandemic using deep learning techniques.” Computers in Biology and
Medicine 145, 105418, 2022.
5. Palwe, Sushila, and Sumedha Sirsikar. “Industry 4.0 Technologies and Their
Applications in Fighting COVID-19.” In Sustainability Measures for COVID-19
Pandemic, pp. 237–251. Springer, Singapore, 2021.
6. Habib, Al-Zadid Sultan Bin, and Tanpia Tasnim. “An ensemble hard voting model for
cardiovascular disease prediction.” In 2020 2nd International Conference on Sustainable
Technologies for Industry 4.0 (STI), pp. 1–6. IEEE, 2020.
7. Chern, Ching-Chin, Weng-U. Lei, Kwei-Long Huang, and Shu-Yi Chen. “A decision
tree classifier for credit assessment problems in big data environments.” Information
Systems and e-Business Management 19, 363–386, 2021.
8. Belgiu, Mariana, and Lucian Drăguţ. “Random forest in remote sensing: A review of
applications and future directions.” ISPRS Journal of Photogrammetry and Remote
Sensing 114, 24–31, 2016.
9. Menard, Scott. Applied Logistic Regression Analysis. No. 106. SAGE, New York, 2002.
10. Ontivero-Ortega, Marlis, Agustin Lage-Castellanos, Giancarlo Valente, Rainer Goebel,
and Mitchell Valdes-Sosa. “Fast Gaussian Naïve Bayes for searchlight classification
analysis.” Neuroimage 163, 471–479, 2017.
11. Dong, Xibin, Zhiwen Yu, Wenming Cao, Yifan Shi, and Qianli Ma. “A survey on
ensemble learning.” Frontiers of Computer Science 14, 241–258, 2020.
12. Breiman, Leo. “Bagging predictors.” Machine Learning 24 (2), 123–140, 1996.
 6 Assessment of Fuzzy
Logic Assessed
Recommender System
A Critical Critique
S. Gopal Krishna Patro
GIET University
Koneru Lakshmaiah Education Foundation

Brojo Kishore Mishra

GIET University

Sanjaya Kumar Panda

NIT Warangal

CONTENTS
6.1 Introduction..................................................................................................... 73
6.2 Recommendation System Approaches............................................................ 74
6.2.1 Collaborative Filtering Method........................................................... 75
6.2.1.1 Cold-Start Problem............................................................... 75
6.2.2 Content-Based Methods....................................................................... 77
6.3 Various Recommendation System Based on E-Commerce............................ 79
6.3.1 Personalized Recommendation System of E-Commerce Based
on Big Data Analysis�������������������������������������������������������������������������� 79
6.3.2 Hybrid Ontology-Dependent RS for E-Commerce.............................80
6.3.3 Trust-Dependent CF Algorithm for E-Commerce RS.........................80
6.3.4 Enhanced Slope One Algorithm.......................................................... 81
6.3.5 Similarity Measures............................................................................. 81
6.4 Conclusions...................................................................................................... 82
References................................................................................................................. 82

6.1 INTRODUCTION
Recommendation systems are referred as information filtering which exhibits the
social elements such as users or groups and the products such as movies, music,
books based on the user interest [1]. It is otherwise stated as based on the user’s
recent behavior such as watching a movie, purchasing a book, or listening to music,

DOI: 10.1201/9781003321149-6 73
74 Intelligent Analytics for Industry 4.0 Applications

the recommender system (RS) or recommendation system works as an information

filter and displayed to the user [2–5]. Various researchers have developed and utilized
effective algorithms for this kind of information filtering process. The recommen-
dation systems are used by many websites such as IMDB, Google, YouTube, and
Amazon. For example, in a movies website, if a user registered and watched adven-
ture movies regularly, then the recommendation system identifies this trend and rec-
ommends adventure movies that he had not watched so far [6,7]. For a company
also, it provides a lot of advantages, and sales are boosted simultaneously. Similar
things are thus found by RS related with user’s likes. In today’s modern world, the
main problem is the information overloading, which leads to navigation problems in
the ocean of all available real choices. RS has been originated as a most generous
big data application that enables the user to detect the exact information in this data
overflowing period. As the e-commerce business is expanded, consumers may be
subjected to a huge amount of data before they make a decision for buying the items
[8]. The system provides a resolution for this data overloading issue thereby suggest-
ing better products to the consumers. The system attempts to improve the e-com-
merce business because it assists in the detection of Internet of Things (IoT) for the
customers they want to purchase but they could not view or detect them among the
overloaded data [9]. RS is generally utilized for enhancing the loyalty and trust of the
customer as it guarantees exclusive personalized service. The interest of the user may
be estimated by directly buying the product based on user preferences or by observ-
ing the past buying behavior [10].
In particular, two other tasks that are linked to the RS are the prediction task and the
recommendation task. With such objectives, various recommendation techniques were
developed on the basis of their working principles and categorized into various kinds.
The contribution toward this paper is listed as follows:

• To review various recommendation system techniques and associated per-

formance metrics.
• To elaborate the recommendation system perspectives based on the 8 meth-
ods such as content-based, collaborative filtering, hybrid RS, social, fuzzy,
genetic, group RS, context-aware, and deep learning.
• To focus on the fuzzy-based recommendation system and analyze the exist-
ing studies accordingly.
• To evaluate the recommendation approaches used in existing studies based
on performance metrics.

6.2  RECOMMENDATION SYSTEM APPROACHES

The formal definition of the RS is as follows:
Consider A and B as the set of items. A function X has been defined, which quanti-
fies the usage of the item A (∈ A) to the corresponding user B (∈ B) through mapping
X : A  X   B →   Ra .
where Ra is referred to as ratings set. The aim is to utilize and learn X to determine the
previously not viewed item a by the user b for the generation of most possible group
Assessment of Fuzzy Logic Assessed Recommender System 75

FIGURE 6.1  Algorithms: recommendation system.

of recommended items. Various conventional recommendation methods ­comprised

collaborative filtering (CF), content-based RS (CBS), and hybrid RS (HRS). Every
method possesses its own merits and demerits. The CF (which is naturally domain
free) has prevented trending works from CBS. But CF has been found to be affected
by sparsity and cold-start problem (CSP), whereas CBD is found to be affected by
overspecialization. Several enhanced methods were suggested for the mitigation of
the disadvantages such as interactive, social, deep learning fuzzy logic, and aware-
based RS, which improve the accuracy and coverage. The recommendation system
overview is shown in Figure 6.1.
The recommendation system purpose is to suggest relevant items to users. To
obtain the above purpose, the major approaches followed by RC are content based
and CF. The brief explanation of these two approaches is as follows.

6.2.1  Collaborative Filtering Method

For RS, the collaborative method is depending on the previously recorded interaction
between the item and user with the intention to generate new recommendations. In the
“user–item interactions matrix” these interactions are stored. The major idea behind
the collaborative methods is the detection of similar users/items by the sufficient past
user–item interactions shown in Figure 6.2. Further, the predictions are made based
on the proximities estimated. The class of the CF algorithms is categorized into two
substrategies named as memory-based and model-based approaches. The approach
under model based refers as generative model, which describes the user–item inter-
actions, and new predictions are made. The memory-based approaches are works
openly with recorded interactions values, not any model assumes and depends on
nearest neighbor search. For example, from user of interest, the closest user is identi-
fied, and among these neighbors, the most famous items are suggested [11].

6.2.1.1  Cold-Start Problem

In recommendation system, the CSP is the common problem. This problem occurs
when the new items or new users arrive in e-commerce platforms. The CF considers
76 Intelligent Analytics for Industry 4.0 Applications

FIGURE 6.2  User/items interactions matrix.

FIGURE 6.3  Representation of cold-start problem in RS.

each item or each user had some ratings so that similar user ratings are derived even
if the ratings are unavailable. It happens when the system does not know the data
about new user with respect to new items [12].
Figure 6.3, represents the CSP, and it happens when new user enters the system
without preference data for that user. In this figure, rows represent the customers or
users and columns represents the products. It had new user problem and new item
problem. CF provides the solution for this CSP [13]. It predicts the ratings to tar-
get user. CF extracts the patterns of similarities and relationships among users and
utilized learn predictive model to generate recommendations. Some predictions are
Assessment of Fuzzy Logic Assessed Recommender System 77

calculated for pair of user items and for every unknown rating by using rating matrix.
The collaborating system filters the items based on highly predicted ratings and it
recommends to user. The CSP is also called as new user problem. The data sparsity
is considered to rectify the issues of cold start. The deep learning and CF are used
for CSP [14]. CF is suffered from “CSP” as for recommendations, it considers past
interactions. To the new users, any kind of recommendation is impossible or to any
user recommending a new item and few interactions corresponding to users or items
to be handled efficiently. This limitation is handled by random strategy – random
items recommendation to new users or to random users with new items, maximum
expectation strategy – recommending the most popular items to new users or to most
active users with new items, exploration strategy – to new users recommending a set
of different items or to set of different users with new item. For the early user life or
the item, the noncollaborative method is used.

6.2.2  Content-Based Methods

The content-based approaches utilized the extra information about users/items like,
for example, in movies RS, this extra information is any other personal data such as
age, sex, and other information about users and also the main actors, movies descrip-
tion, or the duration, whereas the collaborative methods are based only on the inter-
actions between user and item. Based on the available features, the content-based
methods build a model that describes about witnessed user–item interactions, as
shown in Figure 6.4.
The content-based approach has been based on the description of the item and
user preference of the profile. For generating recommendation, potential items have
been compared with the rating items for determining the matching of the item. In
this method, the user interest is oriented on the features based on objects rated. Lee
and Lee presented OTB (outside-the-box) recommendations through the generated
regions of items with respect to the items similarity and the user adherence to the

FIGURE 6.4  Content-based approach.

 78

TABLE 6.1
Comparative Analysis of Existing Recommendation Studies
Reference Technique Description Benefits Limitation
[17] Ontology- Ontology takes different structures in light of the Enhance quality of personalized Need for domain knowledge.
dependent unique situation. Utilizing the Web Ontology recommendation. Distinguishing the right suggestion
recommendation language contrasted and different dialects the Time-consuming approach mix and assessing the
philosophy addressing the XML plot in data set Better performance in hybrid presentation is troublesome in the
setting. recommender system. event of hybrid recommendation-
based ontology.
[18] Context-aware This system comprised of recommendation Recommendations can be adjusted based Integrate contextual information
recommendation provider for gathering the contextual data and on context
feedback data from sensor and learner
application module.
[19] Hybrid filtering This study zeroed in on book proposal No cold-start issue is seen. Combined Time complexity issue is seen.
frameworks in light of crossover approach the Spark big data platform and obtains
contained cooperative and content-based the personalized book recommendation
expectations and compelling mix came about. and further utilization rate is enhanced.
[20] Demographic This study zeroed in on monetary preparation and It does not need rating history of Learner must retrieve personal
recommendation for that proposal cycle needs to change and particular learner. attributes.
worked on in better manner. For various online and offline
applications, this system has used.
[21] Collaborative Different algorithmic techniques featured which is Domain knowledge is not required. Effective utilization of recommender
filtering used to upgrade data recovery to give student a Effectiveness obtained by improved system is not seen in case of
recommendation proficient suggestion by fulfillment level and information retrieval. comparisons with other domains.
execution improvement. Over specification
[22] Fuzzy-based A legitimate watchword extraction method is Efficient classification accuracy resulted. Multimodal data is yet to be
recommendation created in this review for suggesting the items in considered to improve the similar
system a sure and negative way to clients. products search. In an organizational
Intelligent Analytics for Industry 4.0 Applications

setting practical application required

Assessment of Fuzzy Logic Assessed Recommender System 79

regions that overwhelms the chances in the framework [15]. This stated chance from
the returning of the items which are more similar to the existing user rates. This
method possesses a very serious drawback of minimized scope. Due to that, it may
recommend the similar items with the original seed in accordance with the user
altering preference and apart from that it does not recommend the very popular prod-
ucts to the users. Further the naive user issue could not be suggested unless adequate
items were rated by the user for understanding the user preference [16] (Table 6.1).

6.3 VARIOUS RECOMMENDATION SYSTEM

BASED ON E-COMMERCE
The following sections deal with different RSs with e-commerce:

6.3.1 Personalized Recommendation System of

E-Commerce Based on Big Data Analysis
Here, the analysis was segmented into four levels such as presentation layer, business
layer, management layer, and the data layer. All these layers are related closely to big
data. Related to the accuracy of the electronic commerce system, the e-commerce
framework depending on big data was built 5.
Personalized recommendation algorithm (PRA) for e-commerce: personal-
ized recommendation algorithm for commodity n and keyword t in e-commerce is
as follows:

vminimum ≤ 1 (6.1)

qtm ≤ 1 ,where m = 1,2,3,…, M

cnt ≤ 1 (6.2)

Here, cnt represents the similarity of goods. If n and t were similar in M charac-
teristics, then membership of every feature is the same “−” or same “+”, and the
­corresponding product is “+”.

vcnt >= MAXIMUM() * MAXIMUM(qtm )  (6.3)

Generally, the cnt is larger. Suppose if n and t are not same, then it represents that
they are alternate in some characteristics, and membership of them was an alternate
sign, then the corresponding product was “−”, then ntc was smaller. When member-
ship n and keyword t equal to 1 only in characteristic, where other was 0, then cnt =1,
and n and t were entirely the same; where membership of n and t is 1 and −1 in
characteristic, other was 0, then cnt = −1, and n and t were not entirely same. So, for
keyword t, it could recommend suitable PRA by choosing larger items of cnt. This
technique provides a good result for a small range but with a variety of goods the
80 Intelligent Analytics for Industry 4.0 Applications

technique address complexity to grasp while choosing computing membership and

­representative features.

6.3.2 Hybrid Ontology-Dependent RS for E-Commerce

In this study [23], constructs RS with a combination of the K-nearest neighbor (KNN)
algorithm and ontology. Hybrid ontology-dependent RS begins with user profile cre-
ation containing attributes, types, and items purchased by the user [24]. To suggest
items to the user, there was a necessity to predict other users who had purchased
at least 1 common item and chose the one indicating past criteria who have also
purchased other items. These were items suggested to the user [25]. Once creating
the user and his neighbor profile, the KNN algorithm was used to predict the user’s
nearest neighbor. This procedure facilitates extracting items that were not purchased
by the user. Finally, KNN was utilized to predict the nearest item related to the user’s
interest, which contains the overall rating provided by that particular neighbor, price
of items on those items. Additionally, a text review of every product was included
since it acts as a large source to extract needful information [26].

6.3.3 Trust-Dependent CF Algorithm for E-Commerce RS

RS has attained success in resolving the issue of information overload, but still, there
were a few problems such as cold start and data sparseness. In the case of a sparse
rating, it is difficult to receive a satisfying result in RS [27]. One way to resolve this
issue was to present trust within RS. To compute user similarities, Pearson correla-
tion coefficient was utilized to state trust metrics. Available trust metrics depend on
the general guess that information given by the user was accurate, true, and able to
reflect the real interest of the user. However, in numerous cases, this guess was not
reasonable. So, in a way to model good standing measure, more data about rating
and user must be considered. This existing study takes rating data’s reliability and
introduced trust-dependent design depending on the CF algorithm.
Currently, there were numerous fake ratings on e-commerce websites [28]. Here,
the fake rating was classified as follows:

• Because of on-sale activity, users would receive some cashback if they pro-
vide high ratings to items.
• Purposely hiring someone to rate products.
For resolving second type of fake rating, trust-dependent recommenda-
tion design with CF takes the following features such as first, the design
integrates relationship among trusted users, and degree of trust for purpose
of rating. User similarity was a trust relationship among users. On the other
hand, the degree of trust for rating was stated from two features: one was
to spot fraudulent users and eliminate their rating. Second was to give met-
ric for every rating’s trust-dependent strength depending on votes given by
other users. Later, enhanced slope one algorithm was presented depending
on the trust-dependent model.
Assessment of Fuzzy Logic Assessed Recommender System 81

6.3.4 Enhanced Slope One Algorithm

Measures user similarity by recording the rating matrix of the user’s item and then
measures similarity. In the rating matrix of the user’s item, items rated by users
are represented as user vector, then every user was indicated as n dimension rating
­vector, that is Vq = ( sq1 ,  sq 2 ,…, sqn ), sqn is rating user Vq for item Jn. Then, user simi-
larity was measured depending on the user–item rating matrix. This kind of algo-
rithm was utilized in various applications like RS for the social network [27,29–31],
or location-dependent service [32].

6.3.5 Similarity Measures
Similarity measures were utilized to predict the likeliness of two users. Generally,
similarity measures are utilized only when users have rated both items considered.
This creates a bottleneck for new users.
In CF, detection for user j in the nonpersonalized and personalized state is given
as follows:

s ( v, j )
m

Q ( b, j ) = ∑
v =1
m
(6.4)

Here Q is a detection for user b for the item j and detection is made for user b depend-
ing on other users m, s(v, j) is a rating by user v in item j.
Now, there are two considerations, first, if j is not liked by everyone and has vari-
ous opinions about j; second, if the rating utilized for j is not on the same scale. These
cases must be managed differently.
In first case, for detection j to b, if the item was not drawn equally from entire
users, then personalized weight was utilized for each user and finally, it is added.
This is expressed as follows:

∑
m
s ( v, j ) * o ( b,  v )
Q ( b, j ) = v =1
(6.5)
∑
m
o ( b, v )
v =1

Here, s(v, j) is rating done by v for j and o(b, v) is weighted arrangement among users
b and v. It is provided by similarity among users b and v.
In second case, where rating is not on the same scale, the difference of rating of
user v for j item and the average rating for user v is to consider as follows:
m

Q ( b, j ) = ∑s ( v, j ) − s (mv ) + s ( b) (6.6)
v =1

Here, (v, j) denotes user–item pair, and s(b) provides deviation which was added to
the average rating of user b.
82 Intelligent Analytics for Industry 4.0 Applications

CF states that user’s previous agreements are utilized to detect their further
a­ greements. Every user remains stable and cannot modify over time and is also syn-
chronized with one another. So, in CF, the system must work carefully in the area of
agreement. Despite this, there are also cases, where people agree on one domain and
not on other domains [33]. For example, few peoples feel good to discuss politics but
are not interested to have technical discussions. In such a situation, detection for j
was provided by equations 6.2 and 6.3.

6.4 CONCLUSIONS
The evolution of websites results in the data overloading and hence RS offers a com-
prehensive solution for filtering the required information and promotes its usage.
In this chapter, the recommendation system approaches, similarity measures, evalu-
ations, and substantially, the fuzzy logic-based recommendation system are analyzed
and reviewed. The study also discussed the merits and demerits of several RS tech-
niques to offer a panorama by which the researchers, educators, practitioners, and
industrialists could able to quickly understand and learn the scheme behind the RS
to ease the processing. This chapter is then attempted to discuss every key aspect of
the recommendation system for its improvement. This current study is considered as
starting point for the improvement of more recent contributions in the emerging field
of the recommended systems supported by fuzzy tools.

REFERENCES
1 H. Y. Jeong, K. M. Park, M. J. Lee, D. H. Yang, S. H. Kim, and S. Y. Lee, “Vitamin D
and hypertension,” Electrolyte Blood Press., vol. 15, no. 1, pp. 1–11, 2017, doi: 10.5049/
EBP.2017.15.1.1
2. M. Scholz, V. Dorner, G. Schryen, and A. Benlian, “A configuration-based recom-
mender system for supporting e-commerce decisions,” Eur. J. Oper. Res., vol. 259, no.
1, pp. 205–215, 2017, doi: 10.1016/j.ejor.2016.09.057
3. Y. Guo, Y. Liu, A. Oerlemans, S. Lao, S. Wu, and M. S. Lew, “Deep learning for visual
understanding: A review,” Neurocomputing, vol. 187, pp. 27–48, 2016, doi: 10.1016/j.
neucom.2015.09.116
4. C. J. Carmona, S. Ramírez-Gallego, F. Torres, E. Bernal, M. J. Del Jesus, and S. García,
“Web usage mining to improve the design of an e-commerce website: OrOliveSur.
com,” Expert Syst. Appl., vol. 39, no. 12, pp. 11243–11249, 2012, doi: 10.1016/j.
eswa.2012.03.046
5. T. K. Paradarami, N. D. Bastian, and J. L. Wightman, “A hybrid recommender system
using artificial neural networks,” Expert Syst. Appl., vol. 83, pp. 300–313, 2017, doi:
10.1016/j.eswa.2017.04.046
6. J. Lv, B. Song, J. Guo, X. Du, and M. Guizani, “Interest-related item similarity model
based on multimodal data for top-N recommendation,” IEEE Access, vol. 7, no. c,
pp. 12809–12821, 2019, doi: 10.1109/ACCESS.2019.2893355
7. X. Li and J. She, “Collaborative variational autoencoder for recommender systems,”
Proc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., vol. Part F1296, pp. 305–
314, 2017, doi: 10.1145/3097983.3098077
8. G. Shani, D. Heckerman, and R. I. Brafman, “An MDP-based recommender system,” J.
Mach. Learn. Res., vol. 6, pp. 1265–1295, 2005.
Assessment of Fuzzy Logic Assessed Recommender System 83

9. G. van Capelleveen, C. Amrit, D. M. Yazan, and H. Zijm, “The recommender canvas:

A model for developing and documenting recommender system design,” Expert Syst.
Appl., vol. 129, pp. 97–117, 2019, doi: 10.1016/j.eswa.2019.04.001
10. S. Parvatikar and D. Parasar, “Recommendation system using machine learning,”
Int. J. sArtif. Intell. Mach. Learn., vol. 1, no. 1, p. 24, 2021, doi: 10.51483/ijaiml.1.1.
2021.24-30
11. J. Guo, J. Deng, and Y. Wang, “An intuitionistic fuzzy set based hybrid similarity model
for recommender system,” Expert Syst. Appl., vol. 135, pp. 153–163, 2019, doi: 10.1016/j.
eswa.2019.06.008
12. K. Xinchang, P. Vilakone, and D. S. Park, “Movie recommendation algorithm using
social network analysis to alleviate cold-start problem,” J. Inf. Process. Syst., vol. 15,
no. 3, pp. 1–16, 2019, doi: 10.3745/JIPS.04.0121
13. A. K. Pandey and D. S. Rajpoot, “Resolving cold start problem in recommendation
system using demographic approach,” 2016 Int. Conf. Signal Process. Commun. ICSC
2016, pp. 213–218, 2016, doi: 10.1109/ICSPCom.2016.7980578
14. J. Wei, J. He, K. Chen, Y. Zhou, and Z. Tang, “Collaborative filtering and deep ­learning
based recommendation system for cold start items,” Expert Syst. Appl., vol. 69,
pp. 1339–1351, 2017, doi: 10.1016/j.eswa.2016.09.040
15. K. Lee and K. Lee, “Escaping your comfort zone: A graph-based recommender system
for finding novel recommendations among relevant items,” Expert Syst. Appl., vol. 42,
no. 10, pp. 4851–4858, 2015, doi: 10.1016/j.eswa.2014.07.024
16. J. Son and S. B. Kim, “Content-based filtering for recommendation systems using
multiattribute networks,” Expert Syst. Appl., vol. 89, pp. 404–412, 2017, doi: 10.1016/j.
eswa.2017.08.008
17. G. George and A. M. Lal, “Review of ontology-based recommender systems in­
e-learning,” Comput. Educ., vol. 142, p. 103642, 2019, doi: 10.1016/j.compedu.2019.
103642
18. Y. Chen, X. Li, J. Liu, and Z. Ying, “Recommendation system for adaptive learning,”
Appl. Psychol. Meas., vol. 42, no. 1, pp. 24–41, 2018, doi: 10.1177/0146621617697959
19. Y. Tian, B. Zheng, Y. Wang, Y. Zhang, and Q. Wu, “College library personalized rec-
ommendation system based on hybrid recommendation algorithm,” Procedia CIRP,
vol. 83, pp. 490–494, 2019, doi: 10.1016/j.procir.2019.04.126
20. X. Li, M. Gao, W. Rong, Q. Xiong, and J. Wen, “Shilling attacks analysis in collabora-
tive filtering based web service recommendation systems,” Proc. -2016 IEEE Int. Conf.
Web Serv. ICWS 2016, pp. 538–545, 2016, doi: 10.1109/ICWS.2016.75
21. H. F. Nweke, Y. W. Teh, M. A. Al-garadi, and U. R. Alo, “Deep learning algorithms for
human activity recognition using mobile and wearable sensor networks: State of the art
and research challenges,” Expert Syst. Appl., vol. 105, pp. 233–261, 2018, doi: 10.1016/j.
eswa.2018.03.056
22. L. Kolhe, A. K. Jetawat, and V. Khairnar, “Robust product recommendation system
using modified grey wolf optimizer and quantum inspired possibilistic fuzzy C-means,”
Cluster Comput., vol. 24, no. 2, pp. 953–968, 2021, doi: 10.1007/s10586-020-03171-6
23. M. Guia, R. R. Silva, and J. Bernardino, “A hybrid ontology-based recommendation
­system in e-commerce,” Algorithms, vol. 12, no. 11, pp. 1–19, 2019, doi: 10.3390/
a12110239
24. I. Huitzil, F. Alegre, and F. Bobillo, “GimmeHop: A recommender system for mobile
devices using ontology reasoners and fuzzy logic,” Fuzzy Sets Syst., vol. 401, pp. 55–77,
2020, doi: 10.1016/j.fss.2019.12.001
25. M. Nilashi, O. Ibrahim, and K. Bagherifard, “A recommender system based on collab-
orative filtering using ontology and dimensionality reduction techniques,” Expert Syst.
Appl., vol. 92, pp. 507–520, 2018, doi: 10.1016/j.eswa.2017.09.058
84 Intelligent Analytics for Industry 4.0 Applications

26. G. D. Congying Guan, Shengfeng Qin, Wessie Ling, “Apparel recommendation

­system evolution : An empirical review Article information : About Emerald www.­
emeraldinsight.com,” Int. J. Cloth. Sci. Technol., 2016. doi: 10.1108/ijcst-09-2015-0100
27. L. Jiang, Y. Cheng, L. Yang, J. Li, H. Yan, and X. Wang, “A trust-based collabora-
tive filtering algorithm for E-commerce recommendation system,” J. Ambient Intell.
Humaniz. Comput., vol. 10, no. 8, pp. 3023–3034, 2019, doi: 10.1007/s12652-018-0928-7
28. R. Katarya and O. P. Verma, “An effective web page recommender system with fuzzy
c-mean clustering,” Multimed. Tools Appl., vol. 76, no. 20, pp. 21481–21496, 2017, doi:
10.1007/s11042-016-4078-7
29. Y. Zhang, H. Abbas, and Y. Sun, “Smart e-commerce integration with recom-
mender systems,” Electron. Mark., vol. 29, no. 2, pp. 219–220, 2019, doi: 10.1007/
s12525-019-00346-x
30. J. Cai, Y. Wang, Y. Liu, J. Z. Luo, W. Wei, and X. Xu, “Enhancing network capacity by
weakening community structure in scale-free network,” Futur. Gener. Comput. Syst.,
vol. 87, pp. 765–771, 2018, doi: 10.1016/j.future.2017.08.014
31. S. Peng, A. Yang, L. Cao, S. Yu, and D. Xie, “Social influence modeling using informa-
tion theory in mobile social networks,” Inf. Sci. (Ny)., vol. 379, pp. 146–159, 2017, doi:
10.1016/j.ins.2016.08.023
32. T. Peng, Q. Liu, D. Meng, and G. Wang, “Collaborative trajectory privacy preserving
scheme in location-based services,” Inf. Sci. (Ny)., vol. 387, pp. 165–179, 2017, doi:
10.1016/j.ins.2016.08.010
33. J. Iwanaga, N. Nishimura, N. Sukegawa, and Y. Takano, “Improving collaborative filter-
ing recommendations by estimating user preferences from clickstream data,” Electron.
Commer. Res. Appl., vol. 37, p. 100877, 2019, doi: 10.1016/j.elerap.2019.100877
 7 Intelligent Analytics in
Big Data and Cloud
Big Data; Analytics; Cloud
K. Aditya Shastry and B. A. Manjunatha
Nitte Meenakshi Institute of Technology

CONTENTS
7.1 Introduction..................................................................................................... 86
7.2 Cloud-Based “Big Data” Devices.................................................................... 91
7.2.1 “Amazon Web Services (AWS)”.......................................................... 91
7.2.2 Amazon Elastic Search Service (Big Data Analytics Framework)..... 91
7.2.3 “Amazon S3” (Big Data Storage Framework)..................................... 91
7.2.4 Amazon Redshift (Data Warehousing Framework)............................ 91
7.2.5 “AWS Identity and Access Management (IAM)”................................ 91
7.2.6 Google Cloud Platform (GCP).............................................................92
7.2.7 “Google Cloud Storage” (Big Data Storage Framework)....................92
7.2.8 “Google BigQuery” (Big Data Warehousing Framework)..................92
7.2.9 “IBM Cloud”.......................................................................................92
7.2.10 “Analytics Engine” (Big Data Analytics Framework).........................92
7.2.11 “IBM Cloud Object Storage” (Big Data Storage Framework)............ 93
7.2.12 “IBM Db2 Warehouse” on Cloud (Big Data
Warehouse Framework)���������������������������������������������������������������������� 93
7.3 Microsoft Azure.............................................................................................. 93
7.3.1 “Azure HDInsight (Big Data Analytics Solution)”.............................. 93
7.3.2 “Azure Blob Storage (Big Data Storage Solution)”............................. 93
7.3.3 “Azure SQL Data Warehouse (Big Data Warehouse Solution)”.........97
7.4 Assessment of Cloud-Centered “Big Data” Business Solutions Designs........97
7.5 Real-World Case Studies in Big Data and Cloud-Based Analytics.................97
7.5.1 Big Data Case Study-1 – “Walmart”...................................................97
7.5.2 Big Data Case Study 2 – “Uber”.........................................................97
7.5.3 Big Data Case Study 3 – “Netflix”...................................................... 98
7.5.4 Big Data Case Study 4 – “eBay”......................................................... 98
7.5.5 Big Data Case Study 5 – “Procter & Gamble”.................................... 98
7.5.6 Case Study 6: Big Data Predicting the Uncertainties..........................99
7.5.7 Deployment of “Big Data” for Ad Network Assessment in more
Than Ten Regions of the World���������������������������������������������������������� 99
7.5.8 Case Study 8: Design of an IoT Pet Monitor Using Big Data
Strategy����������������������������������������������������������������������������������������������101

DOI: 10.1201/9781003321149-7 85
86 Intelligent Analytics for Industry 4.0 Applications

7.5.9 Case Study 9: Development of a Telecommunication Company’s

Data Analytics Solution���������������������������������������������������������������������102
7.5.10 Case Study 10: Implementation of Data Analytics for a
Multibusiness Organization���������������������������������������������������������������103
7.5.11 Case Study 11: Big Data Consulting and Training for a Satellite
Agency�����������������������������������������������������������������������������������������������105
7.5.12 Case Study 12: Big Data Consulting for a Leading Internet of
Vehicles Company�����������������������������������������������������������������������������106
7.6 Challenges and Issues.................................................................................... 107
7.7 Conclusion..................................................................................................... 108
7.8 Future Research Directions........................................................................... 109
References............................................................................................................... 109

7.1 INTRODUCTION
The volume of information produced, processed, and exchanged has increased since
the dawn of the digital revolution. Large amounts of information can be found in a
variety of places, including spreadsheets, sites, and forums, including acoustic feeds.
As a consequence of increased data collection, vast volumes of widespread and com-
plicated data are generated, that must be effectively generated, saved, distributed,
and evaluated in order to mine meaningful knowledge. Such information has enor-
mous promise, but it is becoming increasingly sophisticated, insecure, and risky, as
well as becoming irrelevant. Given that this research may encompass accessibility
and study of clinical records, engagements, financial documents, and public records,
including genomic patterns, the advantages, and constraints of accessing this infor-
mation are debatable. The notion of big data processing and analytics was born out
of the need for efficient and effective analytics services, apps, development environ-
ments, and platforms [1].
Big data analytics (BDA) is being used in a variety of sectors and areas. Health
research, transportation systems, and worldwide safety, as well as the modeling of
challenges in the cultural and ecological sectors, are just a few of the uses. Aside
from standard applications in industry, commerce, and societal management, current
research is focusing on the potential uses of “big data” in practice [2]. Among them
most important potential uses of “big data analytics” and “cloud computing” is in the
biological sciences [3]. Among the high-impact fields defined are tissue engineering,
structural and posttranslational modification predictions, tailored therapy, and meta-
transcriptomics. Aside from that, among the major uses of “big data analytics” is to
enhance the client satisfaction levels of prevailing business models.
By design, “big data” refers to a variety of information types, including formatted,
quasi, and unstructured information, resulting in a complicated data architecture [4].
Such infrastructure’s complexities necessitate sophisticated administration and tech-
nological solutions. The multi-V paradigm forms one of the most widely used frame-
works for interpreting huge datasets. The multi-V model is depicted in Figure 7.1.
“Variety”, “volume”, “velocity”, “veracity”, “value”, “viability”, “validity”, and
“volatility” are the Vs utilized to describe “big data” [5]. “Variety” is determined
by the many types of information present in a database, whereas “velocity” is
determined by the frequency with which information is produced. The quantity of
Intelligent Analytics in Big Data and Cloud 87

FIGURE 7.1  Six V’s of big data.

information is referred to as volume. Veracity and value are two more attributes that
reflect accuracy of information and usefulness in relation to big data mining, corre-
spondingly. “Viability” is the ability of the model to signify reality. “Validity” refers
to the accuracy and correctness of the data. “Volatility” refers to the time duration
for which the data is useful or relevant. Furthermore, the researchers [6] provided
additional characterization known as the “HACE” theorem. Big data includes two
key properties, as per this hypothesis. To begin with, it has a significant quantity
of complex information from various diverse sources. Secondly, the information is
decentralized and dispersed.
On the Internet, with all the apps created on it, information is the most impor-
tant component of cooperation and interaction. The enormous growing importance
of information-centric services like “Facebook”, “LinkedIn”, “Twitter”, “Amazon”,
“eBay”, and “Google+” is contributing to an increase in the demand for data process-
ing and storage on the cloud. Schouten [7] cites Gartner’s prediction that by 2016,
one-half of all information will be stored in the “cloud”.
Furthermore, the big data analytics data mining methods have powerful compu-
tational demands. As a result, high-performance CPUs are required to complete the
task. The cloud is an excellent platform for storing, analyzing, and interpreting large
amounts of information, since it meets two of the most important needs of “big data
88 Intelligent Analytics for Industry 4.0 Applications

analytics”: high capacity and high-performance computation. Software and data apps
are written, installed, and implemented “as a service” in the “cloud computing” envi-
ronment. “Platform as a service” (PaaS), “software as a service” (SaaS), and “infra-
structure as a service” (IaaS) are three cross-layered architectures. IaaS is a strategy
for renting virtualized resources. Likewise, “PaaS” and “SaaS” represent the “cloud
services” that give “software” platforms or “software as a service” to their customers.
With the introduction of cloud-based solutions, the cost of storage has dropped
dramatically. Furthermore, the pay-as-you-go approach as well as the notion of com-
modities technology enables the effectual handling of enormous amounts of infor-
mation, resulting in the notion of “big data as a service”. “Google BigQuery” is an
illustration of such a platform, that gives real-time observations via huge amounts
of information in the cloud [8]. The authors [9] show how the cloud may be used to
handle “big data” in educational systems, with a specific emphasis on information at
the campus level.
Nonetheless, there were not many effective “cloud-based big data analytics” apps
yet. As a result, researchers are increasingly focusing on cloud-based “big data ana-
lytics”. Information protection and data confidentiality are two issues that are vis-
ible in this setup. Information trustworthiness is also characterized as a service as
part of the cloud solutions. There will be a major drop in credibility as the potential
for security vulnerabilities and privacy violations increases dramatically if big data
methods are implemented in the cloud. There will be a major problem of control
and ownership. The promise of “cloud-based big data analytics”, on the other hand,
has prompted researchers to examine prevailing difficulties and look for solutions.
The numerous angles and characteristics of data mining methodologies’ usage in the
cloud infrastructure for “big data analytics” are discussed in this paper. In addition,
it investigates ongoing studies and vulnerabilities associated with them, including
study instructions prospects in “cloud-based big data analytics”.
Due to the high general amount and intricacy of the information involved, conven-
tional database technologies, including computation or statistical methods, could not
be employed for “big data analytics”. Classic business intelligence solutions involve
“OLAP”, “BPM”, “mining”, and repository systems, including “Relational Data
Base Management System (RDBMS)”, that are associated with traditional statistical
approaches and methods.
“MapReduce” is one of the most prominent models for information processing
on a network of machines. An overview of modeling techniques that facilitate large
data analytics is provided [10]. Although “MapReduce/Hadoop” is identified as one
of the most productive models for “big data analytics”, it also adds other dialects and
modules such as “HiveQL”, “Latin”, and “Pig” that suggest substantial improvements
for this task. “Hadoop” is merely a transparent version of the “MapReduce” context,
which was designed as a shared file system in the first place. As per the work [11],
“Hadoop” has progressed into a whole environment or architecture that operates
with “MapReduce” components and comprises a variety of technological tools such
as the “Hive” and “Pig” languages, a synchronization service known as “Zookeeper”,
and a networked database store known as HBase. “Google MapReduce”, “Spark”,
“Hadoop”, “Twister”, “Hadoop Reduce”, and “Hadoop++” are certain frameworks
available for cloud-based big data analytics [12]. The application of cloud technology
in “big data analytics” is illustrated in Figure 7.2.
Intelligent Analytics in Big Data and Cloud 89

FIGURE 7.2  Application of cloud technology in big data analytics.

90 Intelligent Analytics for Industry 4.0 Applications

Information is stored and processed using these technologies. Repositories like

“HBase”, “BigTable”, and “HadoopDB” can be utilized to hold this data, which could
be of any form. “Pig” and “Hive” tools come into play anymore when it comes to
information processing.
In recent decades, the lack of interaction has been identified as a serious prob-
lem, and numerous initiatives have been launched to address it. The authors improve
the sensitivity of the “HBase” and “Hadoop File System (HDFS)” implementations.
The work [13] assesses the practicality of “OLAP Web” Apps for cloud-centric
­infrastructures, with the goal of enabling transparent and widespread use of online
scientific techniques [50,51].
A massive information management framework for the cloud has been developed
through investigation. The authors [14] offer an information architecture and for-
mat for large datasets on the cloud, with the goal of making data searching easier
for users. Furthermore, efficiency and fast response have been significant research
topics. The authors [15] explore the use of a recommended “Hadoop” and “MPI/
OpenMP” system, as well as how it might enhance the functionality and reliability.
Since information must be transported across datacenters, which are typically sep-
arated by long distances, energy usage has become a critical metric for determining
the process performance. Regarding big storage and computation, a network-based
navigation method known as “GreeDi” can be utilized to locate the best resource
optimum approach to the cloud computing environment [16].
There are several framework analytic devices available. The authors [17] provide
a “direct acrylic graph (DAG)” form related to the technology that was utilized to
forecast the Dengue spread across Singapore. “Aneka” [18] and “CloudComet” [19]
are two examples of live threat analytics and the necessity for an architecture, which
could give consumers with coding facilities and capital to carry out the same. In this
[20], the researchers look at the notion of CAAAS, or “Continuous Analytics as A
Service”, which is used to anticipate a service’s or a participant’s behavior.
Real-time big data analysis is the final focus within “big data analysis”, which has
attracted the interest of academic researchers. Several corporate software solution
vendors offer real-time analytic capabilities. AWS Kinesis [21] is an AWS-based
platform for live stream handling. For this goal, several architectures and computer
platforms were established, including “Apache S4” [22], “IBM InfoSphere Streams”
[23], and “Storm” [24].
Table 7.1 lists the gap between big data and cloud computing.
The balance of this article is laid out as follows: the methodologies in big data
analytics are highlighted in the second section. The final segment uses examples to

TABLE 7.1
Gap between Big Data and Cloud Computing
Big Data Cloud Computing
Huge data increasing rapidly with time On-demand availability of computing resources
Includes structured, unstructured, and semi- Includes IaaS, PaaS, and SaaS
structured data
Purpose is organization of large data Purpose is to store and process data in cloud
Intelligent Analytics in Big Data and Cloud 91

demonstrate how big data and cloud-based intelligence could be used. The chapter’s
synopsis appears at the end.

7.2  CLOUD-BASED “BIG DATA” DEVICES

Three techniques for “big data analytics”, “big data storage”, and “big data ware-
house” were chosen from the stated prominent suppliers in this part, including their
succinct advantages and functions [25].

7.2.1  “Amazon Web Services (AWS)”

AWS offers a variety of BDA frameworks that make developing and delivering BDA
apps simple and fast.

7.2.2 Amazon Elastic Search Service (Big Data Analytics Framework)

“Amazon ElasticSearch (Amazon ES)” is an open-source search and managed
­service for building domains, as well as establishing, managing, and scaling AWS
cloud ElasticSearch clusters [26], with features such as continuous integration man-
agement, log analytics, and click-stream analytics. Some of the benefits include inte-
gration with other AWS services, ease of use, expandable clustering, ease of access,
and compatibility with open-source APIs and applications [28–31].

7.2.3  “Amazon S3” (Big Data Storage Framework)

“Amazon S3” is a density and efficiency, secure, and reliable “big data storage”
­service with a diverse set of algorithms and applications, including big data analytics.
Elastomeric leadership with more individual innovative of administration and stor-
age business solutions, unparalleled reliability, consistency, expandability, in-place
review, and assimilation with more companies can successfully, often these broad
cybersecurity proficiencies, and streamlined and speeded up data transmission are
just a few of the benefits of utilizing this structure [26–28,30,32,33].

7.2.4 Amazon Redshift (Data Warehousing Framework)

“Amazon Redshift” is a well-known, highly scalable “AWS data warehouse” plat-
form that enables rapid, simple, and cost-effective research methodology through
the use of traditional SQL and popular advanced analytics. Rapid data speed, price,
a straightforward and highly secure architecture, a flexible and expandable cluster,
interoperability with a variety of SQL users, and huge extendibility are just a few of
“Amazon Redshift’s” features [26,30,34,35].

7.2.5  “AWS Identity and Access Management (IAM)”

IAM is a web service that helps you securely control access to AWS resources.
You use IAM to control who is authenticated (signed in) and authorized (has per-
mission) to use resources. IAM provides shared access to AWS accounts, granular
92 Intelligent Analytics for Industry 4.0 Applications

permissions, secure access to AWS resources for applications that run on Amazon
EC2, multifactor authentication (MFA), and identity federation, to name a few [52].

7.2.6 Google Cloud Platform (GCP)

“GCP” provides a number of useful tools for a wide range of tasks, including “big
data analytics”, “data warehouses”, “repository and archiving”, and much more is a
list of the most common “big data analytics”, “big data storage”, and “big data ware-
housing” framework technologies. As its name suggests, “Google Cloud Dataproc” is
a completely structured and controlled cloud-centric “Apache Hadoop” and “Spark”
solution for quick, simple, and cost-effective group organizational processes. Rapid
ensemble scalability, cost effectiveness, and an open-source architecture are some of
its characteristics and advantages [30,31,35].

7.2.7  “Google Cloud Storage” (Big Data Storage Framework)

“Google Cloud Storage” is an integrated object storage designed primarily for orga-
nizations and developers that performs a variety of activities, including actual data
processing, data archiving (with “Coldline” and “Nearline” storage solutions), and
analytics. Accessibility with a cheap price, streamlined preservation, as well as stor-
age, fast and easy information transfer, pricing, improved security for corporation
assets, and collaboration with top companies are just a few of the advantages of using
it [30,32,33,35].

7.2.8  “Google BigQuery” (Big Data Warehousing Framework)

It is a fully organized, massively scalable, fast, and low-bandwidth repository for
business intelligence in operation. Rapid architecture installation, easy expansion,
and analysis with fast analytics, including business data and capital security, are just
a few of the benefits of adopting this solution [30,35].

7.2.9  “IBM Cloud”

The “IBM cloud” provides a wide range of cloud-centric “big data solutions”, as well
as the right instrument for the task, including the “big data analytics” framework,
connectivity, and analytics. Like the paradigms chosen by cloud vendors previously
explored, one paradigm for each of “big data analytics”, “big data storage”, and “big
data warehousing” is chosen and described below.

7.2.10  “Analytics Engine” (Big Data Analytics Framework)

It is one of the IBM cloud products that helps businesses with fast information pro-
cessing and the resolution of various big data difficulties. It’s a dedicated clustering
solution that interfaces with “Apache Hadoop” and “Apache Spark” to make it easier
to design and operate analytics solutions. Accessible, expandable groups and a pro-
grammable ecosystem are among its characteristics [31,36,37].
Intelligent Analytics in Big Data and Cloud 93

7.2.11  “IBM Cloud Object Storage” (Big Data Storage Framework)

This IBM-developed cloud-based storage platform allows users to store, manage, and
access data using REST-based APIs and IBM’s self-service site. This platform could
link directly to apps as well as link to certain other IBM cloud services. Endurance
as well as stability, geographic resiliency, memory versatility, rest-based APIs and
SDKs, and IBM cloud administration interface are some of its major characteristics
[33,36,37].

7.2.12 “IBM Db2 Warehouse” on Cloud (Big

Data Warehouse Framework)
This is an enterprise-class cloud-centric “data warehouse” facility powered by IBM
CLU Acceleration for unmatched throughput. This solution is well-organized and
safe, works with “Oracle” and “Nerezza”, is created for the distributed environment,
serves as an information storage for data scientists, and relieves analytical capacity
[36,37].

7.3  MICROSOFT AZURE

“Microsoft Azure” provides a diverse set of cloud programs that benefit designers
and IT experts for building, launching, and trying to manage various applications,
from mobile apps to ISC (Internet-scale computing) keys, all through its global net-
work of datacenters and with the help of “DevOps” and unified tools. “BDA”, “big
data storage”, and “big data warehouse” solutions are also chosen, and their charac-
teristics and advantages are described as follows.

7.3.1  “Azure HDInsight (Big Data Analytics Solution)”

In contrast to other modern facilities, which only provide service level agreement
(SLA) for essential virtual servers, our solution provides 99.99% SLA for a single
example of a “virtual machine”. It, for example, allows for the building of optimized
groups for “Spark”, “Hadoop”, “Kafka”, “HBase Storm”, and “Microsoft R Servers”,
with a 99.9% SLA. Global availability, strong cybersecurity, a very efficient environ-
ment for research and innovation, pricing, and remarkable adaptability are just a few
of its distinguishing characteristics [31,38,39,40].

7.3.2  “Azure Blob Storage (Big Data Storage Solution)”

“Blob Storage” is a simple and price-operative storage solution for “exabytes” of
unstructured data such as audio, movies, photos, and more, organized into three tiers
dependent on the regularity of network connectivity: warm, cooling, and archives.
Numerous blob types, such as site, restrict, and append blob, offer memory optimi-
zation flexibility as well as fully automated spatial for simple equality of enhanced
local and global connectivity [30,38,39].
 94

TABLE 7.2
Frameworks for “Cloud-Based Big Data” Enterprise Solutions [25]
Architecture and
Sl. No. Operations “AWS” “GCP” “IBM Cloud” “MS Azure”
[-5ex]1 “Big Data Analytics” “Amazon Elastic Search “Google Cloud Dataproc” “IBM Analytics Engine” “Azure HDInsight”
Service”
Mode of Application Publicly available Publicly available Publicly available Publicly available
Information types Organized, quasi, and Organized, quasi, and unsupervised unsupervised
unsupervised information unsupervised information
Origins of information “Amazon S3”, “Amazon “Google Bigtable”, IBM “Cloud Object Storage” “Blob Storage”
Kinesis Firehose”, and “Google Cloud Storage”,
“Amazon DynamoDB” and “Google Big Query
“Operating System” “CentOS”, “Ubuntu”, and “Debian 8” “CentOS 7” “Ubuntu 14”, “Ubuntu
“Amazon Linux” 16”, and “Windows
Server 2012 R2”
“Functionality” Web browsing analysis and Machine learning (ML), Advanced analytics, vendor support Information technology in
continuous integration flow analysis, information for several Big data concerns, and streams and batches
tracking extraction, and streaming the creation and implementation of
data mining apps
Unification of services Yes Yes Yes Yes
Location of “Deployment” “Zonal” “Zonal” “Regional” “Regional”
Component of Operation “EC2 cluster” “Cluster” “Cluster” “Cluster”
Scalability of Computing “Manual” “Manual” “Manual” “Manual”
operations
Pricing Structure for Hourly basis payment Payments per second Hourly basis payment Minute based payments
Compute Nodes
(Continued)
Intelligent Analytics for Industry 4.0 Applications
TABLE 7.2 (Continued)
Frameworks for “Cloud-Based Big Data” Enterprise Solutions [25]
Architecture and
Sl. No. Operations “AWS” “GCP” “IBM Cloud” “MS Azure”
[-5ex]2 “Big Data Storage” “Amazon S3” “Google Cloud Storage” “IBM Cloud Object Storage” “Azure Blob Storage”

Form of Memory Archiving of Distributed Archiving of Distributed Archiving of Distributed Entities Archiving of Distributed
Entities Entities Entities
Kinds of entities “Object” “Object” “Object” Blobs can be appended or
blocked
Component for Deployment “Bucket” “Bucket” “Bucket” “Container”
Identity for Deployment Code that is globally unique Code that is globally Unique identifier for everyone Distinctive Identifier at
unique the Account Level
Intelligent Analytics in Big Data and Cloud

Scalability of Storage Automatic scaling Automatic scaling Manual scaling Manual scaling
Location of “Deployment” “Regional” “Regional and “Regional and multiregional” “Zonal” and “Regional”
multiregional”
Metadata for Entities “Yes” “Yes” “Yes” “Yes”
Layout for Information “Any” “CSV, JSON (Newline “Any” “XML”
Storage Delimited Only), Google
Cloud Datastore Backups,
Avro”
Information Storage Capacity Unrestricted Restricted Unrestricted Restricted
Configuration Management Yes Yes Yes No
of Entities
Memory Types “Amazon Glacier”, “Nearline”, “Coldline”, “Standard”, Crypt, as well as Cold Duplicate Categories:
“Standard”, Irregular “Regional”, Crypt Connectivity, Flex Category “LRS”, “GRS”, “ZRS”,
Connectivity “Multiregional” (Varying information) “RA-GRS” Tiers: Hot,
Cool
95

(Continued)
 96

TABLE 7.2 (Continued)

Frameworks for “Cloud-Based Big Data” Enterprise Solutions [25]
Architecture and
Sl. No. Operations “AWS” “GCP” “IBM Cloud” “MS Azure”
Object Versioning Yes, Automatic Yes, Automatic Yes, Manual (Private Cloud Yes, Manual
Implementation), No (Dedicated/
public cloud implementation)
[-5ex]3 “Big Data Warehouse” “Amazon Redshift” “Google BigQuery” “IBM Db2 Warehouse on Cloud” “Azure SQL Data
Warehouse”
Component of Deployment “Cluster” (EC2) Completely regulated “Cluster” “Massively Parallel
Processor” (MPP)
Quantify Scalability Yes, Automated Yes, Automated Yes, Automated Yes, Physical
Allowable range of Compute 1–128 NA, Fully Managed 1–60 (1 head node, 50+ information 1–60
Nodes entities)
Location of Deployment “Region” “Region” “Region” “Region”
Format of storage “Columnar” “Columnar” “Columnar” “Columnar”
Origins of information “Amazon S3”, “Amazon “Google Cloud Storage”, Information file on network, Azure Blob Storage
Dynamo DB”, “Amazon “Google Cloud information stores like Amazon S3
EMR”, “AWS” information Dataflow”, readable data or IBM Cloud Object Storage, and
Channel sources Db2® server
Techniques of Information “COPY” from S3, “Streams” “Google Analytics” Read from the cloud, through a disk Information loading using
loading from “Amazon Kinesis Professional, Live Gps, Twitter information, and “PolyBase”,
Firhose” Uploading, Batch publicly available information can “SQLBulkCopy” Interface,
Streaming all be loaded SSIS (SQL Server
Integration Service) with
BCP Function, with Azure
Information Warehouse
“Query Language” “PostgreSQL” “Standard SQL (Beta), “SQL Reference, CLP+, SQL PL, “PolyBase T-SQL”
Intelligent Analytics for Industry 4.0 Applications

Legacy BigQuery SQL” PL/SQL”

Intelligent Analytics in Big Data and Cloud 97

7.3.3  “Azure SQL Data Warehouse (Big Data Warehouse Solution)”

“SQL Data Warehouse” is a platform for substantially parallelization using SQL
insights that allows for flexible and autonomous computation and storage scaling, as
well as easy interaction utilizing big data repositories to create a center for cube and
data stores. Unlimited scalability, adaptability as well as extension, high cybersecu-
rity, and interoperability with Microsoft as well as other major manufacturers using
other core technologies are among its advantages and benefits of use [30,38,39].

7.4 ASSESSMENT OF CLOUD-CENTERED “BIG

DATA” BUSINESS SOLUTIONS DESIGNS
Table 7.2 lists the relative study of the summarized commodities, in which compari-
sons are made across relevant products within a sector, such as AWS’s chosen “big
data analytics” tool versus big data analytics tools from the other three companies.
The objective of this comparison is not to disparage any of the cloud-based “big data”
tools discussed. The goal is to show that each has a substantial application in many
areas.

7.5 REAL-WORLD CASE STUDIES IN BIG DATA

AND CLOUD-BASED ANALYTICS
This section discusses some relevant real-world case studies in the domain of big
data and cloud-based analytics.

7.5.1 Big Data Case Study-1 – “Walmart”

With more than 2 million workers and 20,000 locations in 28 countries, “Walmart” is
the nation’s biggest retail and revenue-generating enterprise. It began using “big data
analytics” before the phrase “big data” was introduced. “Walmart” utilizes infor-
mation analysis to discover trends that could be utilized to make product sugges-
tions to customers based on which items were combined. “WalMart” has boosted its
user exchange value by implementing effective “Data Mining”. It has been increas-
ing “big data analysis” to produce the best e-commerce technologies and improve
customer service. The main aim of “Walmart’s” massive information space is to
improve services’ retail experience while in a “Walmart store”. “Walmart’s” big data
services were created with the objective of rebuilding international networks and cre-
ating innovative products to personalize clients’ retail experiences while increasing
operational efficiencies. Internally, consumers have access to real-time information
recorded from many resources and organized for efficient usage using “Hadoop” and
“NoSQL” technologies [41].

7.5.2 Big Data Case Study 2 – “Uber”

Whenever it comes to moving passengers, including delivering packages, “Uber” is
the best pick for millions of individuals. It utilizes the person’s private data to track
98 Intelligent Analytics for Industry 4.0 Applications

which parts of the business are most frequently used, analyze usage ­statistics, and
identify areas in which the products would also be targeted more. “Uber” focuses on
the market forces of goods that cause the cost of services to fluctuate. As a result,
dynamic pricing is among Uber’s most essential information applications. For exam-
ple, if you are late for an appointment and reserve a vehicle in a congested area,
you should consider paying double as much. For instance, the traveling expense of
1 mile to a festival could have a high range. Dynamic pricing and economic circum-
stances in the near future could be the difference between preserving and retaining
­consumers. Machine learning (ML) methods are used to assess where there is a large
market [41].

7.5.3 Big Data Case Study 3 – “Netflix”

This is the most popular American media firm, focusing on offering consumers
Internet video on-demand streaming. “Netflix” has set its sights on using big data
to forecast what its users will prefer to watch. As a consequence, “big data analyt-
ics” operates as the fuel for the “recommendation engine” designed for this pur-
pose. “Netflix” has recently begun portraying itself as a content producer rather
than merely a consumer. This technique has, presumably, been heavily influenced
by statistics. Data elements such as what titles users view, the number of times view-
ing is halted, and ratings are provided are input into “Netflix’s” recommendation
algorithms, and fresh content choices. “Hadoop”, “Hive”, and “Pig” are among the
firm’s data structures, along with a variety of certain other standard analytics tools.
“Netflix” demonstrates that understanding precisely what customers want is simple if
businesses do not even assume and instead use big data to make decisions.

7.5.4 Big Data Case Study 4 – “eBay”

As a data-intensive company, “eBay” faces a significant technological barrier in
implementing measures that could quickly analyze and process the information as it
emerges (streaming data). Several ways of supporting stream processing interpreta-
tion have been developed rapidly. “eBay” uses a variety of tools, notably “Apache
Spark”, “Storm”, and “Kafka”. It enables the firm’s technical experts to explore for
informational markers connected with the information (schema) and allow access to
as many individuals as possible while maintaining the additional level of protection
and privileges. The firm is now at the center of things for implementing big data tech-
nologies and plays a significant role in the open-source community [41].

7.5.5 Big Data Case Study 5 – “Procter & Gamble”

“Procter & Gamble” (P&G) is a 170+-year-old corporation providing commodities
used by many. The brilliant corporation saw the potential of big data and applied
it in functional departments globally. P&G has placed a heavy focus on leveraging
big data to make better, faster company choices. International Services Companies
have created products, techniques, and procedures that cross organizational bound-
aries and provide instant access to the most up-to-date data with advanced insights.
Intelligent Analytics in Big Data and Cloud 99

As a result, despite numerous new firms, P&G, as the oldest, continues to have a large
market share [41].

7.5.6  Case Study 6: Big Data Predicting the Uncertainties

A pioneering study in Bangladesh discovered that tracking movements of people
throughout the region utilizing information via mobile telephony can assist anticipate
where infectious diseases like malaria are likely to happen, allowing healthcare com-
panies to take precautionary actions. “Malaria” kills around 4 lakh people worldwide
annually, the majority of which are youngsters. The information, such as details from
the Bangladesh health ministry, is being used to generate risk maps, which exhibit
the likely areas of “malaria” outbreaks, letting regional health authorities get work-
able safeguards like the spritzing pesticides and herbicides and amassing mosquito
nets as well as medications to safeguard the population from the illness [41].

7.5.7 Deployment of “Big Data” for Ad Network Assessment

in more Than Ten Regions of the World [42]

In this case study, the client was a well-known market research firm. The main issue
was that, although possessing a reliable statistical platform, the client feared this
would not be able to meet the firm’s expected demand. Recognizing the predicament,
the client had been on the lookout for a forward-thinking, inventive remedy. The aim
of this system was to deal with the ever-increasing volumes of information, analyze
large information quickly, and provide full marketing network analytics. The client
was looking for a professional and experienced group to execute the change after
settling on the platform’s design. After a lengthy relationship with “ScienceSoft”,
the client asked the specialists to handle the whole transfer from the legacy to the
new analytical model. The solution for the above is summarized as follows. The cli-
ent’s predictive analytics designers teamed up with “ScienceSoft’s big data” group
throughout the work. The client team created an idea, while the “ScienceSoft” team
was in charge of putting it into action. The client’s designers chose the ensuing com-
ponents for the new analytical mechanism:

• pache Hadoop” – for information storage.

• pache Hive” – for consolidation, querying, and assessment of information.
• pache Spark” – for information processing.
• As “cloud computing systems”, “Amazon Web Services”, and “Microsoft
Azure” were chosen.

At the client’s request, the old system and the newer one ran in tandem during the
transition. The following important components were incorporated into the actual
approach:

• Compilation of information.
• taging”.
• ata warehouse-1”.
100 Intelligent Analytics for Industry 4.0 Applications

• ata warehouse-2”.
• esktop application”.

Collected information was sent to the platform from a multitude of different sources,
including television viewing, portable phone browser habits, and Internet traffic
statistics, as well as questionnaires. Information gathering included the following
Python-coded phases to allow the technology to handle over 1,000 distinct categories
of original information (cache, Excel, Text, and so forth):

• Information transition.
• Processing of information.
• Information fusion.
• Information uploaded into the database.
• taging”: the foundation of such a component was “Apache Hive”. The infor-
mation format was comparable to raw database schema at the time, and
there were no documented links among participants from various channels,
such as television and the Internet.
• ata warehouse-1”: this block is built on “Apache Hive”, the same as the
preceding one. Information mapping occurred here. As per the modeling
guidelines, the computer analyzed the responders’ information for audio,
television, web, and paper resources, and connected consumers’ IDs
across multiple data sources. “Python” was used to write the ETL for that
component.
• ata warehouse-2”: utilizing “Apache Hive” and “Spark” at its foundation,
the module ensured real-time data analytics based on business rationale:
summation, averaging, likelihood, and so on. The desktop app’s SQL state-
ments were processed using “Spark’s” Data Frames. “Scala” was used to
create ETL. “Spark” also enabled exhaustive data to be filtered based on the
program’s customers’ authorized access.
• esktop application”: the new approach allowed for cross-analysis of nearly
half a million features and the creation of intersecting vectors that allowed
for cross business intelligence from multiple sectors. The client could pro-
duce informal results in combination with conventional statistics. The plat-
form provided a speedy reply of convenient graphs once the client picked
numerous factors of relevance (for instance, a specific television network, a
class of consumers, or a specific time of day). The prediction may also assist
the consumer. The program, for instance, might anticipate sales compared
to projected reach and anticipated advertising expenditure.
• Findings: the proposed program could handle some requests approximately
100 times quicker than the previous approach at the program’s conclusion.
The client was enabled to do complete promotional campaign research for
several marketplaces using the insightful information gained from the study
of around 1 million parameters.
• Systems and Applications: “Apache Hadoop”, “Apache Hive”, “Apache
Spark”, “Python (ETL)”, “Scala (Spark, ETL)”, “SQL (ETL)”, “Amazon
Web Services (cloud storage)”, “Microsoft Azure (cloud storage)”, “.NET
(desktop application)”.
Intelligent Analytics in Big Data and Cloud 101

7.5.8 Case Study 8: Design of an IoT Pet

Monitor Using Big Data Strategy
In this case study, the client was a corporation located in the United States that oper-
ates in 18 states. Having extensive telecom expertise, the client chose to develop
a new platform that allows pet parents to follow their animals’ movements using
wearables sensors controlled through a smartphone app. The client’s objective was
to find a big data solution, which might help customers to constantly know where
their animals were, receive real-time warnings regarding key occurrences, and
obtain statistics on their animals’ whereabouts. The idea was to allow for the shar-
ing of audio-visual content (voice, film, and images) such that animal parents may
communicate with their animals and know where they have been at any given time.

FIGURE 7.3  Big data deployment on the web.

102 Intelligent Analytics for Industry 4.0 Applications

Because the client anticipated the number of customers to continue to rise, the s­ ystem
needed to be highly adaptable to collect and organize more content [43]. To make
sure that the ­system is easily customizable, “ScienceSoft’s” big data team deployed
it on the web and built it around “Apache Kafka”, “Apache Spark”, and “MongoDB”
(Figure 7.3).
Utilizing the “MQTT” protocol, many global positioning send actual informa-
tion on pet whereabouts as well as occurrences (for example, low power and exiting
a secure zone) toward the mail server. The standard was selected since it ensures
a gadget-friendly interface while also saving battery capacity on cell devices. An
“Apache Kafka”-based dynamic data synthesizer pulls information from several
“MQTT” components, analyzes it instantaneously, and verifies the quality of con-
tent. The “Kafka Streams” module enables push alerts and assures secure informa-
tion transport.
An information integrator built on “Apache Spark” analyzes stored informa-
tion, then combines it by hourly, daily, weekly, and monthly before sending it to an
information store. The “ScienceSoft’s” team recommended “MongoDB” knowledge
for the latter since it enables time-series data occurrences to be stored as a single
document on a daily, hourly, and monthly basis. Furthermore, the report-oriented
design permitted in-place modifications, resulting in significant efficiency gains.
Customers’ identities, credentials, and administrative information were stored in the
existing repository on “PostgreSQL RDS”. “RESTful” services decouple the user
interface from the information storage and also ensure dependability, adaptability,
and system or coding environment independence.
The client got a robust big data solution, which could manage 30,000+ transac-
tions every moment across 1 million sensors. Therefore, users could monitor their
pets’ whereabouts instantaneously as well as transmit and receive images, movies,
and voicemail messages. The consumer gets text alerts if a significant incident occurs
(for example, a pet breached a geographical boundary created by the animal parent or
the animal’s wearable tracker went “out of communication”). Animal holders could
likewise view hourly, quarterly, or periodic statements that are generated routinely,
or adjust the recording phase individually if necessary. The technology devices used
in the project were: “Amazon Web Services”, “MQTT”, “Apache Kafka” (stream
data processor), “Apache Spark” (data aggregator), “MongoDB” (data warehouse),
“PostgreSQL RDS” (operational database), and “RESTful” web services.

7.5.9 Case Study 9: Development of a Telecommunication

Company’s Data Analytics Solution [44]
In this case study, the client was a Texas-based telecommunication organization that
participated in the government Lifeline Recovery Program and provided low-income
persons with pre-paid cellular phones with service packages. The “ScienceSoft’s”
analytics group was tasked with designing and implementing a data processing and
analytics tool that would allow the client to gather data from numerous sources and
gain insights into consumer behavior. The system had to be able to examine historic
information and analysis. An additional challenge to resolve was access permissions,
as the client intended to give their renters exposure to rental statistics.
Intelligent Analytics in Big Data and Cloud 103

Actual data (including user views and buttons, pricing packages, screen sizes,
and applications loaded) were collected across 10+ providers by the data analytics
solution. The “MQTT” protocol was proposed by “ScienceSoft’s” team to capture
the sensor information as well as transfer it into “Apache Kafka”. To save money
on AWS computational power, the team recommended adopting “Amazon” Spot
Instances. Researchers deployed AWS Application Proxy Balancers to guarantee the
analytic game’s adaptability.
“Apache Kafka” served as a medium for streaming data. The gathered informa-
tion was sorted there before being offloaded onto the “Amazon Simple Storage”
Service landing zone. “Amazon Redshift” was selected for information warehousing
and distribution, wherein telemetry information from Android smartphones, as well
as statistics from “Enterprise Resource Planning” and the “Home Location Register”
(HLR), were sent.
The “Science Soft” created ROLAP modules with 30+ measurements and 10+
statistics to facilitate periodic and ad-hoc analysis. For example, the analytic system
provided an overview of the loyalty points gained by measuring a user’s advertise-
ment exposures as well as click-throughs. The client might infer that the client was
unsatisfied with the product centered on the enhanced total of requests to assist.
Without action, this might result in client attrition.
The system was made available not only to the clients but also to their ­renters
­(including telecommunication firms with their own clients and HLRs). A renter,
for example, could view the portion of statistics that pertains to their business. The
“ScienceSoft’s” team devised two methods to do this: public entry (managed at the
data storage tier) and exclusive accessibility using a dedicated AWS account. The client
was able to demonstrate the ability with “ScienceSoft’s” big data solutions as follows:

• Determine how engaged a customer is and what their interests are.

• Recognize patterns in consumer’s behavior.
• Do forecasting regarding how consumers will act.
• Marketers will be invoiced depending on their determined proportion.
• Employ predictive analytics to your advantage. For instance, current earn-
ings, the rate of new subscribers, and customer support statistics, among
other things.
• The client was able to save 80% on AWS computational resources by using
“Amazon Spot Instances”.

The tools used in this case study were “Amazon Web Services (Amazon cloud)”,
“Apache Kafka” for streaming live information, the “Message Queuing Telemetry
Transport” Protocol, “Amazon Simple Storage Service” for continuous ­storage,
“Amazon Redshift” as a “data warehouse”, “Airbnb Airflow”, and “Python (ETL)”.

7.5.10 Case Study 10: Implementation of Data Analytics

for a Multibusiness Organization [45]

The client in this case study was an American-based firm operating a multichannel
commerce, lodging, cafe, and other enterprises. The client’s key goal was to increase
104 Intelligent Analytics for Industry 4.0 Applications

client retention through a personalized technique while also optimizing intrinsic

retail operations. They could not do it, though, because the information was locked
away in numerous programs that were specialized to their operations strategy.
Because the software’s goal was to satisfy all of the company’s business instructions
by collecting and aggregating information from 10+ information resources, including
“CRM”, “Magento”, and “Google Analytics”, as well as devoted restaurant, eatery, and
wellness mechanisms, “ScienceSoft’s” team initially supplied the client with a proto-
type and then formulated a series of test reports centered on the client’s ERP statistics.
The group at “ScienceSoft” developed high-level design modules along with their
primary tasks. The analytic solution is available with scalability in mind. The goal
was to evaluate historical information for 5 years at first, and then to handle informa-
tion expansion in the future. Because the client was concerned about the security of
information stored in a private cloud in a data center, the approach was blended.
“ScienceSoft” suggested a technological architecture, which would meet the
client’s solution requirements, like adaptability, speed, and dependability both for
desktop and mobile consumers. They evaluated whether these technologies and the
accompanying “Microsoft stack” were viable for the possible answer because the
client always had certain old systems operating on “Microsoft SQL Server”. This
might enable the consumers to cut deployment expenses as only a few extra licenses
would be needed.
“ScienceSoft” additionally recommended that the workaround be enhanced with
sophisticated analytics abilities depending on user experience of the retail industry.
They produced a prototype for a recommender system as the forecasting algorithm
underlying the machine was to increase the user’s cross-selling and the up-selling
prospects for the virtual retailer, and a “time-series” projection prototype to estimate
profits, for that.
The elements of the developed analytical solution were as follows:

• A data center that stores regular and semi-structured information from 15

diverse information sources.
• ETL (extract, transform, and load) operations number in the hundreds.
• A “data warehouse” to merge and integrate information.
• A statistical system with five OLAP cubes and around 60 parameters.
• Generation of reports.
• To guarantee that the client may gain through intermediate outputs, the inte-
gration was broken into many phases. They created around 90 modules for
the client’s various enterprise directions and consumer management in total.
Because the information assimilation from different systems is w ­ orthless
without a well-defined information continual improvement procedure, The
“ScienceSoft’s” team devised ETL guidelines that were designed to
– Combine master information from several systems, such as client
profiles.
– Consolidate information into a single format.

The client’s information management procedure was substantially automated thanks

to these guidelines. Administrative manipulations by an information administrator,
on the other hand, were mostly conceivable.
Intelligent Analytics in Big Data and Cloud 105

“ScienceSoft” also expanded on access control restrictions to maintain data

protection. Just before the program started, they examined the client’s extremely
customizable access model and determined that this would be unacceptable, since
it would adversely impair the statistical optimizer efficiency (i.e., the system could
take too much time to generate the needed statistics). As a result, they suggested a
less sophisticated but still very effective three-level access approach (for a particular
enterprise, division, or personnel). The ideas and innovations had no detrimental
effects on the program’s operation.
“ScienceSoft’s” staff offered extensive assistance to the client as part of the sup-
plied business intelligence solutions. For instance, once the client’s third-party ana-
lytical suppliers made some adjustments, they offered training on installing and
dealing with “OLAP cubes” and updated “ETL” procedures.
The client benefitted from a 360° consumer point of view throughout all chan-
nels and company objectives, and strong retail intelligence, which helped in creating
a personalized customer experience, thanks to the built intelligence platform. By
enhancing inventory levels and evaluating employees, the client also was able to
optimize the inner company’s operations.
After integrating all the information, the client was empowered to

• Examine the consumers’ purchasing behavior and patterns.

• Examine the customers’ recent behavior, regularity, and financial worth.
• Determine their most important consumers.

The following data were analyzed by the client:

• The most/least frequented sites, sites with really no traffic, and pages with
traffic jams and low conversions are all examples of traffic and conversion
percentages.
• Participation of Internet shopping consumers.
• Merchandise wish lists, revenues, as well as basket dumping.

Rather than relying on a data repository to show the inventory levels and the need
for continual corrections over the telephone, the client was able to monitor the real
stock position at both the facility and the stores in real time. Such inventory stock
visibility has had a favorable impact on the purchasing and logistical operations. The
client was able to define the overall quality of the workers’ activity using KPIs and
goal management reports. “Microsoft SQL Server”, “Microsoft SQL Analysis” and
“Integration Services”, “Python”, and “Microsoft Power BI” were used.

7.5.11 Case Study 11: Big Data Consulting and

Training for a Satellite Agency [46]
The client in this case study was a geospatial organization that gathers and analyzes
meteorological and environmental data. The difficulty was that the client has a well-
planned training program, pursuant whereby the organization intends to start a new
project involving satellites geosynchronous. They would be gathering and relaying
106 Intelligent Analytics for Industry 4.0 Applications

meteorological, temperature, and ecological survey data. To meet the deadline, the
client hired a contractor ahead of time to develop the structure of the future ­solution
and select the software system. Because the program is designed to last three decades
and having the right design is critical, the client opted to hire other firms to review
the documents that have been acquired and selected “ScienceSoft” for the job. Those
who also asked “ScienceSoft” to develop and execute training courses on the big
data environment in general and the recommended solutions in specific, as the work-
around depends on big data.
“ScienceSoft’s” big data consultancy group starts by visiting the client on website
to know about their company and the product they were considering. They also held
group discussions with client’s staff. Our specialists then dug into the offered infor-
mation on the recommended design and computing environment. In much less than
a week, our crew had to consider carefully 1,000+ pages, highlight the good aspects,
and, if necessary, bring out whatever omissions or flaws.
The staff at “ScienceSoft” has created educational materials that will be provided
over the next initial appointment. The experts provide a summary of current big
data technologies to expose the client to the big data ecosystem going far beyond
those recommended by the third-party contractor. The presentation also includes
application scenarios, with an emphasis on those that would be relevant to the cli-
ent’s industry. The experts designed entirely different sets of learning resources for
every potential customer because “ScienceSoft’s” team was really to conduct train-
ing courses both for senior knowledge and professional staff.
The “ScienceSoft’s” team was prepared not only to lead seminars but also to cre-
ate experience and understanding resources and offer advice on how to evaluate test
takers. Nevertheless, the client preferred a conference approach with Q&A sessions.
Besides that, “ScienceSoft” offered technical workers with on-the-job assistance,
with the experts conducting coding workshops during the second onsite visit.
The client gets an assessment report detailing the optimizer’s strengths and weak-
nesses. The study also included suggestions for how to enhance the system. For
instance, substituting one of the recommended technologies with the other resulted
in improved quality and durability for the system. “ScienceSoft” has created over
300+ slides of training manuals that present the big data idea, technology, and real-
world experiences. NoSQL technology, and broadcasting and queuing analysis, were
one of the subjects addressed. Seminars with panel discussions and on-the-job help
were used to practice regularly. After the session, the client gets all of the documents,
which became their intangible asset.
Q&A meetings, quantitative performance evaluations (accessibility, integration,
expandability, serviceability, ease of maintenance, vendor support, and achieve-
ment), and remedy evaluations related to technological features (consumer experi-
ence, technology maturity, digital innovation, and supplier assistance) were some of
the methodologies used in this case study.

7.5.12 Case Study 12: Big Data Consulting for a

Leading Internet of Vehicles Company [47]
The client is a transportation and investment management company, which is one of
the biggest Internet-of-things firms in the European Union (EU). To allow customers
Intelligent Analytics in Big Data and Cloud 107

of their services to make informed decisions, the client gathers, saves, and analyzes
IoT data from 1 million automobiles linked to their platforms. The client sought to
enhance their big data analytics abilities after implementing a comprehensive IoT
information gathering and storage system. “Apache Cassandra” is used in their main
approach. “Apache Cassandra”, described as the key, does not build sustainable large
amounts of data analytics very well. To download information for 70,000 vehicles,
for instance, the client must dynamic responses 70,000 distinct queries, each of
which will receive 70,000 unique reports, necessitating more attempts to obtain a
comprehensive picture.
The big data specialists visited the client for a 3-day onsite visit, where they
thoroughly analyzed the existing schema, including its design, relevant documen-
tation, current evidence streams, and existing data standard operating procedures.
Following that, we organized a workshop devoted to the future solution, wherein they
addressed the optimizer’s estimated launch date, current and chosen technology, and
accessible licenses. They additionally established structurally required specifications
for the optimizer’s reliability, speed, privacy, and adaptability during the session.
Next, they classified every demand as crucial, moderate, mid, or minor in terms of its
influence on the firm’s success.
The experts created the design concept for the future solution based on the conclu-
sions of the session. They discussed the responsibilities of high-level design modules
such as continuous integration as well as a cloud service. Following the presentation
of the design approach, the experts conducted a Conversation during which they
offered definitive solutions to any questions the client had. For example, they dis-
cussed the benefits and drawbacks of a datastore vs a centralized data warehouse.
They further contrasted on-premises deployment with cloud-based solutions, cover-
ing several public scenarios – including “Amazon Web Services” and “Microsoft
Azure” at the forefront. Furthermore, the “Apache Cassandra” consultants gave the
client ideas about how to increase “Apache Cassandra’s” performance, including
table layout, partitioning variables, and the type of information that can be stored.
The following information was included in the client’s visit document:

• Architecturally significant criteria for the future solution, as well as their

business impact. The responsibilities of essential architectures are outlined
at a high level in this high-level design.
• The client gets detailed responses to their inquiries about the prospective
solution’s deployment, the benefits and drawbacks of various technologies
and architectural aspects, the complexities of disk space and connection, and
other key topics. Analysis of business needs, function breakdown, workshop
on structurally critical aspects, Q&A session, and technology comparison.

7.6  CHALLENGES AND ISSUES

Certain hurdles must be addressed in order to progress above current ML and big data
approaches and methods. The following criteria are viewed as crucial by NESSI [20].

• A real scientific foundation must be established before selecting an appro-

priate approach or concept.
108 Intelligent Analytics for Industry 4.0 Applications

• Novel scalable and efficient methods must be created.

• Suitable development abilities and technology frameworks must be devel-
oped and implemented to properly apply conceived remedies.
• Finally, the solutions’ value creation must be investigated just as thoroughly
as the data format and accessibility.

Ongoing problems, such as acceptance and application of quality big data solutions
employing cloud infrastructure, as well as addressing vulnerabilities, arise in the
context of virtualized “big data analytics”. One of the most pressing challenges
when combining big data analytics with cloud technology is security. This is why
­cloud-based big data analytics, as well as its application in practice and execution,
has gotten so much interest.
The researchers summarized, assessed, and compared authenticator-based infor-
mation integrity testing strategies for “cloud” and “Internet-of-things” data [48]. Any
further advancements in this field, according to this article, should consider three
major facets: effectiveness, reliability, and expandability. On “cloud” and “Internet-
of-things” information, the authors [48] give an overview, evaluation, and com-
parison of authenticator-based file integrity verification methodologies. Any future
advancements in this field, according to this article, should consider three primary
aspects: efficiency, reliability, and expandability.
The authors [49] advise that the following topics must be investigated further:

• Hierarchical programming or extensible high-level methodology.

• Interchange of information and technological solutions.
• Combining several big data analytics platforms.
• Proprietary data mining methods.

In a decentralized computing environment known as fog computing, data, compute,

storage, and applications are spread between the information source and the cloud.
A decentralized computing system called edge computing puts enterprise applica-
tions closer to sources of data like IoT gadgets or regional edge servers. Strong com-
mercial advantages, including quicker insights, quicker response times, and greater
bandwidth availability, can result from being close to the source of the information.
Durability, complexity, dynamicity, diversity, delay, and privacy are a few issues with
fog computing that are related to big data. The issues of edge computing related to
big data include storage costs, lost information, and privacy.
Big data peregrination is a long journey. It involves network hurdles and interdis-
ciplinary interactions, dealing with expensive data to name a few [53].

7.7 CONCLUSION
The advent of this subject of study has grabbed the attention of several profession-
als and scholars in this age of big data. Big data analytics and analysis have grown
increasingly important as the pace at which information is generated in the virtual
environment has increased. Furthermore, most of this information is already on the
Intelligent Analytics in Big Data and Cloud 109

cloud. As a result, relocating “big data analytics” to the “cloud” is a feasible choice.
Furthermore, the cloud structure meets the data analytics techniques’ storing and
processing needs. On the other hand, there are conflicts such as safety, confidential-
ity, and a loss of ownership and control. This chapter surveyed some of the important
aspects of intelligent analytics in big data and cloud. Several significant tools in “big
data” and “cloud” are compared based on their performance. Numerous real-world
case studies (12 in particular) on intelligent analytics in big data and cloud are dis-
cussed in detail. The challenges and issues faced while developing big data analytics
in cloud along with future research directions are summarized in this chapter.

7.8  FUTURE RESEARCH DIRECTIONS

There are a variety of open-source data mining approaches, tools, and applica-
tions. “R”, “Gate”, “Rapid-Miner”, and “Weka” are just a few of the tools available.
“Cloud-based big data analytics” solutions must have the permission to connect these
cheaper data analyses on the cloud system to provide expense and quality service.
The availability, expense, and convenience of setup and evaluation of cloud-based
statistics are the primary reasons for its popularity. Considering this, researchers [50]
outlined the following important study guidelines:

• Statistics and data management have evolved in relation to cloud-based

analytics.
• Adapting and evolving approaches and approaches to boost efficiencies and
minimize hazards.
• Develop plans and tactics to address security and privacy issues.
• Examining and adapting ethical and legal practices to reflect shifting per-
spectives, impacts, and consequences of technological advancements in this
domain.

The study possibilities, however, really are not confined to the aforementioned points.
The primary objective would be to shift the cloud from an information management
and infrastructure base to a flexible and scalable analytics solution.

REFERENCES
1. Khan, S., Shakil, K. A., and Alam, M. (2008). Cloud-Based Big Data Analytics-A
Survey of Current Research and Future Directions. Advances in Intelligent Systems
and Computing, 654. https://doi.org/10.1007/978-981-10-6620-7-57
2. Chen, C. L. P. and Zhang, C. Y. (2014). Data-Intensive Applications, Challenges,
Techniques, and Technologies: A Survey on Big Data. Information Sciences. Advances
in Intelligent Systems and Computing, 275, 314–347. https://doi.org/10.1016/j.
ins.2014.01.015
3. Driscoll, A., Daugelaite, J., and Sleator, R. D. (2013). ‘Big Data’, Hadoop and Cloud
Computing in Genomics. Journal of Biomedical Informatics, 46(5), 774–781. https://
doi.org/10.1016/j.jbi.2013.07.001
4. Manekar, A. and Pradeepini, G. (2015). A Review on Cloud-based Big Data Analytics.
ICSES Journal on Computer Networks and Communication, 1(1), 1–4.
110 Intelligent Analytics for Industry 4.0 Applications

5. Assuncao, M. D., Calheiros, R. N., Bianchi, S., and Netto, M. A. S. (2015). Big
Data Computing and Clouds: Trends and Future Directions. Journal of Parallel and
Distributed Computing, 79, 3–15. https://doi.org/10.1016/j.jpdc.2014.08.003
6. Wu, X., Zhu, X., Wu, G.-Q., and Ding, W. (2014). Data Mining with Big Data.
IEEE Transactions on Knowledge and Data Engineering, 26(1), 97–107. https://doi.
org/10.1109/TKDE.2013.109
7. Pashazadeh, A. and Jafari Navimipour, N. (2018). Big Data Handling Mechanisms
in the Healthcare Applications: A Comprehensive and Systematic Literature Review.
Journal of Biomedical Informatics, 82, 47–62. https://doi.org/10.1016/j.jbi.2018.03.014
8. Challita, S., Zalila, F., Gourdin, C., and Merle, P. (2018). A Precise Model for Google
Cloud Platform. IEEE International Conference on Cloud Engineering, 177–183.
https://doi.org/10.1109/IC2E.2018.00041
9. Shakil, K. A., Sethi, S., and Alam, M. (2015). An Effective Framework for Managing
University Data Using a Cloud Based Environment. 2nd International Conference on
Computing for Sustainable Global Development (INDIACom), 1262–1266.
10. Jackson, J. C., Vijayakumar, V., Quadir, M. A., and Bharathi, C. (2015). Survey on
Programming Models and Environments for Cluster, Cloud and Grid Computing That
Defends Big Data. 2nd International Symposium on Big Data and Cloud Computing,
517–523.
11. Neaga, I. and Hao, Y. (2014). A Holistic Analysis of Cloud Based Big Data Mining.
International Journal of Knowledge, Innovation and Entrepreneurship, 2(2), 56–64.
12. Borthakur, D., Gray, J., Sarma, J. S., Muthukkaruppan, K., Spiegelberg, N., Kuang,
H., Ranganathan, K., Molkov, D., Menon, A., Rash, S., Schmidt, R., and Aiyer, A.
(2011). Apache Hadoop Goes Real-Time at Facebook. ACM SIGMOD International
Conference on Management of Data, 1071–1080.
13. Strambei, C. (2012). OLAP Services on Cloud Architecture. Journal of Software and
Systems Development. https://doi.org/10.5171/2012.840273
14. Khan, I., Naqvi, S. K., Alam, M., and Rizvi, S. N. A. (2015). Data Model for Big Data
in Cloud Environment. Computing for Sustainable Global Development (INDIACom),
582–585.
15. Ortiz, J. L. R., Oneto, L., and Anguita, D. (2015). Big Data Analytics in the Cloud:
Spark on Hadoop vs MPI/OpenMP. INNS Conference on Big Data, 53, 121–130.
16. Baker, T., Al-Dawsari, B., Tawfik, H., Reid, D., and Nyogo, Y. (2015). GreeDi:
An Energy Efficient Routing Algorithm for Big Data on Cloud. Ad Hoc Networks, 35,
1–14.
17. Li, X., Calheiros, R. N., Lu, S., Wang, L., Palit, H., Zheng, Q., and Buyya, R. (2012).
Design and Development of an Adaptive Workflow-Enabled Spatial-Temporal Analytics
Framework. 18th International Conference on Parallel and Distributed Systems,
862–867.
18. Calheiros, R. N., Vecchiola, C., Karunamoorthy, D., and Buyya, R. (2012). The Aneka
Platform and QoS-Driven Resource Provisioning for Elastic Applications on Hybrid
Clouds. Future Generation Computer Systems, 28(6), 861–870.
19. Kim, H., Abdelbaky, M., and Parashar, M. (2009). CometPortal: A Portal for Online
Risk Analytics Using CometCloud. 17th International Conference on Computer Theory
and Applications.
20. Chen, Q., Hsu, M., and Zeller, H. (2011). Experience in Continuous analytics as a
Service (CaaaS). 14th International Conference on Extending Database Technology,
509–514.
21. Srivastava, M. and Yadav, P. (2021). Build a Log Analytic Solution on AWS. 5th
International Conference on Information Systems and Computer Networks, 1–5.
https://doi.org/10.1109/ISCON52037.2021.9702374
Intelligent Analytics in Big Data and Cloud 111

22. Neumeyer, L., Robbins, B., Nair, A., and Kesari, A. (2010). S4: Distributed Stream
Computing Platform. IEEE International Conference on Data Mining, ICDM, 170–177.
https://doi.org/10.1109/ICDMW.2010.172
23. Biem, A., Bouillet, E., Feng, H., Ranganathan, A., Riabov, A., Verscheure, O.,
Koutsopoulos, H., and Moran, C. (2010). IBM InfoSphere Streams for Scalable, Real-
Time, Intelligent Transportation Services. ACM SIGMOD International Conference on
Management of Data, 1093–1104. https://doi.org/10.1145/1807167.1807291
24. Zhang, Z., Liu, Z., Jiang, Q., Chen, J., and An, H. (2021). RDMA-Based Apache Storm
for High-Performance Stream Data Processing. International Journal of Parallel
Programming, 49, 671–684. https://doi.org/10.1007/s10766-021-00696–0.
25. Saif, S. and Wazir, S. (2018). Performance Analysis of Big Data and Cloud Computing
Techniques: A Survey. Procedia Computer Science, 132, 118–127. https://doi.
org/10.1016/j.procs.2018.05.172
26. Kumar, V. D. A., Divakar, H., and Gokul, R. (2017). Cloud Enabled Media Streaming
Using Amazon Web Services. IEEE International Conference on Smart Technologies
and Management for Computing, Communication, Controls, Energy and Materials
(ICSTM), 132, 118–127. https://doi.org/10.1016/j.procs.2018.05.172
27. Gulabani, S. (2017). Practical Amazon EC2, SQS, Kinesis, and S3 A Hands-On
Approach to AWS. eBook: Springer. https://doi.org/10.1007/978-1-4842-2841-8
28. Pradhananga, Y., Karande, S., and Karande, C. (2016). High Performance Analytics
of Bigdata with Dynamic and Optimized Hadoop Cluster. International Conference
on Advanced Communication Control and Computing Technologies (ICACCCT), 715–
720. https://doi.org/10.1109/ICACCCT.2016.7831733
29. Dawelbeit, O. and McCrindle, R. (2014). A Novel Cloud Based Elastic Framework for
Big Data Preprocessing. 6th Computer Science and Electronic Engineering Conference
(CEEC), 23–28. https://doi.org/10.1109/CEEC.2014.6958549
30. Ugia Gonzalez, J., and Krishnan, S. P. T (2015). Building Your Next Big Thing with
Google Cloud Platform: A Guide for Developers and Enterprise Architects. eBook:
Springer. https://doi.org/10.1007/978-1-4842-1004-8
31. Ramesh, B. (2015). Big Data Architecture. In: Mohanty, H., Bhuyan, P., and Chenthati,
D. (eds) Big Data. Studies in Big Data, 11. https://doi.org/10.1007/978-81-322-2494-5_2
32. Patil, A., Rangarao, D., Seipp, H., Lasota, M., Marcelo dos Santos, R., Markovic, R.,
Casey, S., Bollers, S., Gucer, V., Lin, A., Richardson, C., Rios, R., Van Alstine, R., and
Medlin, T. (2020). Cloud Object Storage as a Service: IBM Cloud Object Storage from
Theory to Practice - For developers, IT architects and IT specialists. IBM Redbooks.
ISBN: 9780738442457
33. Das, N. S., Usmani, M., and Jain, S. (2015). Implementation and Performance Evaluation
of Sentiment Analysis Web Application in Cloud Computing Using IBM Blue mix.
International Conference on Computing, Communication & Automation, 668–673.
https://doi.org/10.1109/CCAA.2015.7148458
34. Nakhimovsky, A. and Myers, T. (2004). Google, Amazon, and Beyond: Creating and
Consuming Web Services. eBook: Springer. https://doi.org/10.1109/CCAA.2015.7148458
35. Krishnan, S. P. T. and Ugia Gonzalez, J. L. (2015). Building Your Next Big Thing
with Google Cloud Platform: A Guide for Developers and Enterprise Architects eBook:
Springer. https://doi.org/10.1007/978-1-4842-1004-8_10
36. Serrano, N., Gallardo, G., and Hernantes, J. (2015). Infrastructure as a Service and
Cloud Technologies. IEEE Software, 32,2, 30–36. https://doi.org/10.1109/MS.2015.43
37. Shovic, J. C. (2021). Connecting an IoT Device to a Cloud Server: IoTPulse. Raspberry
Pi IoT Projects. Apress, Berkeley, CA. https://doi.org/10.1007/978-1-4842-6911-4_5
38. Soh, J., Copeland, M., Puca, A., and Harris, M. (2020). Microsoft Azure. Apress
Berkeley, https://doi.org/10.1007/978-1-4842-5958-0
112 Intelligent Analytics for Industry 4.0 Applications

39. Moemeka, E. (2019). Azure in the Enterprise: Cloud Architecture, Patterns, and
Microservices with Azure PaaS and IaaS. Springer.
40. Reagan (2018). Web Applications on Azure: Developing for Global Scale. Springer.
https://doi.org/10.1007/978-1-4842-2976-7
41. 5 Big Data Case Studies – How big companies use Big Data. Retrieved from https://
data-flair.training/blogs/big-data-case-studies/ on May 2, 2022.
42. Big Data Implementation for Advertising Channel Analysis in 10+ Countries. Retrieved
from https://www.scnsoft.com/case-studies/big-data-implementation-for-advertising-
channel-analysis on May 3, 2022.
43. Development of a Big Data Solution for IoT Pet Trackers. Retrieved from https://www.
scnsoft.com/case-studies/big-data-solution-for-iot-pet-trackers on May 3, 2022
44. Implementation of a Data Analytics Platform for a Telecom Company. Retrieved
from https://www.scnsoft.com/case-studies/data-analytics-platform-for-a-us-telecom-­
company-operating-in-18-states on May 4, 2022.
45. Data Analytics Implementation for a Multi-business Corporation. Retrieved from https://
www.scnsoft.com/case-studies/data-analytics-implementation-for-a-­multibusiness-
corporation on May 5, 2022.
46. Big Data Consulting and Training for a Satellite Agency. Retrieved from https://www.
scnsoft.com/case-studies/big-data-consulting-and-training-for-a-satellite-agency on
May 5, 2022.
47. Liu, C., Yang, C., Zhang, X., and Chen, J. (2015). External Integrity Verification for
Outsourced Big Data in cloud and IoT: A Big Picture. Future Generation Computer
System, 49, 58–67.
48. Talia, D. (2013). Clouds for Scalable Big Data Analytics. IEEE Computer Society,
46(5), 58–67. https://doi.org/10.1109/MC.2013.162
49. Neaga, I. and Hao, Y. (2014). A Holistic Analysis of Cloud Based Big Data Mining.
International Journal of Knowledge, Innovation and Entrepreneurship, 2(2), 56–64.
50. Saha, M., Panda, S.K., and Panigrahi, S (2021). A Hybrid Multi-Criteria Decision
Making Algorithm for Cloud Service Selection. International Journal of Information
Technology, 13, 1417–1422. https://doi.org/10.1007/s41870-021-00716-9
51. Saha, M., Kumar Panda, S., and Panigrahi, S. (2022). A Survey on Applications of
Multi-Attribute Decision Making Algorithms in Cloud Computing. ECS Transactions,
107(1), 12887–12900.
52. Pandit, P. (2021). Case Study on AWS Identity and User Management. https://doi.
org/10.13140/RG.2.2.18227.55844.
53. Behura, A. and Kumar Panda, S. (2022). Role of Machine Learning in Big Data
Peregrination. Handbook of Research for Big Data, pages 42.
 8 Various Audio
Classification Models
for Automatic Speaker
Verification System
in Industry 4.0
Sanil Joshi, Mohit Dua, and Shelza Dua
National Institute of Technology

CONTENTS
8.1 Introduction................................................................................................... 113
8.2 Back-End Classification Models.................................................................... 115
8.2.1 Machine Learning Models................................................................ 116
8.2.2 Deep Learning Models...................................................................... 118
8.3 Validation....................................................................................................... 121
8.3.1 Hold Out Validation........................................................................... 122
8.3.2 k-Fold Cross Validation..................................................................... 122
8.3.3 Stratified k-Fold Cross Validation..................................................... 123
8.3.4 Monte Carlo Cross Validation........................................................... 123
8.4 Evaluation Metrics......................................................................................... 123
8.4.1 Equal Error Rate (EER)..................................................................... 123
8.4.2 Accuracy............................................................................................ 124
8.4.3 Precision (Pre)................................................................................... 124
8.4.4 Recall (Re)......................................................................................... 124
8.4.5 Area Under Curve (AUC).................................................................. 124
8.4.6 F1 Score............................................................................................. 125
8.4.7 Mean Average Precision (AP)........................................................... 125
8.5 Analysis of Different Classification Models.................................................. 125
8.6 Conclusion..................................................................................................... 126
References............................................................................................................... 128

8.1 INTRODUCTION
Automatic speaker verification (ASV) technologies have advanced to the point that
many industries and businesses such as banks are attracted in using them in actual
security systems. However, the susceptibility of these systems to a variety of direct
DOI: 10.1201/9781003321149-8 113
114 Intelligent Analytics for Industry 4.0 Applications

and indirect access threats reduces the effectiveness of the ASV a­uthentication
­mechanism. ASV system provides the verification and the authorization to the
­original users of the system [1].
Like other security systems, these ASV systems are also prone to various spoof-
ing threats. With respect to ASV system, spoofing is the security attack in which
imposter gain access to the original user’s private information by creating the audio
sample similar to the voice of the original user. In spoofing, different types of spoof-
ing threats like replay attacks, voice conversion (VC), speech synthesis (SS), and
mimicry attacks are very popular. Speech synthesis [2] and voice conversion [3] are
the types of logical access attacks and are performed using advanced algorithms and
functions. These attacks are performed by skilled person with deep knowledge in
speech processing.
The replay attacks [4] are the simplest form of spoofing attacks. In order to per-
form these attacks, we just require the recording and replay devices. With the help of
recording device, the attacker records the voice of the original user, and with the help
of replay device, the attacker gets access to the user’s private system after performing
verification. These replay attacks can be performed by any person because no techni-
cal expertise is required to perform these attacks. Another important spoofing attack
is mimicry attacks [5], and these attacks can be performed by skilled voice mimicry
artists. Mimicry artist mimics the voice of the original user and then performs spoof-
ing attack using mimic audio. Figure 8.1 shows the important components of ASV
system.
The ASV systems comprise three main phases. First important phase is front
end feature extraction phase. In feature extraction phase, the important and useful
information is extracted from the input audio signal to produce the audio feature vec-
tor. Second important phase is back-end classification model. The extracted feature
vector is then fed to classification model to classify the audio samples into bonafide
(original) or the spoofed (cloned) audio sample. Another important phase of ASV
system is the dataset because for speaker verification tasks, various speech datasets
such as ASVspoof 2015 [6], ASVspoof 2017 [7], ASVspoof 2019 [8], and voice spoof-
ing detection corpus (VSDC) [9] datasets are available.

FIGURE 8.1  Components in automatic speaker verification (ASV) system.

Various Audio Classification Models for Automatic Speaker 115

Various feature extraction techniques such as mel frequency cepstral ­coefficients

(MFCCs) [10], Gammatone cepstral coefficients (GTCCs) [11], constant Q ceps-
tral coefficients (CQCCs) [12], and acoustic ternary patterns (ATP) [13] are g­ etting
­popularity for the extraction of the useful feature from the input audio signal.
Traditionally, MFCC feature vectors were popular for the feature extraction tasks;
however, nowadays, GTCC, ATP, and CQCC-based feature extraction techniques are
also getting popularity.
For classification of the input audio samples into spoofed or bonafide audios, ini-
tially, the focus of the researchers was toward the use of Gaussian Mixture Model
(GMM) [14] based classifier. However, after advancement in artificial intelligence,
various machine learning techniques such as random forest (RF) [13], Naïve Bayes
(NB) [13], and K-nearest neighbor (KNN) [13], and different deep learning algo-
rithms such as long short-term memory (LSTM) [15], convolutional neural network
(CNN) [16], and recurrent neural network (RNN) [17] are becoming popular for
speaker verification tasks. For any biometric system, the equal error rate (EER) [18]
value is considered as the optimal evaluation criterion to measure the performance of
the speaker verification system. EER value generally defines as the rate at which the
system is accepting the false values and rejecting the true values. The main focus of
the chapter is to analyze the different classification models for the speaker verifica-
tion tasks. The detailed description of various back-end classification models used in
verification of speakers in ASV systems is given in the next section.
The organization of the rest of the chapter is as follows: Section 8.2 discusses the
various classification models used to classify audio samples into cloned or original
audios in speaker verification task. In Section 8.3, detailed discussion related to the
need of validation and types of validation performed in machine and deep learning
is done. Section 8.4 explores the various assessment criteria available for evaluating
the efficiency of ASV system. Section 8.5 performs the analysis of different classifi-
cation models that are used by different researchers to improve the performance of
ASV system. Section 8.6 includes the conclusion and future scope related to different
classification models used in ASV system.

8.2  BACK-END CLASSIFICATION MODELS

When the audio feature vector is fed to classification model, it classifies the audio
samples into original human audio or cloned audio sample. First, the classification
model is created. After creation of classification model, the hyper parameters are
set. Hyper parameters are those parameters that control the learning process. After
this, training of the classifier using training feature vector is performed. While train-
ing, the validation is also performed. In next step it is checked that whether the
results show improvement over previous settings of hyper parameters. If no, then
hyper parameters are set again and all the previous steps will run again. If yes, then
the testing of model is performed using testing audio feature vector. At last, model
gets evaluated using different evaluation criteria and classifies the audio samples into
bonafide or spoofed audio samples. Figure 8.2 shows the steps during classification
of audio samples. Classification models are divided into two parts. One is machine
learning models and other is deep learning models. Different machine learning
116 Intelligent Analytics for Industry 4.0 Applications

FIGURE 8.2  Steps during the classification of audio samples.

and deep learning classification models such as RF, NB, KNN, LSTM, CNN, and
­bidirectional LSTM [19,20] are widely used for classification of the audio samples
into spoofed and original audio samples. The fundamentals related to different clas-
sification models are as follows.

8.2.1 Machine Learning Models

Machine learning models are widely used in classification of text, audios, etc. The
brief introduction of popular machine learning algorithms that are used for classifi-
cation of audio samples is as follows:
Various Audio Classification Models for Automatic Speaker 117

• Naïve Bayes (NB)

NB classification algorithm is based on Bayes theorem. It is one of the
most common machine learning algorithms that is used for the classifica-
tion task of audio samples in speech processing domain. Bayes rule cal-
culates the probability of an object by considering the prior knowledge. It
depends on the conditional probability. The notation P( M | N ) means that
probability of event M when event N has already been occurred [13,21]. The
­following equation shows the formula of Bayes rule:

P ( N | M )* P ( M )
P(M | N ) = (8.1)
P(N )

where P ( M | N ) represents posterior probability, P ( M | N ) represents like-

lihood probability, P(M) denotes prior probability, and P(N) shows ­marginal
probability.
• K-Nearest Neighbor (KNN)
KNN is a supervised learning algorithm for classification and widely
used for binary classification. KNN classification algorithm depends on the
similarities between the new data and the available data. When the audio
feature vector is fed to the KNN model then according to the similarities, the
algorithm divides the new data in feature vector into the different classes.
In  KNN, first, the number of neighbors (N) is selected. And Euclidean
­distance of all k neighbors is calculated. Then in all categories, the num-
ber of data samples is calculated. Whenever new data (feature vector) for
the new audio sample comes, on the basis of the Euclidean distance it gets
divided into the different categories [13,21]. Euclidean distance between
two points A and B having data points ( X1 , Y1 ) and ( X 2 , Y2 ), respectively, can
be calculated as follows:

Euclidean distance = ( X 2 − X1  )2 + (Y2 − Y1 )2 (8.2)

• Random Forest
RF is based on an ensemble learning algorithm. Ensemble learning is
the technique in which output from the different classifiers is taken into
consideration to classify the input audio sample into bonafide or spoofed
audio sample. In random forest, output from multiple decision trees is con-
sidered instead of single decision tree which results in higher accuracy and
solves the problem of overfitting which is more prevalent in decision trees
[13,22]. The main advantage of RF algorithm over other classifiers is that
it can easily handle the large datasets and provides good performance as
compared to decision trees. Figure 8.3 shows the random forest classifier
for classifying the audio samples for speaker verification task in speech
processing.
118 Intelligent Analytics for Industry 4.0 Applications

FIGURE 8.3  Block diagram of classification using random forest classifier.

8.2.2 Deep Learning Models

In artificial intelligence, deep learning-based classifiers are getting popularity to
classify the audio samples into bonafide (original human audio) or spoofed (cloned)
samples. Various popular deep learning-based models used for the classification of
audio samples are explained as follows:

• Convolutional Neural Network (CNN)

CNN is one of the most popular deep learning algorithms for classifying
images and audio samples. The images and audio samples are considered
in matrix form. On these, matrix the operation of mathematics called con-
volution is performed, which is why the name CNN name is given to this
technique. CNN mainly consists of three main layers, i.e., convolutional
layer, pooling layer, and fully connected layer [23]. In convolutional layer,
the convolution operation is performed between the matrix generated either
from the input image or audio sample and the filter having the size K*K,
i.e., square matrix. The next layer is the pooling layer. The main task of this
layer is to reduce the size of the output of the convolutional layer. In simple
Various Audio Classification Models for Automatic Speaker 119

words, it generally lowers the size of the feature vector. Bias and weights
associated with different neurons are considered in a fully connected layer.
First, input flattening is performed and then passed to the fully connected
layer for the classification task [23,24]. When the data are fed to the fully
connected layer, there might be overfitting. So, to prevent the effect of over-
fitting, drop-out layers are used. The drop-out layer randomly drops out
some neurons to avoid overfitting. Figure 8.4 shows the different layers of
the CNN. Different activation functions such as reLU, tanh, SoftMax, and
sigmoid are widely used in classification.
• Recurrent Neural Network (RNN)
RNN is the type of neural network in which the output of one layer is
fed as the input to another layer. RNN is very useful in the cases where
the prediction of the subsequent work of any given sentence needs to be
performed. Nowadays, researchers are also using RNN in speech process-
ing tasks. RNN consists of an input layer, hidden layer, and output layer. It
provides the same bias and weights to all the layers. Hence, it reduces the
increasing parameter complexity and delivers the output of the previously
hidden layer to the next layer as input to improve the memorization of the
output of each layer. Figure 8.5 shows the block diagram of the RNN.
• Long Short-Term Memory (LSTM)
LSTM model was developed to overcome the disadvantage of RNN.
RNN mainly suffers from vanishing gradient problems, i.e., when training
the neural network on backpropagation, there is no updating in the value
of the weights. The layer of LSTM consists of the connection of memory
block in the recurrent fashion. Hence it becomes suitable for sequence pre-
diction tasks and also for speech processing. In LSTM network, there are
three gates. First is input gate, and other two are output gate and forget
gate, respectively. The input gate’s job is to evaluate the new information

FIGURE 8.4  Layers in convolutional neural network.

120 Intelligent Analytics for Industry 4.0 Applications

FIGURE 8.5  Block diagram of recurrent neural network.

carried by the input and determine its significance. The forget gate is used
to ­determine whether the data from the preceding time stamp should be
kept or dismissed. The value of the next hidden state is determined by the
output gate.
The output of each gate is represented by the following equations [25,26]:

ft = ∅(Wt fg * ht −1 * it + b f (8.3)

itp = ∅(Wtig * ht −1 * it + bi (8.4)

Vi = tanh(Wc * ht −1 * it + bc   (8.5)

Vt = ft * Vt −1 + it * Vi (8.6)

where t is the time stamp, ft is the forget gate at t, it is the input, ht-1 is the
prior hidden state, Wtfg is the weights matrix, and bf is the bias in equa-
tion (8.3). In equations (8.4–8.6), Wtig represents the weight matrix between
Various Audio Classification Models for Automatic Speaker 121

the output gate and the input gate, Vi is the result generated after ­applying
the tanh  activation function, Wc is the weight matrix between cell state
­information and the previous output of the network, and Vt is cell state
information.

ot = ∅(WO * ht −1 * it + bo (8.7)

Opt = ot * tanh (Vt ) (8.8)

where ot is the output gate at t, Wo and bo denote the weight and bias of ot,
respectively, and Opt denotes the final output of LSTM.
• Gated Recurrent Unit (GRU)
The complexity of GRU model is less as compared to LSTM model. It
provides faster and better results as compared to LSTM and is widely used
when training samples are small in amount [27]. GRU consists of two gates
one is reset gate and other one is update gate. The role of update gate is
to determine how much past information is to be retained for making the
predictions for future. The reset gate helps to determine how much previous
information need to forget. The equation for output of update and reset gates
is as follows:

U ( t ) = ∂(Wu xt + Uu ht −1 ) (8.9)

R ( t ) = ∂(Wr xt + Ur ht −1 ) (8.10)

where U(t) and R(t) represent the output of update gate and reset gates,
respectively, Wu, Uu, Wr, and Ur show the weights, and xt is the input, and ht-1
is the output of previous t−1 states.

8.3 VALIDATION
In artificial intelligence, validation [28] often entails testing the system on the testing
dataset. This testing dataset is not a fresh dataset, but rather a subset of the training
dataset. Some of the training dataset is put aside for training, while others are placed
aside for testing [29]. In order to classify the audio samples into bonafide or spoofed
audios, the classification model is trained with different hyper parameters. With the
help of validation, the best hyperparameters are selected for the particular model
that gives high validation accuracy. Validation testing helps to select the best hyper
parameters for the model. The validation techniques also determine the underfit-
ting and overfitting of the classification model. Underfitting is the situation in which
the model is neither providing good accuracy on training data nor on testing data.
Overfitting occurs when the model generalizes good on the training data but does
not show good performance on the testing data. Different validation techniques that
are popular to improve the performance of the machine learning and deep learning
models are as follows:
122 Intelligent Analytics for Industry 4.0 Applications

8.3.1 Hold Out Validation

In holdout cross validation, the training speech feature vector is divided into two
parts. One is training speech feature vector part, which will be used for training of
the classification model, and another part is testing speech vector part, which will
be used for the testing of the classification model [30]. As training is performed on
larger dataset, the for-train test split, 80% of data will be reserved for training and
remaining 20% of data will be sued for testing. This validation process is the easiest
to understand but it has some drawbacks. The major drawback is that the testing data
may contain the useful information needed for training of the model and training
dataset missed out all the essential information required for training. It may lead to
low accuracy during training and validation.

8.3.2  k-Fold Cross Validation

In order to overcome the drawbacks of holdout validation, k-fold cross validation is
used [31]. In K-fold cross validation, subsets of the training data are formed. K gener-
ally defines number of the subsets to be formed. These subsets contain the random
data, i.e., subsets may have imbalance classes [29]. It means that one subset may have
higher number of bonafide samples as compared to spoofed samples and vice versa.
Figure 8.6 shows the block diagram of k-fold cross validation when the number of
folds is set to 5.
First, during iteration one, five folds of training data are formed. Four folds all
together are reserved for training of the classification model and first subset set is
used for testing. Similarly for the second iteration, the second subset will get used
for testing of the classification model and remaining four subsets all together will

FIGURE 8.6  k-fold cross validation when the number of folds is set to 5.
Various Audio Classification Models for Automatic Speaker 123

be used for training of the system. And this process will get repeated till the fifth
­iteration. It is important to note that only one subset will be reserved for testing at a
time, and the remaining subsets all together will be used for training.

8.3.3 Stratified k-Fold Cross Validation

In k-fold cross validation, each subset contains the random distribution of the data.
It means that one class may have more samples as compared to another class. It may
lead to low accuracy. In order to solve this problem, stratified k-fold cross validation
is used [32]. In these techniques, between each subset, all the classes are equally
distributed. It means that each subset contains a generalized form of the dataset. It
may lead to improve the validation accuracy. This is the only difference between
k-fold and stratified k-fold cross validation. Rest all the steps in both the techniques
are same.

8.3.4 Monte Carlo Cross Validation

In Monte Carlo cross validation [33], the training audio feature is divided into train
test split. This split can be 70:30, 80:20, and 75:25, i.e., different train test split for all
the iterations. Roughly, a large number of iterations such as 500 or 1,000 are used in
this process. At the end, average accuracy of all the iterations is taken to calculate
the final validation accuracy. This process is the enhancement of holdout validation,
as in this approach, different iterations are there and each iteration contains differ-
ent train test splits. The major disadvantage of this approach is that some of the data
points may occur many times in the testing set but some might never occur in the
testing set.

8.4  EVALUATION METRICS

Evaluation metrics are the performance evaluation criteria that are used to measure
the performance of machine learning and deep learning algorithms. While evaluat-
ing the classification model, there can be four possibilities: false positive (FPos), true
negative (TNeg), true positive (Tpos), and false negative (FNeg). TPos samples are
those that are correctly predicted, FPos samples are those that are mistakenly fore-
casted, TNeg samples are those that are correctly predicted, and FNeg samples are
those that are incorrectly predicted. The brief introduction to different evaluation cri-
teria that are used to measure the performance of the proposed system is as follows.

8.4.1 Equal Error Rate (EER)

EER value is one of the superlative evaluation criteria for the evaluation of the
ASV system [34]. It generally outlines the proportion at which system discards the
bonafide audio data, i.e., false rejection rate (FRR), and admits the spoofed audio
data, i.e., FRR. EER usually sets the threshold value of FRR and FAR. The point at
which FRR and FAR becomes equals that point is entitled as the ERR value. The
formula for FAR and FRR is as follows:
124 Intelligent Analytics for Industry 4.0 Applications

Number of False Acceptance possibilties 
FAR = (8.9)
Total number of the utterances

Number of False Rejection  possibilties 
FRR = (8.10)
Total number of the utterances

8.4.2 Accuracy
Accuracy is another evaluation criterion that is used to measure the performance of
classification model. Accuracy is the measure of number of the samples that model
correctly predicts out of the total number of the samples [34]:

Number of correctly predicted samples
Accuracy = (8.11)
Total number of the samples

TPos+TNeg
Accuracy = (8.12)
TPos+ FPos+TNeg+FNeg

8.4.3  Precision (Pre)

Precision is the fraction of number of the audio samples that are correctly predicted,
i.e., Tpos to the total number of the correctly predicted and incorrectly predicted
samples of positive class i.e., Tpos and FPos [34].

TPos
Precision = (8.13)
TPos + FPos

8.4.4 Recall (Re)
Recall basically defines out of all the predicted samples of one class how many of the
samples are correctly predicted by our classification model. Recall is the fraction of
number of TPos to TPos and FNeg [34].

TPos
Recall =   (8.14)
TPos + FNeg

8.4.5 Area Under Curve (AUC)

Area under ROC curve in machine learning is the performance evaluation criteria
that is used to measure the performance of the classification algorithm. It defines
the rate at which our classification algorithm is able to find differences between the
binary classes of the samples. More the AUC value better our system classification
would be and vice versa. The ROC curve is plotted between the Tpos rate (recall) and
FPos rate [35]. Formula for FPos rate is as follows:
Various Audio Classification Models for Automatic Speaker 125

FPos
False Positive Rate =   (8.15)
TNeg + FPos

8.4.6 F1 Score
F1 score is the evaluation criterion widely used in the binary classification problem.
It is the harmonic mean between the precision and recall values. In order words, we
can say it is the weighted average between recall value and precision value. It has
actual application when our dataset contains the irregular distribution of the classes
[34]. The formula for the calculation of F1 score is as follows:

2* Pre * Re
F1  Score = (8.16)
Pre + Re

8.4.7 Mean Average Precision (AP)

AP is another popular evaluation metric that is used for measuring efficiency of
the classification algorithm. It is defined as taking the mean of precisions over all
the data sample groups reliant on diverse trials that happen [36]. The formula for the
calculation of AP is as follows:

∑
n
Precision of given class
AP = i =1
(8.17)
Total number of samples of particular class

8.5  ANALYSIS OF DIFFERENT CLASSIFICATION MODELS

The work of Sadhu et al. [34] mainly focused on the hybrid of LSTM and CNN model
at back end to classify the audio samples into bonafide or spoofed audios. First, the
authors have extracted Mel spectrograms from the audio samples and these spectro-
grams are fed to CNN classification model. From the same audio samples, CQCC
features are extracted and fed to LSTM based classification model. The output of
both the models gets combined to provide the final output as bonafide or spoofed
audio. The proposed system by them provides 3.6% EER in 1PR attacks and 2.96%
EER in 2PR attacks, respectively.
The work of Mittal and Dua involves different deep learning models and combi-
nation of different deep learning models to classify the audio samples into bonafide
or spoofed audios. First, the authors [37] have proposed CQCC–2DCNN (two-
dimensional convolutional neural network)-based ASV system. For extraction of
CQCC features the authors have used 30 static + 30 first order + 30 second order fea-
tures. Their first proposed system provides 5.5% EER, 0.101 min t-DCF value against
logical access (LA) attacks, and 6.2% EER and 0.122 min t-DCF value against PA
attacks. Second, the authors [37] have proposed CQCC–LSTM-based system. The
proposed system provides 0.42% EER against LA attacks and 0.51% EER value
against PA attacks. In the third proposed system [12], the authors have combined dif-
ferent deep learning models such as LSTM with time-distributed wrappers and CNN.
126 Intelligent Analytics for Industry 4.0 Applications

The authors conclude that hybrid deep learning models improve the c­ lassification
performance of ASV system.
Malik et al. [13] proposed ATP+GTCC-based feature extraction technique.
The extracted feature vectors are fed to different classification models such as
­Error-Correcting Output Coding-Support Vector Machine (ECOC-SVM), bagged
trees, decision trees, Naïve Bayes, KNN, and bidirectional LSTM. The authors con-
clude that hybrid feature extraction at front end and machine learning model at back
end improves the efficiency of ASV system. The proposed system by them provides
0.6% EER against 1PR and 2PR attacks and 1% EER value under PA attacks.
Li et al. [38] proposed attention-based LSTM to classify the audio samples into
bonafide or cloned audio samples. The authors conclude that the attention-based
LSTM provides nearly 13.81% improvement in EER value as compared to LSTM.
Faisal et al. [39] used MFCC feature extraction to extract the important features
from the audio samples. After that, the extracted feature vectors are given to DNN
and GMM classifiers individually. The authors conclude that the DNN-based classi-
fier provides better results (18.1% EER) as compared to GMM-based classifier (19%
EER).
Dua et al. [40] have proposed a hybrid of deep neural networks such as LSTM
with time-distributed dense layers, and spatial convolution (SC) and temporal con-
volution (TC)-based DNN to perform the classification of the audio samples into
bonafide or spoofed audios. The hybrid deep neural network model provides 0.6%
EER against both LA and physical access (PA) (ASVspoof 2019 dataset) attacks.
Kumar and Bharathi [41] proposed FBCC feature extraction technique to extract
the useful information from the audio samples. The extracted audio feature vector
is fed to GMM-based classifier. The proposed system provides 0.16 min t-DCF value
against LA attacks and 0.25 min t-DCF value against PA attacks. Table 8.1 provides
the performance analysis of ASV systems that use different feature extraction and
classifiers to classify the audio samples into original human audio or spoofed audio.
Further, from the performance analysis of different classifiers, it can be concluded
that the classification model that performs the validation of the obtained results using
different validation techniques, that classifier provides better performance as com-
pared to others.

8.6 CONCLUSION
For industry automation, voice technologies are very helpful to automate the produc-
tion process. Speaker verification performs the classification of voice samples into
bonafide or spoofed audios. The back-end classification model plays a significant role
to improve the accuracy of ASV system. Traditionally, machine learning-based mod-
els were popular for spoofing detection task in speaker verification system. However,
today deep learning models and hybrid of various deep learning models are get-
ting researchers’ attention. Although the hybrid deep learning models at back end
improve the performance of ASV system, it also increases the overall complexity of
the system. Further to increase the performance of classification model to classify the
audio samples into bonafide or spoofed, the validation procedure can be applied. For
evaluating the ASV system, along with EER value different performance evaluation
TABLE 8.1
Analysis of Different Classification Models Used at Back End of ASV System
Work Front End Back End Validation Results Remarks
Sadhu et al. [34] Mel LSTM, CNN K-fold cross 3.6% EER (1PR) Hybrid deep learning model of LSTM+CNN at
spectrograms, validation 2.96% EER (2PR) back end improves the performance of overall
CQCC 97.6% Accuracy (1PR) system as compared to system that uses
97.78% Accuracy (2PR) standalone CNN and LSTM models.
Mittal and Dua Static dynamic LSTM with time-distributed Validation using 0.9% EER Hybrid deep learning models improve the
[12] CQCC features wrappers, CNN, Voting development set performance of ASV system.
protocol
Dua et al. [40] MFCC, IMFCC, LSTM with time-distributed — 0.6% EER using ensemble CQCC features with ensemble deep neural
CQCC dense layers, SC and model network provide better EER as compared to
TC-based DNN, Ensemble other models and existing systems.
model of all these models
Mittal and Dua 30 static + 30 first 2D CNN — 5.5% EER and 0.101 t-DCF Combination of 2D CNN with CQCC features
[37] order + 30 for LA attacks increases the efficiency of ASV system.
second order 6.2% EER and 0.122 t-DCF
CQCC features for PA attacks
Malik et al. [13] Hybrid ECOC-SVM, Decision K-fold Cross 0.6% EER (VSDC) When hybrid feature extraction technique is used
ATP+GTCC Trees, KNN, NB, Bagged Validation 1% EER (PA) at front end, machine learning algorithms show
features Trees, Bidirectional LSTM great performance as compared to deep learning
techniques.
Various Audio Classification Models for Automatic Speaker

Li et al. [38] CQCC features Attention-based Long — 20.32% EER (Traditional Attention-based LSTM provides better results as
Short-Term Memory LSTM) compared to other models.
16.86% EER (Attention-
based LSTM)
Mittal and Dua CQCC LSTM classification model — 0.42% EER (LA attacks) LSTM model outshines other baseline models
[15] 0.51% EER (PA attacks) when CQCC features are fed to it.
Faisal et al. [39] MFCC DNN, GMM — 18.1% EER (DNN) Deep neural networks provide good results as
19.0% EER (GMM) compared to GMM.
127

Kumar and FBCC GMM classifier 0.16 t-DCF (LA attacks) GMM-based classifier improves the min t-DCF
Bharathi [41] 0.25 t-DCF (PA attacks) value when FBCC features are fed to it.
128 Intelligent Analytics for Industry 4.0 Applications

criteria such as F1 score, precision, recall, and accuracy can be used. Further in
future scope to increase the performance and to decrease the overall complexity of
ASV system, the ASV system with hybrid feature extraction at front end and less
complex classification model such as light convolutional neural network (LCNN)
can be used.

REFERENCES
1. Z. Wu, N. Evans, T. Kinnunen, J. Yamagishi, F. Alegre, and H. Li, “Spoofing and coun-
termeasures for speaker verification: A survey,” Speech Commun., vol. 66, pp. 130–153,
2015.
2. A. Indumathi and E. Chandra, “Survey on speech synthesis,” Signal Process. An Int. J.,
vol. 6, no. 5, p. 140, 2012.
3. A. Kain and M. W. Macon, “Design and evaluation of a voice conversion algo-
rithm based on spectral envelope mapping and residual prediction,” in 2001 IEEE
International Conference on Acoustics, Speech, and Signal Processing. Proceedings
(Cat. No. 01CH37221), 2001, vol. 2, pp. 813–816.
4. M. Liu, L. Wang, Z. Oo, J. Dang, D. Li, and S. Nakagawa, “Replay attacks detection
using phase and magnitude features with various frequency resolutions,” in 2018 11th
International Symposium on Chinese Spoken Language Processing (ISCSLP), 2018,
pp. 329–333.
5. V. Vestman, T. Kinnunen, R. G. Hautamäki, and M. Sahidullah, “Voice mimicry attacks
assisted by automatic speaker verification,” Comput. Speech Lang., vol. 59, pp. 36–54,
2020.
6. Z. Wu et al., “ASVspoof 2015: the first automatic speaker verification spoofing and
countermeasures challenge,” 2015.
7. H. Delgado et al., “ASVspoof 2017 Version 2.0: meta-data analysis and baseline
enhancements,” 2018.
8. Y. Yang et al., “The SJTU robust anti-spoofing system for the ASVspoof 2019 chal-
lenge,” in Interspeech, 2019, pp. 1038–1042.
9. R. Baumann, K. M. Malik, A. Javed, A. Ball, B. Kujawa, and H. Malik, “Voice spoof-
ing detection corpus for single and multi-order audio replays,” Comput. Speech Lang.,
vol. 65, p. 101132, 2021.
10. S. Nakagawa, L. Wang, and S. Ohtsuka, “Speaker identification and verification by com-
bining MFCC and phase information,” IEEE Trans. Audio. Speech. Lang. Processing,
vol. 20, no. 4, pp. 1085–1095, 2011.
11. M. Dua, R. K. Aggarwal, and M. Biswas, “GFCC based discriminatively trained
noise robust continuous ASR system for Hindi language,” J. Ambient Intell. Humaniz.
Comput., vol. 10, no. 6, pp. 2301–2314, 2019.
12. A. Mittal and M. Dua, “Static–dynamic features and hybrid deep learning models
based spoof detection system for ASV,” Complex Intell. Syst., vol. 8, pp. 1–14, 2021.
13. K. M. Malik, A. Javed, H. Malik, and A. Irtaza, “A light-weight replay detection frame-
work for voice controlled IoT devices,” IEEE J. Sel. Top. Signal Process., vol. 14, no. 5,
pp. 982–996, 2020.
14. A. Kuamr, M. Dua, and T. Choudhary, “Continuous Hindi speech recognition
using Gaussian mixture HMM,” in 2014 IEEE Students’ Conference on Electrical,
Electronics and Computer Science, 2014, pp. 1–5.
15. A. Mittal and M. Dua, “Constant Q cepstral coefficients and long short-term memory
model-based automatic speaker verification system,” in Proceedings of International
Conference on Intelligent Computing, Information and Control Systems, 2021,
pp. 895–904.
Various Audio Classification Models for Automatic Speaker 129

16. B. Chettri, S. Mishra, B. L. Sturm, and E. Benetos, “Analysing the predictions of

a ­cnn-based replay spoofing detection system,” in 2018 IEEE Spoken Language
Technology Workshop (SLT), 2018, pp. 92–97.
17. T. Tan et al., “Speaker-aware training of LSTM-RNNs for acoustic modelling,” in 2016
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP),
2016, pp. 5280–5284.
18. J.-M. Cheng and H.-C. Wang, “A method of estimating the equal error rate for automatic
speaker verification,” in 2004 International Symposium on Chinese Spoken Language
Processing, 2004, pp. 285–288.
19. S. Duraibi, W. Alhamdani, and F. T. Sheldon, “Replay spoof attack detection using deep
neural networks for classification,” in 2020 International Conference on Computational
Science and Computational Intelligence (CSCI), 2020, pp. 170–174.
20. C. Hanilci, T. Kinnunen, M. Sahidullah, and A. Sizov, “Spoofing detection goes noisy:
An analysis of synthetic speech detection in the presence of additive noise,” Speech
Commun., vol. 85, pp. 83–97, 2016.
21. T. G. Dietterich, “Ensemble methods in machine learning,” in International Workshop
on Multiple Classifier Systems, 2000, pp. 1–15.
22. G. Biau and E. Scornet, “A random forest guided tour,” Test, vol. 25, no. 2, pp. 197–227,
2016.
23. S. Albawi, T. A. Mohammed, and S. Al-Zawi, “Understanding of a convolutional neural
network,” in 2017 International Conference on Engineering and Technology (ICET),
2017, pp. 1–6.
24. A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep
convolutional neural networks,” Adv. Neural Inf. Process. Syst., vol. 25, 2012,
­
pp. 1097–1105.
25. K. Greff, R. K. Srivastava, J. Koutník, B. R. Steunebrink, and J. Schmidhuber, “LSTM:
A search space odyssey,” IEEE Trans. Neural Networks Learn. Syst., vol. 28, no. 10,
pp. 2222–2232, 2016.
26. A. Graves and J. Schmidhuber, “Framewise phoneme classification with bidirectional
LSTM and other neural network architectures,” Neural Networks, vol. 18, no. 5–6,
pp. 602–610, 2005.
27. R. Dey and F. M. Salem, “Gate-variants of gated recurrent unit (GRU) neural net-
works,” in 2017 IEEE 60th International Midwest Symposium on Circuits and Systems
(MWSCAS), 2017, pp. 1597–1600.
28. P. Refaeilzadeh, L. Tang, and H. Liu, “Cross-validation,” Encycl. Database Syst., vol. 5,
pp. 532–538, 2009.
29. T. Fushiki, “Estimation of prediction error by using K-fold cross-validation,” Stat.
Comput., vol. 21, no. 2, pp. 137–146, 2011.
30. S. Yadav and S. Shukla, “Analysis of k-fold cross-validation over hold-out validation on
colossal datasets for quality classification,” in 2016 IEEE 6th International Conference
on Advanced Computing (IACC), 2016, pp. 78–83.
31. J. Wieczorek, C. Guerin, and T. McMahon, “K‐fold cross‐validation for complex sample
surveys,” Stat, vol. 11, p. e454, 2022.
32. S. Alrabaee, “A stratified approach to function fingerprinting in program binaries using
diverse features,” Expert Syst. Appl., vol. 193, p. 116384, 2022.
33. P. Smyth, “Clustering using Monte Carlo cross-validation,” Kdd, vol. 1, pp. 26–133,
1996.
34. M. Dua, A. Sadhu, A. Jindal, and R. Mehta, “A hybrid noise robust model for mul-
tireplay attack detection in automatic speaker verification systems,” Biomed. Signal
Process. Control, vol. 74, p. 103517, 2022.
35. A. P. Bradley, “The use of the area under the ROC curve in the evaluation of machine
learning algorithms,” Pattern Recognit., vol. 30, no. 7, pp. 1145–1159, 1997.
130 Intelligent Analytics for Industry 4.0 Applications

36. S. Ding, Q. Wang, S. Chang, L. Wan, and I. L. Moreno, “Personal VAD: Speaker-
conditioned voice activity detection,” arXiv Prepr. arXiv1908.04284, 2019.
37. A. Mittal and M. Dua, “Automatic speaker verification system using three dimensional
static and contextual variation-based features with two dimensional convolutional
­neural network,” Int. J. Swarm Intell., vol. 6, no. 2, pp. 143–153, 2021.
38. J. Li, X. Zhang, M. Sun, X. Zou, and C. Zheng, “Attention-based LSTM algorithm for
audio replay detection in noisy environments,” Appl. Sci., vol. 9, no. 8, p. 1539, 2019.
39. M. Y. Faisal and S. Suyanto, “SpecAugment impact on automatic speaker verification
system,” in 2019 International Seminar on Research of Information Technology and
Intelligent Systems (ISRITI), 2019, pp. 305–308.
40. M. Dua, C. Jain, and S. Kumar, “LSTM and CNN based ensemble approach for spoof
detection task in automatic speaker verification systems,” J. Ambient Intell. Humaniz.
Comput., vol. 13, pp. 1–16, 2021.
41. S. Rupesh Kumar and B. Bharathi, “A novel approach towards generalization of
­countermeasure for spoofing attack on ASV systems,” Circuits, Syst. Signal Process.,
vol. 40, no. 2, pp. 872–889, 2021.
 9 Trending IoT Platforms
on Middleware Layer
Neha Katiyar and Priti Kumari
Noida Institute of Engineering & Technology

Surabhi Sakhshi
Indian Institute of Management

Jyoti Srivastava
Madan Mohan Malviya University of Technology

CONTENTS
9.1 Introduction................................................................................................... 131
9.2 Related Work................................................................................................. 132
9.3 IoT Platforms................................................................................................. 135
9.3.1 Functional Requirements................................................................... 137
9.3.2 Nonfunctional Requirements of IoT Platforms................................. 137
9.4 Categories of IoT Management Platform for Industries................................ 138
9.4.1 Device Management Platforms.......................................................... 138
9.4.2 IoT Application Management Platforms............................................ 139
9.4.3 Application Development Platforms.................................................. 139
9.5 Trending IoT Platform for Industry Revolution 4.0....................................... 139
9.6 Connectivity of IoT Platforms Middleware Layer Using 6G Technology..... 144
9.7 Conclusion..................................................................................................... 146
References............................................................................................................... 146

9.1 INTRODUCTION
Internet of Things (IoT) is the word that was first introduced to this world by Kevin
Mshton in 1999. IoT means ‘Internet’ is the network of networks, and connectivity
is needed to connect the things autonomously. It is expected that by 2020 around 60
million devices may be connected to the Internet. The IoT architecture consists of
different layers in it. The sensors and actuators layers, transit and communication
layer, middleware layer, and application layer. Middleware act as a bridge between
the network and applications. It is linked between the different software and is hid-
den between the OS and application layers. The application layer includes different
types of applications that can run on various IoT platforms. The basic principle IoT
follows is as we give commands by speaking, the device performs the same task as

DOI: 10.1201/9781003321149-9 131

132 Intelligent Analytics for Industry 4.0 Applications

we give to them. It can be achieved by various technologies such as Bluetooth, WiFi

(wireless fidelity), Zigbee, and Z-Wave. They constantly maintained undisputed
proof of the progress of IoT platforms achieving that technical revolution.
IoT platforms enable interoperability for heterogeneous devices and technolo-
gies. There are so many platforms for IoT that can be efficiently designed for
the literature and industry, mainly healthcare fields for monitoring and energy
conservation and psychological fields for child monitoring. In middleware layer
solutions, programmers must know about the new software specifications because
they are implemented on the latest software packages. These tasks consumed a lot
of time. In IoT, different corporate persons and industrialists will use various soft-
ware that might be incompatible. Billion devices are connected with IoT (sensors,
printers, mobiles, laptops, cameras, and projectors), and their heterogeneous IoT
platforms can perform essential tasks. Prominent technology companies created
their own IoT platform, the IoT activity extensive framework. The top companies
have strong competencies and their name mentioned are Amazon web services,
Ayla Agile IoT platform, Azure IoT suite, Bosch IoT suite, Jasper control center,
Oracle IoT cloud services, GE Predix, SAP cloud platform, PTC Thing work, and
Watson IoT platform.
As the IoT, as mentioned above, platforms nowadays cover various sectors. IoT
platforms have gained more importance in the last decades. The criteria of these
platforms’ functionality, strategic techniques, easy availability, and cloud connectiv-
ity will increase the platform’s position in the market. The word IoT consists of the
word Internet. Due to this reason, the entire thing that works on the Internet currently
uses significantly lighter IoT versions. Big data analysis is the most frequent wire-
less method for connecting the Internet. It performs such that the Internet protocol
stack does not apply any restrictions on IoT in the account. Bluetooth 5, Zigbee, and
WiFi are these technologies used for better connectivity with the Internet in IoT.
Bluetooth 5.0 is the trending variant used for communication between various intel-
ligent homes, IoT platforms, and IoT devices. (Bluetooth range is limited; it supports
only 5 m in outdoor environments, and 200 m and 40 m indoors).
Perceiving the decisive role of platforms in IoT platforms, we focused on that
paper as IoT middleware platforms or an instance model of IoT architecture. This
chapter’s proposal of IoT security features in IoT middleware is mainly designed for
a protected environment for IoT. This chapter also highlights the significance of IoT
middleware and platforms and their essential role with 6G network connecting tech-
nology. The remaining piece consists of the following criteria: Section 9.2 provides
an analysis of security in IoT middleware, IoT networks, and related work. Section
9.3 IoT platforms. Section 9.4 presents the categories of IoT platforms for industry.
Section 9.5 presents the trending IoT platform for Industry Revolution 4.0. Section
9.6 proposes a technique for IoT platforms connectivity with the 6G network. Finally,
Section 9.7 is the conclusion of the chapter.

9.2  RELATED WORK

In general, based on various literature research smart network connectivity, we
observed that IoT platforms are categorized into two sections:
Trending IoT Platforms on Middleware Layer 133

1. Cloud-based systems with IoT devices.

2. Device gateways are directly connected to the cloud through networks.

In 2020, Viswanathan et al. [13] focused on the 6G network is gradually taking

momentum. This chapter discusses different interface connectivity with 6G networks
such as man–machine interfaces, ubiquitous interfaces, multilevel sensor data, or IoT
platform connectivity. The article also shows the foundation of the 6G interface air
network and its connectivity with different devices. It is the most beneficial net-
working technology for connecting IoT devices with the middleware layer platform.
Nevertheless, a 6G network is undoubtedly taking the shape of IoT platforms.
In 2021, Barakat et al. [14] also highlighted the 6G connecting technology with
IoT platforms. 6G will continue to provide its multiple feature network services for
all digital devices. Smart cities will have the future of flying taxis already operated in
limited areas in various cities such as Dubai. The regulation, as well as connectivity
requirements of flying taxis, required a fully connected IoT platform.
In 2021, Nkenyereye and Hwang [17] mainly highlighted the specialization of
IoT systems. They focused on the visual network functions of the IoT platform on
different layers. They give an idea about IoT platform communication infrastructure
method with network slicing involving 5G and 6G techniques. They designed an
algorithm for network slicing of IoT platforms. The algorithm name is virtual IoT
slice services functions (VIoT-SSFs). This algorithm improves the network function
of IoT platforms. The IoT ecosystem consists of a stratified architecture model and
uses this representation to provide more imaginative solutions.
In 2018, da Cruz et al. [1] discussed the recommended dummy for the IoT middle-
ware. It shows designate the correct platform for a specific task and the differentiate
between the better and worse IoT solutions because it is a long-term process to make
the platform user-friendly. They discussed the features of the IoT platform and made
them more compatible without any security point drawbacks. IoT middleware plat-
forms are facing many challenges in providing security.
In 2015, Fersi [2] discussed and classified the challenge facing middleware IoT
platforms. They organized the middleware platform approach depending on the
generic techniques. They highlighted and explored the significant issues of security,
privacy, configuration, and bootstrapping maintenance issues.
Several algorithms are designed to maintain IoT platforms’ security and network
connectivity. In 2016, Pallavi et al. [3] discussed the security algorithm that protects
the IoT data on the middleware layer. The security in middleware is done by apply-
ing different algorithms in the fog layer to secure the information it will send to the
cloud. In 2021, Nawaz et al. [15] focused on various techniques of IoT platforms.
A secure IoT system is made when we apply several transfer protocols and use the
most secure connecting technology. 6G is the fastest and connecting technology.
In 2018, Kokkonis et al. [4] proposed that several IoT middleware protocols have
been used to transfer the data to the cloud. The prominent representatives of IoT
are the COAP (constrained application protocol), SOAP (simple object access pro-
tocol), MQTT (MQ telemetry transport), and HTTP (hypertext transfer protocol).
It concerns the various protocol techniques such as delay, packet loss, and process-
ing of IoT devices. In addition, the categorization of an advanced security layer in
134 Intelligent Analytics for Industry 4.0 Applications

middleware platforms that may generate additional features overhead, depending on

the restraint of the environmental resources, and these costs cannot be confirmed
without security measures. A secure network is formed on these platforms.
In 2019, Silva et al. [5] implemented management for devices and networks in
IoT (M4DN.IoT) – a network and device management platform for the IoT. M4DN.
IoT network provides a solution for automatic IoT network management and a user
compatibility interface that includes information about the network devices. This
­platform may be used in any smart device (desktop, laptop, and smart meter) and
­provide access to any location. In 2021, Nkenyereye et al. [17] focused on the work-
ing of the IoT platform. The IoT platforms are providing good quality services.
Interactions have been bounded due to heterogeneous creation of devices, commu-
nication protocols, data types, data rates, quality of service (QoS) requirements, and
trust measures between applications. It is also used for maintaining the security of
the home environment. Sandor et al. [6] proposed the operational requirements of
protection for an IoT platform called security-enhanced interoperability middleware
for the IoT.
The proposed IoT platform is named ghost – safeguarding home IoT environments
with personalized real-time risk control. IoT devices provide that much comfort to
the industries so industries can adapt quickly, and it designed the Industrial IoT (IIoT)
procedure. In this IIoT procedure, in 2018, Bakhshi et al. [7] discussed the IIoT and
investigated the security and concern of IoT. For the data accumulation and abstrac-
tion levels of IIoT, they chose Cisco and Microsoft Azure IoT because the two layers
of the Cisco reference architecture model have been selected to determine the vari-
ous security measures. IIoT is used in industries to develop the best quality products.
The establishment of IIoT has provided a new platform designed in the middleware
layer called IOP (Internet of production). In 2018, Jorg Hoffman et al. [8] proposed
the method of Internet of production for the IoT platforms. The IoT platforms in the
merchandise are not fully customized to meet the particular requirements of users
and organizations – specifically in the manufacturing industry; to manage these
issues, the Internet of production is needed. Additionally, IoT creates a paradigm
for a required set of features that represents an architecture of reference models.
In designing the IoT middleware platforms and models, many challenges nowadays
occur. Some researchers also focused on the challenges of IoT platforms. In 2018,
Yadav et al. [9] discussed IoT and its challenges and issues in Indian perspective. It
familiarizes to show the growth of IoT in India, its risk factors, and security issues
challenge an Indian view. The most critical challenge is maintaining the security
of IoT platforms. In 2019, Jin-Young Yu et al. [10] analyzed IoT platform security.
Different settled IoT platforms are customized based on media on which similar-
ity is underway, such as one Machine-to-Machine (M2M) and Open Connectivity
Foundation (OCF). Different settled IoT platforms are customized based on media
on which similarity is underway, such as one M2M and OCF for the analysis and
comparison of security elements of national and international IoT platforms.
In 2017, Mario Frustaci et al. [11] proposed a new paradigm in SIOT (social
Internet of things), which merged with socialization and allowed humans to connect
with devices and share information. They study the following three significant lay-
ers: (i) perception layer, (ii) transportation layer, and (iii) application layer. The most
Trending IoT Platforms on Middleware Layer 135

unsafe level of the IoT system is the perception layer due to the physical disclosure
of the IoT. The IoT middleware layer platforms are also used in the healthcare sys-
tems for sending and detecting data of patients to forensics. Macro Conti et al. [12]
introduced the significant security and forensics issues in IoT with their potentially
promising solutions. They introduced special security, privacy, and protection chal-
lenges in IoT environments. In 2021, Jang et al. [16] focused on nonintrusive load
monitoring techniques. These techniques are beneficial and efficient for intelligent
energy systems. In this technique, they designed IoT system architecture that consists
of an application layer, perception layer, appliances layer, communication layer, and
middleware layer. They compare this generation technique with ILM (intrusive load
monitoring technique) because the energy consumption is more than the NILM (non-
intrusive load monitoring) technique. Moreover, the IoT middleware layer provides a
sustainable, innovative system that consumes low power. The computational require-
ment of the middleware layer is very high. In 2020, Swamy et al. [18] focused on
cloud, fog, and edge computing connectivity with IoT platforms. In this paper, they
summarized the IoT platform and IoT application architecture. They also focused on
real-time-based IoT systems. It uses various computing techniques. The middleware
layer is the abstract layer between the user interface and deployed devices.

9.3 IoT PLATFORMS
At a high level, an IoT platform supports software that interconnects each hardware,
access point, and data network to additional quality of the value chain. An IoT plat-
form is often referred to as a middleware solution, which is the plumbing of the IoT.
These platforms are a package that integrates devices and applications, as shown in
Figure 9.1. Middleware is the link between different software and applications. The
middleware system integrates the distributed application system with heterogeneous
network environments. The middleware layer connectivity with hardware and appli-
cations is shown in Figure 9.2. The layer is hidden between the OS and application
layers. It is connected to all types of devices. The kinds of middleware layer plat-
forms are formed, and their specifications given are as follows:

Message-Oriented Middleware: this type of middleware software

1.
allows additional software components to send and receive messages.
This ­software also provides connectivity to the front and back end of the
procedure.
Object Middleware: it is a dissent masterminded system that sends inqui-
2.
ries about the request of organizations.
Database Middleware: this middleware layer can be used to communicate
3.
with database storage. The data middleware layer can combine if the user
wants to extract and retrieve various information from the database.
Exchange Middleware: the trades exchanging with applications and
4.
screens are used in the durability of middleware. It covers the web-applica-
tion servers.
Implanted Middleware: this type of middleware layer is considered resem-
5.
blance and coordination associated with programming and interfaces.
136 Intelligent Analytics for Industry 4.0 Applications

FIGURE 9.1  IoT platform and its applications.

FIGURE 9.2  IoT middleware.

The platform hides the complex infrastructure and supports the enabled IoT ­solutions.
IoT platform increases the market value rate of the IoT sector. These platforms are
also called players of the market. The IoT platform market continues to provide
strong marketing businesses for IoT data-driven companies. The IoT platform mar-
ket period from 2015 to 2021. The Club and Group Activity Report (CGAR)’s annual
revenue was reaching $1.6 billion in 2021. IoT platform has some functional as well
as non-functional requirements in various industries. Functional requirements in the
industry such as healthcare and cloud storage.
Trending IoT Platforms on Middleware Layer 137

9.3.1 Functional Requirements
Functional requirement is the declaration of the intended operations of a system
and its parts. A plan should be designed in the form of functional requirements.
Functional requirements can be an instruction set for a device that gives you an idea
about assembling a device, which are explained as follows:

• Diverse Connections: Ethernet is needed to connect the Internet in the IoT.

IoT devices can use various technologies for connectivity over wide applica-
tions. The connectivity of IoT supports, various ways of connectivity wired
and wireless. Wireless options are Bluetooth, RFID tags, GPRS devices,
WLAN, Zigbee, and Z-Wave.
• Leverage Applications: when you enter the home, the door is automati-
cally open, lights are on automatically, and AC depends upon automation
capabilities that make IoT solutions valuable. These applications provide
efficiency with application to the IoT.
• Industry Range Management: the number of devices connected in this
world of IoT is 28–50 billion. According to the area of devices, the range is
for monitoring. IoT sensors can collect the data from where the sensors were
placed, such as temperature and moisture level. Actuators perform specific
tasks, and it makes things turn on and off. In addition, IoT has different
kinds of wearable devices, such as a health-tracking bracelet that can moni-
tor your health positioning at your location. The IoT platform can manage a
range of heterogeneous sets of machines.
• Massive Data: as we know, about billions of devices are connected with
IoT. They will also report the tasks that they performed. Their connection
with the IoT platform and with different devices. They will have to trans-
mit each detail of action to one estimation of forces in 2018. IoT generates
around 400 terabytes of data. IoT platforms may be able to store and support
massive amounts of data.
• Powerful Analytics: the massive amount of data stored by IoT platforms.
We require powerful analytics tools to recognize which data is helpful
for consumers. A compatible platform analytics solution will translate
­significant amounts of data into valuable and actionable insights.

9.3.2 Nonfunctional Requirements of IoT Platforms

Nonfunctional requirements substantially show how the system should work and that
it depends on the system’s behavior. The nonfunctional requirement consists of all
those left requirements detained by functional requirements, which is explained as
follows:

• Security: the most compelling concern in each platform and application

is security. In IoT, it is essential to avoid and perform against all sorts of
attacks, such as brute-force and Dos, or not to disclose sensitive informa-
tion, such as user locations, photos, documents, and regular schedules. The
138 Intelligent Analytics for Industry 4.0 Applications

significance of data being exposed without any limit platform should do

the best protection of user data while providing the mechanism of intru-
sion detection. Some IoT security solutions are Blackberry, Cisco, Subex,
CUJO F-Secure, ZingBox, Luma, etc. These platforms offer better security
services to the users.
• Scalability: connectivity of billions of devices and Zettabytes of data is
discussed, so scalability is the explicit requirement of the IoT platform. IoT
platforms should become scalable to grow the network exponentially. Many
IoT solutions can achieve their potential only at a scale.
• Availability: the reason is that the Internet is needed for connecting IoT
devices. The availability of the Internet makes devices work efficiently.
Due to the IoT devices being connected to the real-world impact, the IoT
­platform must provide the highest availability.
• Maintainability: the main reason for applying IoT to manage your predic-
tive maintenance. Maintainability means maintaining the performance of
IoT devices, compatibility, and maintenance of overall platforms. In this
maintenance, more and more systems shipping connectivity is needed.
• Serviceability: IoT devices and platforms can work together, and their
speed should be maintained. Serviceability means diagnosing problems
that should help a rise in the IoT platform while working or providing better
services to the users, so users are tangled with media and techniques.
• Usability: this means how an IoT platform should be used. The IoT plat-
form must be utilized. It should be used by users. It may not be complex for
such technical and non-technical users to operate efficiently.
• Data Integrity: we mentioned that terabytes of data should be transferred
over to the cloud and data used by the devices. The data are movable to
protect information from modification while on its journey from the sensor
to the cloud application. Data integrity is one of the most important key
aspects of implementing security policies.

9.4 CATEGORIES OF IoT MANAGEMENT

PLATFORM FOR INDUSTRIES
The best IoT platforms support functional as well as nonfunctional requirements.
The three main categories of IoT industry management platforms are as follows:

1. Device management platforms.

2. Application management platforms.
3. Application development platforms.

9.4.1 Device Management Platforms

One of the topmost concerns of the management of IoT device management is an
end-to-end device security provider. IoT devises management includes IoT security.
The management of IoT platforms encompasses device positioning administration
Trending IoT Platforms on Middleware Layer 139

and monitoring and diagnostics in various industries. IoT creates management func-
tions consisting of enrollment or provisioning association, configuration software
updates, and overall management and control in industries. The device management
platform provides class-leading lifecycle management for industry development.

9.4.2 IoT Application Management Platforms

IoT Application Enablement Platform (AEP) as a technology essentially offers raise to
best of breed, and industry extensible middleware core builds up a set of independent
and interconnected solutions for the customers. IoT application enablement: a com-
prehensive set of devices and enterprise backend should be connected by Software
Development Kit (SDKs) and Application Programming Interface (APIs) and the
location of well-defined documents and developer resources for IT and ­electrical
industries.

9.4.3 Application Development Platforms

The IoT application development method is a compound of a large-scale techno-
logical modernization process. The IoT confirms that it is an extension of interaction
between applications and people through new dimensions of “things” for commu-
nication. The main principle for developing the IoT application is to ensure the safe
collection of data. At the early stage of IoT deployment, deriving domain-specific
applications is the primary development strategy for the healthcare industry. In the
developing phase, control and manage the high performance of data streaming. IoT
platforms connect with the cloud and increase data management efficiency in various
healthcare industries.

9.5  TRENDING IoT PLATFORM FOR INDUSTRY REVOLUTION 4.0

IoT platforms set up a new era of cloud computing. IoT middleware solutions are
available in the market to increase compatibility worldwide. There are open and
closed source middleware platforms given, and the topmost IoT platforms for 2021
are as follows:

Google Cloud Platform: it provides multifunction secured infrastructure.

1.
It offers an efficient solution for conjecturing maintenance of devices for
smart cities and real-time asset tracking. It combines AI capabilities. It also
provides support to the wide range of latest operating systems.
Website: Google cloud platform.
Particle: the particle is called the next-generation IoT platform. This plat-
2.
form is designed by spark care. This platform is compatible with IoT prod-
ucts and WiFi or 6G connectivity. This is the leading middleware layer
platform for cellular connectivity to the cloud. This is the only online plat-
form to bring out machines, equipment, and assets. It can give us firewall-
protected clouds.
Website: particle.io.
140 Intelligent Analytics for Industry 4.0 Applications

Cisco IoT Connect Cloud: Cisco IoT products provide various solutions to
3.
the IoT platforms users. This platform increased the production efficiency
of the commerce sector. This platform secures the commerce sector with
edge intelligence. This platform modernized the edge infrastructure. This
platform also reduces network congestion for a safer transport system.
Website: cisco.com.
Amazon AWS IoT Core Platforms: AWS core IoT will help device con-
4.
nectivity to the cloud. It manages all the cloud services AWS IoT connects
with the cloud and interacts with the other devices and cloud applications. It
provides a lightweight protocol such as MQTT and HTTP. It processes vast
amounts of data and messages. It is a secure and reliable platform for trans-
mitting messages to AWS endpoints and other services. These platforms
help users to communicate and track when not connected. Users can use the
different AWS services such as AWS Lambda and Amazon quick sight.
Website: Amazon AWS IoT core.
Salesforce IoT Cloud: this platform will help to set the targeted customers,
5.
partners, devices, and sensors with relevant actions. This platform is mainly
designed for digital marketing, automation, and market analytics. It allows
users to test ideas for business without any programming skills. It can easily
interpret the data from any device. It can give real-time traffic.
Website: Salesforce.com.
IBM Waston IoT: this platform will provide the facility to capture and
6.
interrogate data and devices, machines, and equipment and find out the
understanding for better decisions. IBM platform will work on optimiza-
tion processes and resources. It can provide a bidirectional communication
facility. It can have features of AI or analytics. It is the expertise of domains.
It offers better security solutions and captures real-time data access.
Website: ibm.com/Internet-of-things IBM WastonIoT.
Microsoft Azure IoT Suite: the platform is mainly designed according to
7.
industrial requirements. This platform is primarily used by manufactur-
ing to transportation to retail. It provides solutions for intelligent spaces,
remote monitoring, and connected products. This platform is very much
compatible. Skilled and unskilled people can use this platform. This
­
­platform provides two solutions to open as an IoT or Sass and open-source
IoT templates.
Website: Microsoft Azure IoT suite, azure.microsoft.com.
Voracity Platforms Manage IRI Voracity: the massive, large databases. It
8.
is a fast, the adaptable platform for data searching, navigation, and reporting
device data. It is a small imprint for data manipulation in rapid aggregation
on the edges, it connects and integrates sensors, logs, and many more data
sources. It is used for data filtering, modification, mining, and reporting.
It runs on various platforms such as Linux, UNIX, and windows, from a
Raspberry Pi to a Z/Linux mainframe. Universal forwarder option for agile
composition and direct indexing techniques on Splunk for cloud analytics
and actions on the cloud IoT data.
Website: iri.com.
Trending IoT Platforms on Middleware Layer 141

9. Samsung Artik Cloud: this platform has the distinct feature of interoper-
ability. This platform dependency is based on hardware as a service. This
platform depends upon cloud Monetization with IoT devices. It can have
in-built security features for devices, applications, and a user-friendly envi-
ronment. It keeps the connectivity data with all the devices of the cloud. In
the future, it will be used in smart homes and IIoT.
Website: Samsung Artik.
10. ThingWorx: it maintains the management cycle of IoT applications. It can
provide features for quickly accessing data on IoT from off-premises and
off-premises from the hybrid environment. ThingWorx is compatible in
various aspects such as costs, visibility control, and improved connectivity
IIoTs. It can be used in IIoTs and cloud applications and can access the web
server’s data on-premises.
Website: ptc.com, ThingWorx.
11. Altair Smart Works: Altair brilliant works till the end of a process. It pro-
vides an end-to-end platform. It provides the forum as a service. It can have
an open architecture. Using this platform or intelligent networks, we con-
nect the devices such as sensors, features detection, and gateways machines.
It can provide good features and functionalities.
Website: altairsmartworks.com.
12. Oracle IoT: this is a much-preferred middleware platform used more effi-
ciently. Oracle IoT platform is more important for innovation as well as
animation of devices with IoT. It is simple and easy to access. Cloud systems
provide a two-way connection between the device and the cloud developed
with IoT technology. This platform architecture is designed to connect vari-
ous devices at the same time. This platform provides a very high-security
level.
Website: Oracle İoT.
13. SAP: SAP IoT platforms make business technology smarter. This IoT plat-
form uses power-driven business tools with a large volume of data and is
easily managed. SAP runs intelligent business processes autonomously at
the edge and orchestrates from the cloud. This platform also performs flex-
ible and reliable message processing. SAP has included features such as IoT
devices and integration services and provides scalable data for integration.
Website: www.cloud4c.com.
14. Itron: Itron platform uses the products and purpose of making energy-­
efficient intelligent devices. This platform provides power-safe, reliable, and
resilient energy and water systems. This platform offers services to the cus-
tomer at a meager cost. This platform effectively manages all the resources.
Website: www.itron.com.
15. Predix: Predix is an intelligent IIoT platform used in the industries for
cloud formation and digital applications. This platform is used to secure
and scalable IoT. This platform is used for asset and operations performance
management in IIoT. This platform provides edge-to-edge connectivity,
data analytics, and data processing to many IoT initiatives.
Website: www.ge.com.
142 Intelligent Analytics for Industry 4.0 Applications

16. Deutsche Telecom: Deutsche telecom is the most versatile IoT platform.
This platform works on the SAAS; it is dependent upon the software as
a service solution. This platform is used for the transformation of classic
companies into digital service providers and enables them into new business
models in the easiest way. This platform is rapidly used by the automotive,
manufacturing, retail, and logistics industries.
Website: www.telekom.com.
17. Kaa IoT Platforms: Kaa is an open-source middleware platform. This
­platform has the feature to complete at the end IoT solutions, interacting
application, and intelligent digital devices. The Kaa IoT platform comprises
a Kaa cluster control service, bootstrap service, operation service, and third-
party components such as zookeeper, SQL database, and No SQL database.
Kaa IoT platforms are also called next-generation platforms because of
internode communication methods.
Website: www.kaaiot.com.

In Table 9.1, the IoT platform is taken from 2008 to 2021. Their working procedure
and websites and the development procedure. Reference model for IoT middleware –
IoT is done with the makeover of beautiful scenarios and presented where gadgets
are used to study user requirement and act accordingly to improve experience and
quality of life. Middleware platform that integrates data from several devices and acts
accordingly. Due to this reason, the IoT middleware layer is present in IoT scenarios.
The middleware platform is the most intelligent and responsible part of IoT. The
requirements of IoT platforms should be represented by IoT architecture. Some prob-
lems that are faced by the middleware layer on IoT are as follows:

Interoperability: the IoT is the combination of heterogeneous devices to

1.
cooperate to realize a familiar task or transmit messages to each other. The
middleware layer required a platform that should be interoperable for het-
erogeneous things and objects predicted in future. The global sensor net-
work is used to fulfill the requirements of middleware IoT platforms. The
main drawback is that it is not compatible with all types of sensors, which
can contradict the interoperability purposes.
Trust: As we know that IoT can be used in healthcare industries. Many
2.
surgeries are performed at a distance using IoT healthcare tools and devices.
In these cases, any unexpected situation can occur which causes the patient
death.
Scalability: IoT is a trending technology in 2020. Sixty million devices may
3.
be connected to it with the help of the Internet. The proposed middleware
platform should handle the increasing number of devices and ensure that
they function appropriately from different locations.
Heterogeneity Abstraction: IoT is a combination of heterogeneous things
4.
having different processing, sensing, and communication strategies of vari-
ous software. They are used in many applications for multiple inputs and
outputs. These middleware platforms provide a higher abstraction level to
provide low-level communication complexity.
TABLE 9.1
IoT Platforms Year-Wise Development and Website
S. No. Platform Invented Year Area Website
1. Google Cloud IoT Platform 2008 Government Healthcare, Retail, Media, Energy, Finance Cloud.google.com
2. Particle 2013 OTA Software Updates, Sim and device, Data Governance Particle.io
3. Cisco IoT Connect Cloud 2014 Mobile Service, First virtualized packet core, Virtual Network Service Cisco.com
4. Amazon AWS IoT Core 2015 Analytics services, Device Software, Connectivity control Aws.amazon.com
platform
5. Salesforce IoT 2015 CRM, Cloud Service, Field Service, Community Salesforce.com
6. IBM Waston IoT 2015 Market-leading AI and analytics security, Dedicated Privacy ibm.com
7. Microsoft Azure IoT 2015 Energy azure.microsoft.com
Process Manufacturing
Automotive Retail
8. IRI Voracity 2016 Big Data Analytics, ETZ Modernization Iri.com
Data Governance
Trending IoT Platforms on Middleware Layer

9. Samsung Artik Cloud 2016 Device Simulator, API Console arktik.io

10. Thingworx 2017 Real-time visibility, Plant Benchmarking, Digital Work ptc.com
11. Altair Smart Works 2017 Performance optimization, Expands product life Altairsmartworks.com
12. Oracle IoT 2018 Asset Monitoring, Fleet Monitoring, Production Monitoring, Connected Worker, oracle.com
Cloud Service
13. SAP 2019 SAP has provided the best services to cloud IoT platforms over the last couple of www.cloud4c.com
years. They offer the most trending technology, machine learning, and big data.
14. Itron 2019 It is used for company specialization in renewable energy resources. It is used for www.itron.com
resource monitoring and optimization.
15. Predix 2020 It is an IoT platform that decided to take the steps to pass the game. It provides www.ge.com
connectivity and analytics to the aviation sector.
16. Deatsche Telekom 2021 This platform provides better connectivity. These platforms are experts in mobile IoT. www.telekom.com
They have the advantage of 5G Connectivity and the ability to update the new
standards
143

17. Kaa IoT Platform 2021 Kaa IoT platform provides better speed in the case of IoT implementation. It provides www.kaaiot.com
a comprehensive set of IoT Features. It includes onboard IoT analytics.
144 Intelligent Analytics for Industry 4.0 Applications

5. Spontaneous Events: spontaneous events occur randomly, and there

should be unpredictable and abrupt connections between objects at a given
time. If the randomness of the thing is increased, then mobility factors also
increase.
6. Multiplicity: smart objects have minimal filtering and memory storage.
Multiple services have been designed for a given bright object and nodes.
Every node can choose and process the situation for different results and
benefits. The middleware platform should be flexible, but it can partially
support multiplicity.
7. Security and Privacy: IoT devices have mainly two types of security:
­operational safety and data security. Data security process exchanging
things that should maintain the security policy, data integrity, nonrepu-
diation, confidentiality, authorization, authentication, availability, opera-
tional security, and performance, which should never lose data whether the
­network should be lost.
8. Bootstrapping: It is a crucial procedure in IoT. It ensures that intelligent
things compute the network to collaborate with currently available objects. In
IoT, bootstrapping should be done in an appropriated manner. Smart objects
transmitted their message to their neighboring nodes that should be present.
Its adjacent nodes have the responsibility that they can independently com-
bine the network. That technique must be added to the middleware.
9. Modularity: Modularity is mainly connected to security. It provides two-
way communication in Internet-connected things and cloud management
such as Amazon AWS IoT. Modularity is a further significant component
implemented in complicated applications. Middleware in IoT platforms
should be managed with the help of ubiquitous computing and heteroge-
neous program for study similarity to provide a smooth performance which
is extremely tough.
10. Extensibility: Around 60 million devices are connected with the IoT plat-
forms. It is a transformative forwarding technology that is emerging. Its pro-
gression added advanced challenges to the middleware platform. It should
grant the latest devices to connect independently to IoT devices. With over
60 million devices connected to IoT platforms, this transformative technol-
ogy is rapidly emerging and bringing new challenges to middleware plat-
forms. To ensure the latest devices can connect seamlessly to IoT devices,
middleware platforms must continuously evolve to meet these advanced
challenges. The IoT presents both challenges and opportunities for organi-
zations and researchers in all fields. The powerful devices that interact with
IoT platforms can perform tasks with precision and accuracy, delivering
reliable results.

9.6 CONNECTIVITY OF IoT PLATFORMS MIDDLEWARE

LAYER USING 6G TECHNOLOGY
The IoT platform and the IoT have a multidisciplinary approach to connecting with
devices. The broadband connectivity is going on with geospatial and terrestrial
Trending IoT Platforms on Middleware Layer 145

satellites. The continuous connectivity required a high bandwidth network. The high
bandwidth network is coming with the advancement of many features, so this net-
work is called the future network. In this era, the future network is 6G and 5G. The
6G network connectivity maintains the connectivity of the IoT platform in the proper
way. IoT platform connectivity with this network makes it more efficient, relevant,
and accurate. The connectivity of a 6G network with IoT platforms speeds up the
process and provides the best services to the user. This chapter presents a proposed
structure of 6G network connectivity with the IoT platform’s middleware layer. This
proposed structure helps to perform more experiments with IoT platforms, their net-
work connectivity, bandwidth, and changes in the venue. This model allows you to
accomplish the performance objective, which are little packets of network and faster
channel access for low latency and multidimensional links involving multiple access
channel reliability and connectivity. The 6G networks have subnetworks, hyper-spe-
cialized slicing, and rancor coverage. This network has advanced paradigms regard-
ing new security, protection, privacy, and policy. The 6G network blindly supports
long-tailored solutions and mobile network features as a platform. The 6G network
is compatible with IoT platforms using future state prediction, correlation recovery,
proactive exploration, and self-learning. The proposed model shows the 6G net-
work’s extensive connectivity, consisting of satellites, towers, fiber optic, telecom-
munication orders, and IoT platforms connectivity with the network and end users
with the ­middleware layer in Figure 9.3. This model consists of IoT platforms used
by smart cars.

FIGURE 9.3  IoT middleware layer connectivity with 6G network Internet access procedure.
146 Intelligent Analytics for Industry 4.0 Applications

9.7 CONCLUSION
This chapter examined the main issues faced by the IoT middleware. Thus, we high-
light the middleware approach depends on their functional and nonfunctional tech-
niques and discuss the different platform approaches. This research provides an idea
about the pros and cons of the existing middleware approaches. It highlights the main
problem to deal with yet not well explored, such as security, privacy bootstrapping,
and reliability. The middleware platform developers provide extra time to make them
more user-friendly without compromising safety, as usability with a certain degree of
quality might be the key to prosperity in the crowded market. In the future, the con-
nectivity of IoT platforms is increased with the latest networks such as 7G or 8G to
make the ecosystem and society smarter. In future, the industry become fully auto-
mated these fully automated industries gave the birth to Industry 4.0. In automated
industries, the advanced network is required such as 6G, 7G, and 8G. These networks
are installed in the industries by applying routers at different locations, RFID tags,
sensors, etc. In this chapter, we have only given the idea of network connectivity, but
many more experiments are performed in this regard.

REFERENCES
1. da Cruz, M.A., Rodrigues, J.J.P., Al-Muhtadi, J., Korotaev, V.V., and de Albuquerque,
V.H.C., 2018. A reference model for internet of things middleware. IEEE Internet of
Things Journal, 5(2), pp. 871–883.
2. Fersi, G., 2015 June. Middleware for the internet of things: A study. In 2015 International
Conference on Distributed Computing in Sensor Systems (pp. 230–235). IEEE.
3. Pallavi, K.N., Kumar, R., and Kulal, P., 2019. Study of security algorithms to secure
IOT data in middleware. In Proceedings of the 2nd International Conference on Green
Computing and Internet of Things, ICGCIoT 2018. VVCE.
4. Kokkonis, G., Chatzimparmpas, A., and Kontogiannis, S., 2018 September. Middleware
IoT protocols performance evaluation for carrying out clustered data. In 2018 South-
Eastern European Design Automation, Computer Engineering, Computer Networks
and Society Media Conference (SEEDA_CECNSM) (pp. 1–5). IEEE.
5. Silva, J.D.C., Rodrigues, J.J.P., Saleem, K., Kozlov, S.A., and Rabêlo, R.A., 2019.
M4DN. IoT-A networks and devices management platform for internet of things. IEEE
Access, 7, pp. 53305–53313.
6. Sandor, H., Genge, B., Haller, P., and Bica, A., 2019 June. A security-enhanced interop-
erability middleware for the internet of things. In 2019 7th International Symposium on
Digital Forensics and Security (ISDFS) (pp. 1–6). IEEE.
7. Bakhshi, Z., Balador, A., and Mustafa, J., 2018 April. Industrial IoT security threats
and concerns by considering Cisco and Microsoft IoT reference models. In 2018
IEEE Wireless Communications and Networking Conference Workshops (WCNCW)
(pp. 173–178). IEEE.
8. Hoffmann, J.B., Heimes, P., and Senel, S., 2018. IoT platforms for the Internet of pro-
duction. IEEE Internet of Things Journal, 6(3), pp. 4098–4105.
9. Yadav, E.P., Mittal, E.A., and Yadav, H., 2018 February. IoT: Challenges and issues in
Indian perspective. In 2018 3rd International Conference on Internet of Things: Smart
Innovation and Usages (IoT-SIU) (pp. 1–5). IEEE.
10. Yu, J.Y. and Kim, Y.G., 2019 January. Analysis of IoT platform security: A survey.
In  2019 International Conference on Platform Technology and Service (PlatCon)
(pp. 1–5). IEEE.
Trending IoT Platforms on Middleware Layer 147

11. Frustaci, M., Pace, P., Aloi, G., and Fortino, G., 2017. Evaluating critical security issues
of the IoT world: Present and future challenges. IEEE Internet of Things Journal, 5(4),
pp. 2483–2495.
12. Conti, M., Dehghantanha, A., Franke, K., and Watson, S., 2018. Internet of Things
Security and Forensics: Challenges and Opportunities.
13. Viswanathan, H. and Mogensen, P.E. (2020). Communications in the 6G era. IEEE
Access, 8, pp. 57063–57074. https://doi.org/10.1109/ACCESS.2020.2981745
14. Barakat, B., Taha, A., Samson, R., Steponenaite, A., Ansari, S., Langdon, P. M., Wassell,
I. J., Abbasi, Q. H., Imran, M. A., and Keates, S. (2021). 6G Opportunities Arising from
the Internet of Things Use Cases: A Review Paper, 1–29.
15. Nawaz, S.J. and Member, S. (2021). Non-coherent and backscatter communications:
Enabling ultra-massive connectivity in 6G wireless networks. IEEE Access. https://doi.
org/10.1109/ACCESS.2021.3061499
16. Jang, H. S., Jung, B. C., and Member, S. (2021). Resource-hopping based grant-free
multiple access for 6G-enabled massive IoT networks. IEEE Internet of Things Journal,
8(20), pp. 15349–15360. https://doi.org/10.1109/JIOT.2021.3064872
17. Nkenyereye, L. and Hwang, J. (2021). Virtual IoT service slice functions for multi-
access edge computing platform. IEEE Internet of Things Journal, 8(14), pp. 11233–
11248. https://doi.org/10.1109/JIOT.2021.3051652
18. Swamy, S. N. and Kota, S. R. (2020). An empirical study on system level aspects of
internet of things (IoT). IEEE Access, 8, pp. 188082–188134. https://doi.org/10.1109/
access.2020.3029847
 10 A Factual and Feasible
Healthcare IoT

Application of Industrial IoT

Subasish Mohapatra, Amlan Sahoo,
and Subhadarshini Mohanty
Odisha University of Technology and Research

Munesh Singh
PDPM IIITDM

CONTENTS
10.1 Introduction................................................................................................... 150
10.2 Applications................................................................................................... 151
10.2.1 Smart Grids........................................................................................ 151
10.2.2 Smart Flights..................................................................................... 151
10.2.3 Smart Agriculture and Farming........................................................ 151
10.2.4 Smart Manufacturing........................................................................ 151
10.2.5 Smart Cities....................................................................................... 151
10.2.6 Shopping and Retail........................................................................... 153
10.2.7 Smart Healthcare............................................................................... 153
10.2.8 Smart Vehicles................................................................................... 154
10.3 Related Works................................................................................................ 154
10.4 Healthcare IoT............................................................................................... 154
10.5 IoT Analytics in Healthcare........................................................................... 156
10.6 Benefits.......................................................................................................... 158
10.6.1 Real-Time Reporting and Monitoring............................................... 158
10.6.2 End-to-End Connectivity and Affordability...................................... 158
10.6.3 Data Assortment and Analysis.......................................................... 158
10.6.4 Tracking and Alerts........................................................................... 158
10.6.5 Remote Medical Assistance.............................................................. 158
10.7 Challenges..................................................................................................... 158
10.7.1 Data Security and Privacy................................................................. 158
10.7.2 Integration of Multiple Devices and Protocols.................................. 159
10.7.3 Data Overload and Accuracy............................................................. 159
10.7.4 Cost.................................................................................................... 159
10.8 Suggested Framework.................................................................................... 159
10.9 Conclusion..................................................................................................... 162
References............................................................................................................... 162

DOI: 10.1201/9781003321149-10 149

150 Intelligent Analytics for Industry 4.0 Applications

10.1 INTRODUCTION
Modern social and economic trends, such as the Sustainable Agenda 2030 granted
by the United States [1], aim to boost the nature of well-being conditions to increase
life expectancy. Real-time implementation of this agenda has been made possible by
recent technological breakthroughs. To do this, several focused technological efforts
have been implemented. For instance, one such program is “Healthcare 4.0,” which
is being created in response to the advent of maintaining the digital well-being stan-
dard for healthcare services [2]. Additionally, recent technological advancements
have made it possible to remotely and automatically monitor healthcare services
using medical equipment that are designed to keep track of a patient’s numerous
health issues. It is important to note that these gadgets operate independently and
are devoted to particular health issues. Industry 4.0, which over the previous decades
had made the healthcare industry increasingly digital, gave rise to “Healthcare 4.0.”
For instance, X-rays, MRI, and ultrasound scans have all evolved into computer
tomography and electronic medical records, respectively [3]. Caretakers and medical
professionals utilize these user-centered gadgets to monitor and treat patients’ medi-
cal issues as well as provide preventative care and wellness solutions. The usage of
things in the Internet of Things (IoT) technology and their respective applications
are expanding rapidly. As a result of recent technological advancements, such as
industrial Internet of Things (IIoT) and industrial cyber-physical systems (ICPSs) [4]
predicted that by the end of 2021, there will be 212 billion IoT devices and apps, with
healthcare accounting for around 41% of their total usage. In addition, Grand View
Research [5] estimates that by 2025, the healthcare IoT industry would be valued at
USD 534.3 billion. The major motive behind this technology-driven application in
healthcare is to provide better and reliable assistance along with dynamic healthcare
operations by remotely and continuously regulating the patients’ varied health prob-
lems [6–8].
This study has provided the essential components of HIoT (Healthcare IoT) to
enable medical personalization of the traditional healthcare system as opposed to
the current ones that merely support any technical assistance. Thus it offers various
contributions; to promote clinical personalization, the authors contextualized HIoT
(Healthcare 4.0) to define unique complete customized healthcare services (CPHS).
Various measuring parameters of the HIoT are determined. As an illustration of con-
temporary HIoT, a use-case scenario has been provided. To establish the identified
HIoT requirements in every layer of the suggested architecture, proper investigation
and analysis points have been noted. Finally, dependable, robust, real-time, com-
plete, individualized healthcare services, which solve the stated shortcomings of the
current methodologies, are discussed in the framework section.
The remaining sections are structured as follows: various applications of the IIoT
are discussed in Section 10.2. Section 10.3 derives some of the related works. The
introduction to the personalization of healthcare IoT depicted is in Section 10.4. The
analytics for the IoT-based healthcare systems, their many levels, and current ini-
tiatives to define criteria for contemporary customized services are all described
in Section 10.5. Major advantages and disadvantages are derived in Section 10.6.
Similarly, Section 10.7 elaborates on the suggested framework. Finally, Section 10.8
brings our chapter to a close with the conclusion and some of the future aspects.
Healthcare IoT 151

10.2 APPLICATIONS
It is the height of automation. Every tangible item that exists becomes empowered
and linked in an enterprise setting. A new generation of appealing automation is
coming. The IT landscape with cloud technology has taken an important place
thanks to its unstoppable ability to rapidly and simply accomplish the long-standing
aim of infrastructure optimization. The integration of diverse systems and sensors is
the main challenge. In order to make decisions, it is also critical to draw conclusions
from streaming data from all sources. IoT applications are not constrained to a single
category since they may be implemented in any type of equipment depending on the
situation [9]. Electrical, mechanical, physical, and electronic components with smart
labels, barcodes, LED lights, beacons, and pads are the main elements present in any
IoT ecosystem. The sections that follow explain the most well-known and significant
business, consumer, and individual use cases that have emerged from all the many
IoT-related innovations (Figure 10.1).

10.2.1 Smart Grids
It is a networked gadget that measures a building or home’s use of energy, water, or
natural gas. Due to the effectiveness of smart grids, power providers began using
them instead of traditional meters to measure overall usage.

10.2.2 Smart Flights
The fusion of artificial intelligence (AI) and IoT works together to maximize advan-
tages while minimizing expenses for the passengers. Adding AI to jet engines,
the new C-series airliner featured the geared turbofan (GTF) engine from Pratt &
Whitney with 5,000 sensors that could generate up to 10 GB of data per second.

10.2.3 Smart Agriculture and Farming

Farmers may prepare for meteorological conditions like rain, drought, snow, or wind
by forecasting them in advance. Microclimate conditions may be managed using
Heating, ventilation, and air conditioning (HVAC) and greenhouse environment sen-
sors. Utilizing remote monitoring and wireless sensors, smart farming reduces the
resource consumption of IoT-based pest control and smart tractors.

10.2.4 Smart Manufacturing
Supply chain management and intelligent logistics are both used by the IIoT.

10.2.5 Smart Cities
Smart cities, which are described as municipalities that employ data and communi-
cation technology to improve operational efficiency, exchange information, and the
standard of public services, are made possible by the IoT. The IoT can be used for
the several problems such as long wait times to find parking; exceeding emissions
152 Intelligent Analytics for Industry 4.0 Applications

Smart Grids

Smart Flights
Real-time
Monitoring
Smart Farming
Real-time
Maintenance

Efficient
Monitoring

Traffics and
Smart Cities
Pollution Control

Internet of Things Shopping and Real-time

Retail Communication

Real-time
Smart Home
Monitoring

Tracking Staff,
Patients and
Inventory
Smart Healthcare
Remote Health
Monitoring

Auto Insurance

Smart Vehicles Fleet Management

Transports and
Logistics

FIGURE 10.1  Various applications based on the industrial IoT.

thresholds; and unauthorized parking. To save needless emissions and traffic, one
must choose to park straight away in a nearby garage rather than driving about
the city center looking for an on-street spot. Better monitoring cities can automate
Healthcare IoT 153

operations that can be time-consuming or expensive, thanks to ingenious parking

services and management solutions.

10.2.6 Shopping and Retail

The IoT enables users to link their mobile devices to storefront windows, discover
what they need, and receive helpful advice or perks in exchange for their loyalty.
Smart homes: If a room is empty, smart home sensors adjust the thermostat to
lower the temperature and switch off the lights.

10.2.7 Smart Healthcare
IoT improves supply chain management and rationalizes asset tracking. The use of an
inventory management system allows individuals to locate necessary tools quickly.
Patients may be able to receive remote health monitoring instead of going to the
emergency department in some circumstances. The most common use in healthcare
is remote health monitoring.

FIGURE 10.2  Various use cases of the Internet of Things.

154 Intelligent Analytics for Industry 4.0 Applications

10.2.8 Smart Vehicles
Making driving safer by using sensors, businesses created technology that incorpo-
rates sensors into smartphones to analyze driving habits. Data collection allows the
app to provide training for safer driving (Figure 10.2).

10.3  RELATED WORKS

The growth of healthcare system might raise the healthcare of anybody at anytime
and anywhere, therefore minimizing other limits with the rise in coverage and
­quality. Remote consultation, also known as teleconsultation, has gained a lot of rel-
evance these days. This process makes use of a broadband network and bidirectional
control technologies with high resolution.
There are certain gaps in the research work of the various IoT layers within the
framework of the needs for inclusive customized healthcare services [10]. These
resulting gaps identify research issues related to various architectural levels and
accompanying essential healthcare requirements. The identified gaps result in the
conclusion are consequently reliable for a patient with various diagnostic condition as
the existing architecture fall short of providing dependable, resilient, individualized,
and application-specific healthcare. Similarly, it fails to comprehend the connection
and impact of a person’s many health issues [11]. The various comparison demon-
strates unequivocally that only a small number of the aforementioned techniques
fully or partially enable the personalization icon of healthcare services taking into
account a patient’s various health problems.

10.4  HEALTHCARE IoT

Healthcare is a type of medical treatment that involves keeping an eye on patients
to see if they are acting normally or abnormally and adopting preventative measures
to keep them healthy. Two components of the technology employed in healthcare
are, healthcare software and computers allowing it 24/7. The majority of healthcare
technologies improve the way healthcare is now delivered, but ubiquitous healthcare
uses a paradigm that shifts the focus from doctors to patients. Figure 10.3 shows a
sample of healthcare architecture. Medical technologies are constantly crucial since
it is vital in managing health and quality of life. IoT has access to health monitoring,
which enhances quality while lowering costs. Patients and doctors can accurately
identify the ailment with the help of smart devices [12]. The methods used now,
which include mobile devices and external sensors, are archaic. When compared to
other systems, healthcare systems may occasionally be the least efficient, yet there
are several approaches to make them more effective. Currently, diagnostic mistakes
account for 20% of medical errors, 20% of which are caused by inaccurate diagnosis
methods, and 20% of which are caused by treatment delays.
By gathering analysis using patient data and medical resources, connected
healthcare devices can improve patient monitoring and medical decision-making.
There are connectivity problems with many types of equipment, but there are some
­cutting-edge items on the market that are already demonstrating the trend of IoT in
Healthcare IoT 155

FIGURE 10.3  Characteristics of the smart healthcare solution.

Embedded
to body
devices

Components
External
smart of Solutions
for health
medical
appliances Healthcare monitoring

IoT

Stationary
smart
medical
devices

FIGURE 10.4  Components of healthcare IoT.

healthcare. The IoT components in healthcare are shown in Figure 10.4. According
to current customized healthcare is a service that adapts medical care to specific
people by identifying common traits as such genetics, heredity, and lifestyle [13]. The
patient’s total health, which necessitates an awareness of the biological relationships
156 Intelligent Analytics for Industry 4.0 Applications

FIGURE 10.5  Illustration of partial and comprehensive healthcare architecture.

Home-
Mobile
based
Devices
devices

Health Web
Records Resources

FIGURE 10.6  IoT used scenarios in smart health care.

among many health disorders, cannot be personalized using this approach, which
only assistances to personal specific health issues. Additionally, the specifics of
personalized healthcare vary depending on the patient because every person has
a different set of health condition determinants and traits, which are essential for
maintaining a person’s personalized health. By minimizing the negative impacts of
the patient’s many medical problems (such as heart disease, high blood pressure,
or insulin resistance), element-based optimization seeks to enhance the patient’s
overall long-term health. Establishing a comprehensive customized healthcare ser-
vice for patients and enhancing their enduring health may be made feasible through
Healthcare 4.0 (Figures 10.5 and 10.6).

10.5 IoT ANALYTICS IN HEALTHCARE

We live in a world where data are more valuable than the rarest minerals and more
lethal than the most powerful weaponry. But it is essentially useless in its unprocessed
Healthcare IoT 157

state. It must first be treated before it can be used. There are sensor, network, and
decision layers in a typical IoT design. Data from the environment are collected by
the sensor layer and sent to the processing layer across the network [14]. Data analysis
is carried out in this instance to assist the decision layer in drawing judgments. Data
analytics is hence the IoT’s foundation. IoT researchers have always found the health-
care industry to be an attractive business. Additionally, the global lack of medical
experts has fueled the expansion of such studies.
The following reasons for the numerous healthcare procedures that may be
­optimized as a result of a strong and rapid data analysis of an IoT system:
Disease diagnosis: the physiological data, including blood pressure, body tem-
perature, and others, acquired by wearable IoT devices can be utilized to diagnose
diseases faster, in addition to other test results (urine tests, blood tests, magnetic
resonance imaging [MRI] scans, etc.).
Health prevention and lifestyle management: the treatment of a chronic patient’s
lifestyle is aided by routine analysis of data about behavioral elements of a person,
such as sleep patterns, eating habits, and daily physical activity. A healthy person
may be informed of the diseases they may be at risk for owing to their present life-
style by such an examination.
Medical emergency: if an abnormality in the physiological data is quickly identi-
fied, an emergency (such as an old person falling and fainting in his or her home) can
be effectively managed. The elderly and those with terminal illnesses will benefit the
most from this (Figure 10.7).
Remote real-time health monitoring: real-time health monitoring is one of the
main analysis applications in the healthcare industry. IoT sensors gather and transmit
the patients’ physiological data. The data are then immediately examined. Every
time a deviation from the usual pattern is found, a thorough report is sent to the
patient’s healthcare provider for additional investigation.
To enable IoT systems and applications to behave adaptively, these collected
and aggregated insights may be looped back into them. In other words, data-driven
insights enable hardware and software to act intelligently. IoT data must be meticu-
lously collected since it conceals a range of useful patterns, correlations, recommen-
dations, information, knowledge, etc.

FIGURE 10.7  Utility of IoT-based data analytics in healthcare.

158 Intelligent Analytics for Industry 4.0 Applications

10.6 BENEFITS
10.6.1 Real-Time Reporting and Monitoring
In the case of a medical emergency, real-time monitoring via associated devices can
save a million lives. Connected devices can gather important medical and health-
related data when real-time condition monitoring is implemented using a smart
medical device connected to a smartphone app. The IoT-connected gadget gathers
and transmits health data, including measurements of various body parameters. The
information is kept in the cloud and can be sent to a person who has been permitted
to view it which is made available ubiquitously.

10.6.2 End-to-End Connectivity and Affordability

Through healthcare mobility solutions, IoT may assist in transferring the health-
care data, which greatly reduces the cost of a result, and this technology-driven
arrangement  can reduce expenses by reducing needless trips and employing
­
resources of higher quality enhancing resource allocation and planning throughout
the process.

10.6.3 Data Assortment and Analysis

It is not as simple as it seems for healthcare practitioners to manage a large volume
of data. Through mobility solutions driven by IoT, real-time data gathered by IoT-
enabled mobile devices may be analyzed and segmented.

10.6.4 Tracking and Alerts

With continual notifications and real-time alerts for appropriate analysis, tracking
and alerts in life-threatening circumstances might prove to be a lifesaver to protect
a crucial patient’s health. Real-time tracking, notification, and monitoring are made
possible by IoT-powered healthcare mobility solutions.

10.6.5 Remote Medical Assistance

When a patient needs medical care but is unable to reach a doctor because of obsta-
cles like distance or ignorance, it is a horrible scenario.
Through healthcare delivery networks connected to patients through IoT devices,
patients can take prescription medications right at home [15].

10.7 CHALLENGES
10.7.1 Data Security and Privacy
This is a major issue that IoT is now facing. Real-time data collection is not possible
unless it is complying with data standards with protocols. Regarding ownership and
control of data, there is a lot of confusion. False health claims and the fabrication of
Healthcare IoT 159

phony IDs for prescription purchases and sales are two instances of how IoT device
data have been misused.

10.7.2 Integration of Multiple Devices and Protocols

The application of IoT in the healthcare industry is hampered by the integration of
many device types. This obstacle exists because device makers cannot agree on a
communication protocol with a set of standards.

10.7.3 Data Overload and Accuracy

The inconsistency of various methods makes it challenging to aggregate data for crit-
ical understanding and analysis. IoT gathers data in large quantities, thus for efficient
data analysis, the data need to be divided into manageable chunks without being
overloaded with fine precision.

10.7.4  Cost
IoT app development for healthcare mobility solutions, costs are one of the bigger
difficulties. However, if the IoT solution addresses a real problem, the expenses are
justified. Even though the investment of a lot of resources and money into develop-
ing an IoT application, the benefits will be enormous when a company uses IoT to
improve business processes, generate more income streams, and open up new busi-
ness options while also saving time and labor costs.

10.8  SUGGESTED FRAMEWORK

The terms “Internet” and “things” are combined to form the phrase “Internet of
Things” (IoT). Smart gadgets, instruments, and ubiquitous, context-aware things that
can communicate are all referred to as things since they can connect to the internet.
For IoT to be truly widespread, IoT devices need to have the following three funda-
mental characteristics:

• Sensing Capability: IoT devices need to be able to perceive and integrate

the data they collect. For instance, numerous specialists are used in the
healthcare industry to sense and collect data on various body parameters.
The data collection is self-contained.
• Communicable: IoT devices should be able to transmit data to the vari-
ous desired data centers after sensing and aggregating using a variety of
communication mediums, such as wireless technologies, mobile networks,
Wireless Local Area Networks (WLANs), Wireless Sensor Networks
(WSNs), and Mobile Ad Hoc Network (MANET), among others.
• Actionable: The aggregated data are meaningless on their own, thus for
IoT devices to respond, they must be able to comprehend the data. Until
processing technology is utilized to detect whether a body parameter level
exceeds or deviates from the normal range.
160 Intelligent Analytics for Industry 4.0 Applications

In this process, three-tiered HIoT architecture has been taken into

c­ onsideration. The things layer comes first in this three-tier system, trailed
by the communication and application layers.

• Things Layer: This stage in the HIoT refers to sensing systems and appa-
ratus that, depending on the application, record the different values seen
from various sources (for example, apparatus monitoring patients’ various
body parts or health issues). These sensors are designed to gather data from
a patient’s body related to their ailment. The devices and sensors mentioned
above collect information from the patient’s body and, in some cases, send
it to the data center. Some responsive sensors have devices that keep track
of the relevant body parameters.
• Communication Layer: The communication layer is viewed as the founda-
tion of IoT systems. IoT devices generate a substantial amount of continu-
ous data that are challenging to transmit and store. All of the data are sent
from the things layer to the application layer through the communication
layer. An appropriate communication medium is required to convey data
from all of the Things layer nodes to the other layers of the system. The
­communication medium may be wireless (WiFi, Bluetooth, RFID, 5G, etc.)
or wired, depending on the communication protocol developed (Ethernet,
USB, etc.).
• Application Layer: The application layer focuses on delivering top-notch
healthcare services and directly connects the end user to the cloud-based
IoT platform. HIoT is a network of interconnected hardware, software,
devices, and services. Their use is expanding tremendously as a result
of network improvements. These are employed by healthcare workers to
process and analyze data to make decisions and treat patients. The IoMT
in hospitals often focuses on enhancing patient safety and/or streamlin-
ing procedures. It makes it possible for practitioners to collaborate across
disciplinary lines to provide tailored patient care. The application layer
serves two purposes by offering: On the one hand, monitoring the body
parameters and aiding in the supervisory decisive outcome; on the other,
support applications serve as user interfaces for a variety of health moni-
toring equipment. To enable tailored healthcare services, many methodolo-
gies have been created.

To achieve the objective of the overall system there is a fusion made between wear-
able and mobile technology. When needed, the system provides both residents with
cognitive difficulties with immediate help through their wearable gadgets. Automatic
alerts are sent to carers in potentially dangerous circumstances. The discussed meth-
odology consists of different elements to guarantee the operational dependability of
correlated healthcare IoT services and applications to provide individual services.
Figure 10.8 depicts the workflow of the personalized HIoT system for all-encom-
passing personalized healthcare. Data are gathered at the IoT cloud and delivered to
the concerned predictive model, where it is stored, from various monitoring devices
used by patients, such as heart rate, temperature, blood pressure monitor, etc. Data
Healthcare IoT 161

FIGURE 10.8  Flowchart of the suggested framework.

that have been stored are processed using various predictive methods and disease
models. To create the personalized HIoT system for all-encompassing personalized
healthcare, there are several obstacles to overcome.
Additionally, the monitor offers clinical reasoning to precisely pinpoint the ori-
gin of the disparity. Once the disparity has been located, the monitor may treat it
appropriately, and the treatment’s results can then be observed. The monitor remotely
and automatically manages an individual’s numerous physical conditions based on
clinical features. The clinical medication enables the monitoring process to be self-
conscious, which promotes the provision of healthcare services free from incorrect
diagnoses.
Modeling health problems is a difficult undertaking since it calls for deter-
mining the many clinical relationships between the various patient health states.
However, these dependencies are frequently complex and rely on knowledge of the
circumstances’ underlying variable-level biological processes, which are mostly
unknowable. To focus the various healthcare services based on their clinical fea-
tures, a library of distinct health problems (models) will be created that may sub-
sequently be employed in any computerized healthcare system. The library will
also aid in discovering further connections between multiple fitness disorders
which are normally outside the scope of concerned professionals. Additionally,
based on the models, it will be possible to describe the newly discovered depen-
dencies in a way that both practitioners and machines can comprehend. Similarly,
predictive approaches can be used to compare the data of many patients with
comparable illnesses to better comprehend health issues, leading to the recom-
mendation of a more individualized prognosis and therapy that will efficiently
enhance patients’ health. In several facets of healthcare, AI is essential. AI is
now employed for data gathering, communication, and monitoring at the things,
communication, and navigation levels, respectively. Due to its expandability and
trust concerns, machine-based prediction is not yet totally safe for diagnostic and
treatment purposes. As a result, essential healthcare fields do not yet completely
trust smart diagnosis systems.
162 Intelligent Analytics for Industry 4.0 Applications

10.9 CONCLUSION
Numerous industrial segments are seeing advancements because of the Internet of
Things. Healthcare is one of the sectors that has embraced this prospect the quickest,
opening up a new market-centered on HIoT. This fact inspired the authors to create
a thorough survey to evaluate the condition of the field. The most recent HIoT arti-
cles and products were found, explained, and examined to achieve this aim. Several
HIoT services and apps address societal demands but are becoming more isolated.
Based on the collected sensor data, healthcare IoT is a collection of various IoT tech-
nologies and devices used to support individual health. Healthcare IoT offers indi-
vidualized services using a single person’s gadgets. Users require a comprehensive
system to fulfill the strict requirements of health monitoring. Since the integration
of diagnostic and therapeutic processes is crucial, information should be maintained
in a manner that is simple to access and convey. An effective IoT framework for the
healthcare environment describes how to combine patient information for caregivers
and report patient progress using the technologies now employed in the healthcare
system. Patients’ physical and mental health may improve as a result of less expense
and in less time.
Additionally, healthcare professionals, experts, and the general public with an
interest in HIoT can use this document as a resource for information. However, this
analysis does not offer a thorough grasp of several essential subjects, such as HIoT
topologies, architectures, and platforms, as well as the requirements, difficulties, and
suggested security models. Other technologies, such as big data, augmented reality,
and cognitive systems, which are not included in this analysis but might be investi-
gated further, exist.

REFERENCES
1. Lee, B. X., Kjaerulf, F., Turner, S., Cohen, L., Donnelly, P. D., Muggah, R., & Gilligan,
J. (2016). Transforming our world: implementing the 2030 agenda through sustainable
development goal indicators. Journal of Public Health Policy, 37(1), 13–31.
2. Gupta, R., Bhattacharya, P., Tanwar, S., Kumar, N., & Zeadally, S. (2021). GaRuDa: a
blockchain-based delivery scheme using drones for healthcare 5.0 applications. IEEE
Internet of Things Magazine, 4(4), 60–66.
3. Bercovich, E., & Javitt, M. C. (2018). Medical imaging: from roentgen to the digital
revolution, and beyond. Rambam Maimonides Medical Journal, 9(4), e0034.
4. Giri, A., Dutta, S., Neogy, S., Dahal, K., & Pervez, Z. (2017, October). Internet of
Things (IoT) a survey on architecture, enabling technologies, applications and chal-
lenges. In Proceedings of the 1st International Conference on Internet of Things and
Machine Learning (pp. 1–12).
5. Hanson, K., & Goodman, C. (2017). Testing times: trends in availability, price, and
market share of malaria diagnostics in the public and private healthcare sector across
eight sub-Saharan African countries from 2009 to 2015. Malaria Journal, 16(1), 1–16.
6. Ramson, S. J., Vishnu, S., & Shanmugam, M. (2020, March). Applications of Internet of
Things (IoT)–an overview. In 2020 5th International Conference on Devices, Circuits
and Systems (ICDCS) (pp. 92–95). IEEE.
7. Zhan, K. (2021). Sports and health big data system based on 5G network and Internet of
Things system. Microprocessors and Microsystems, 80, 103363.
Healthcare IoT 163

8. AlShorman, O., AlShorman, B., Alkhassaweneh, M., & Alkahtani, F. (2020). A review
of internet of medical things (IoMT)-based remote health monitoring through wearable
sensors: a case study for diabetic patients. Indonesian Journal of Electrical Engineering
and Computer Science, 20(1), 414–422.
9. Vyas, S., Gupta, M., & Yadav, R. (2019, February). Converging blockchain and
machine learning for healthcare. In 2019 Amity International Conference on Artificial
Intelligence (AICAI) (pp. 709–711). IEEE.
10. Vergutz, A., Noubir, G., & Nogueira, M. (2020). Reliability for smart healthcare: a
network slicing perspective. IEEE Network, 34(4), 91–97.
11. Zhang, J., Li, L., Lin, G., Fang, D., Tai, Y., & Huang, J. (2020). Cyber resilience in
healthcare digital twin on lung cancer. IEEE Access, 8, 201900–201913.
12. Yan, C., Duan, G., Zhang, Y., Wu, F. X., Pan, Y., & Wang, J. (2020). Predicting drug-
drug interactions based on integrated similarity and semi-supervised learning. IEEE/
ACM Transactions on Computational Biology and Bioinformatics.
13. Singh, D., Mishra, P. M., Lamba, A., & Swagatika, S. (2020). Security issues in dif-
ferent layers of IoT and their possible mitigation. International Journal of Scientific &
Technology Research, 9(4), 2762–2771.
14. Qadri, Y. A., Nauman, A., Zikria, Y. B., Vasilakos, A. V., & Kim, S. W. (2020). The
future of healthcare internet of things: a survey of emerging technologies. IEEE
Communications Surveys & Tutorials, 22(2), 1121–1167.
15. Fani, M., Rezayi, M., Meshkat, Z., Rezaee, S. A., Makvandi, M., & Angali, K. A.
(2020). A novel electrochemical DNA biosensor based on a gold nanoparticles-reduced
graphene oxide-polypyrrole nanocomposite to detect human T-lymphotropic virus-1.
IEEE Sensors Journal, 20(18), 10625–10632.
 11 IoT-Based Spacecraft
Anti-Collision HUD
Design Formulation
Bibhorr
IUBH International University

CONTENTS
11.1 Introduction................................................................................................... 165
11.2 Related Work................................................................................................. 166
11.3 Cyber-Physical Networks in Space................................................................ 166
11.4 Spacecraft Anti-Collision System Design..................................................... 167
11.4.1 Traffic Identification Design.............................................................. 168
11.4.2 Density Evaluation Model................................................................. 169
11.4.3 Risk Assessment Model..................................................................... 170
11.4.4 Safe Trajectory Planning Model........................................................ 171
11.4.5 Heads-Up Display System................................................................. 173
11.5 Conclusion and Future Scope........................................................................ 173
References............................................................................................................... 174

11.1 INTRODUCTION
The Earth’s orbit is becoming increasingly congested. Reusable rockets and
­increasingly powerful smallsats are decreasing the obstacles to accessing and using
space. In the previous decade, the cost of commercial launches has decreased by a
factor of 20 [1]. Increased space activity has resulted in a significant increase in the
number of objects in space and this is only expected to grow exponentially in the
future. As the number of objects in Earth’s orbit grows, the chances of a satellite col-
liding with space junk or another satellite are increasing [2]. The distribution of space
objects whether vehicles, satellites, debris, etc. is also varied with some areas in the
space with a high density of these space objects while others with moderate to low
density. The density of space traffic elements is therefore a crucial factor in determin-
ing the traffic scenario and subsequent maneuvering of vehicles. In practical applica-
tions of AI-controlled spacecrafts and trajectory planning integrated autonomous
systems, maneuvering these in an area with obstructions in the path is challenging
and requires advanced capability in these systems other than the very basic feature
of ability to avoid collisions with other space elements, viz., other spacecrafts, debris,
or space devices orbiting for any other experimental purposes. Increased space activ-
ity can significantly aggravate the space debris situation [3]. After 60 years of space
DOI: 10.1201/9781003321149-11 165
166 Intelligent Analytics for Industry 4.0 Applications

activities, the orbital population appears to be unevenly distributed, with more than
90% of the recorded objects being nonfunctional i.e., orbital garbage. Furthermore, a
large number of active spacecrafts lack a propulsion engine and hence are unable to
maneuver to prevent accidents [4].
Due to technological and political concerns, the space traffic control model has
historically been presented in a vague, confusing, and imprecise way [5]. The notion
of space traffic management is complicated. It is critical to design a model capable
of successfully and efficiently regulating flightpath for collision prevention before
constructing a comprehensive space traffic management system [6].
With an increased number of space objects, the collision probability has increased
manifold. Because of the growing quantity of space objects, space is becom-
ing increasingly congested. It is a given fact that the number of objects in space
­surrounding the Earth will continue to grow in the near future, increasing the like-
lihood of a collision, communication failure, or other undesired occurrences [7].
To mitigate the collision risk, it is therefore crucial to assess the collision probability
and associated collision risk with high precision and much before the elements come
close enough to appropriately maneuver and change the trajectories of the involved
controllable elements.

11.2  RELATED WORK

Min et al. [8] detailed a collision avoidance method for satellites with formation
flight highlighting the formation initialization and formation reconfiguration opera-
tions. Omar et al. [9] discussed a spacecraft collision avoidance technique employing
aerodynamic drag. In order to avoid collision with inactive space objects, Lee et al.
[10] proposed a semianalytic and suboptimal guidance control for an active space-
craft. Wang et al. [11] discussed a two-spacecraft collision avoidance problem. Hobbs
et al. [12] proposed four types of collision avoidance. Similar other lone methods and
techniques are being researched and analyzed rarely highlighting artificial intelli-
gence and machine learning. Such works are all lone techniques based on condition-
alities and conventionally propagated models, lacking the desideratum elements of
futuristic aspects and parameters involving artificial intelligence, machine learning,
and Internet of things. Moreover, such works are not able to typically establish inclu-
sive solutions for the collision avoidance and management to incorporate multifac-
eted models/scenarios. Hence, this work has been incepted to put forth multimodel
solutions and all-inclusive approach to manage space traffic accidents in an effective
way irrespective of engineering conditionalities.

11.3  CYBER-PHYSICAL NETWORKS IN SPACE

Aerospace industries are aspiring to station data center networks into the Earth’s
orbit and in deep space, the reason for which is progressing feasibility as the cost of
manufacturing and launching satellites is decreasing; they are eyeing on new busi-
ness models by accomplishing this concept. The space industry aspires to replicate
the IT sector’s approach of colocation services regarding server hardware, edge
IoT-Based Spacecraft Anti-Collision HUD Design Formulation 167

FIGURE 11.1  Cyber-physical networks in space.

computing, running applications remotely in the cloud, etc., founded upon satellite-
based futuristic models.
The work takes into consideration this model where a heavy server-mounted satel-
lite is theoretically hypothesized to interact with other satellites to gather and ana-
lyze their data and conduct overhead edge computing. This model guarantees the
Earth Observation (EO) satellite data to be distributed, stored, and processed into
instantly useable images and then sent directly to grass root users. Figure 11.1 depicts
such a cyber-physical network that comprises the integration of IoT-integrated satel-
lites, AI-enabled high-altitude platforms, terrestrial network communication, ground
communication, remote-sensing satellite networks, communication, and astronomi-
cal satellite networks.

11.4  SPACECRAFT ANTI-COLLISION SYSTEM DESIGN

A collision avoidance system is a safety technology that is meant to avoid or mitigate
the severity of an accident during the occurrence of the accidental circumstances.
The spacecraft anti-collision system (SACS) is based on the concept of establish-
ing data centers in space, as shown in Figure 11.1. Anti-collision system design
constitutes five design subsystems, viz., traffic identification design (TID), density
evaluation model (DEM), trajectory planning model (TPM), risk assessment model
(RAM), and finally heads-up display (HUD) system.
168 Intelligent Analytics for Industry 4.0 Applications

11.4.1 Traffic Identification Design

The traffic identification design uses a video imaging process based on an IoT-
integrated video camera docked with a mathematical computing system. The video
camera sends raw image data feed as a first source impression for the system to
further handle the computational processes. The system works by recording image
frames and sending them to the computer unit. The video feed gets processed via a
series of computers and the resultant information of the visual silhouettes detected
in the form of traffic elements is transmitted over the controller area network (CAN)
bus that also gets displayed on the HUD system. The system’s performance is pre-
dicted to be high, and it is dependent on the image analysis algorithm, which should
accordingly match the onboard hardware implementation. The system must be capa-
ble of distinguishing between different sorts of barriers as well as different types of
terrain, viz., corrugated (as in the case of asteroids) and smooth (as in the case of
rockets, missiles, and spacecrafts).
As shown in Figure 11.2, the original visual B is captured by the camera and then
subsequently scanned resulting in output A. The scanned data are communicated to
the IoT-integrated satellite networks, together establishing data center C, which can
later be retrieved in case of any incident sparks.
The intensity reduction feature extraction technique is used to divide a big col-
lection of raw data into smaller, more tractable data sets for subsequent processing.
These huge data sets include a great number of variables, requiring a significant

FIGURE 11.2  An illustration depicting traffic identification design where the image is first
scanned, and the visual elements are recognized postcomputing.
IoT-Based Spacecraft Anti-Collision HUD Design Formulation 169

FIGURE 11.3  Progression at various layers in computing processes over a time period (t).

amount of processing resources to compute. When there is a huge data collection and
it is needed to decrease the amount of physiognomy without accommodating any loss
of critical or relevant information, extracting the features is a handy strategy. Most
of the information in the original set of features is summarized through a new and
smaller collection of features, which are further classified.
Figure 11.3 shows the progression at various layers in the computing process,
which initiates from the point of image scanning and results in the identification of
the image. Meanwhile, the features are extracted and the data obtained are matched
with predefined elements and parameters.

11.4.2 Density Evaluation Model

A three-dimensional dilated spatial frame consisting of n traffic elements is rendered.
The frame that encapsulates the elements has magnitude a as three of its dimen-
sions. As the density is dynamic in nature, it changes over time t. The traffic den-
sity is a function of time, magnitude of three-dimensional spatial cube, and n­ umber
170 Intelligent Analytics for Industry 4.0 Applications

FIGURE 11.4  Plot for dynamic density; the region zoomed in the figure depicts three-time
frames with reduced traffic elements at each point.

of ­traffic elements. This dynamic traffic density dT is mathematically expressed as

follows:

dT = f ( t ,  a,  n ) (11.1)

The formula for evaluating dynamic traffic density dT is as follows:

n   +   ∆n
dT = (11.2)
a3
where Δn is the change in traffic elements observed over a period of time t.
The dynamic density dT is visualized, to the pilot, in HUD screen in the form of a
continuously varying graphical plot whose Y-axis denotes the density variations and
X-axis denotes the time. This graphical plot is depicted in Figure 11.4.

11.4.3 Risk Assessment Model

The loss of human life or injuries, environmental damage, and economic losses are
the primary hazards associated with spacecraft accidents. The consideration of risk
studies is valuable in the preparation of strategic tools for developing and deploying
emergency control and management scenarios in space.
In this model, the maximum collision probability risk P(C) is given as follows:

na
P(C ) = (11.3)
nt
where na implies the number of traffic elements tested for collision and nt indicates
the total number of traffic elements present in the rendered frame.
IoT-Based Spacecraft Anti-Collision HUD Design Formulation 171

All na elements account for maximum collision risk since they confirm their
p­ resence in the congested zone. The total array of elements present in the congested
zone is given as follows:

L = {e1 , e2 , e3 , e4 , e5 ,…, en } (11.4)

while the total of all the elements present in the frame is indicated as follows:

M = {e1 , e2 , e3 , e4 , e5 , e5+ k ,e5+ 2 k ,…, e5+ mk } (11.5)

Subtracting na from nt results in ne depicts the number of traffic elements that

remained absent from being present in the congested zone.
Further to this, combination-based risk assessment probability can also be com-
puted for precise risk possibilities. The combinations could be generated using the
following formula:

n!
Cr = (11.6)
r ! ( n − r )! 
 n

where n = na and r < na

11.4.4 Safe Trajectory Planning Model

As shown in Figure 11.5, dt and Dt are the diagonals emanating from N0 and touching
the vertices of the overview projection representing the congested zone.

FIGURE 11.5  Depiction of traffic elements na and ne in the rendered frame.

172 Intelligent Analytics for Industry 4.0 Applications

The empty zone is located by identifying the free-from-traffic (FFT) vertices at dt

and Dt. The empty zone ze guarantees the zero-traffic movement for time period T0
sufficient enough to maneuver the spacecraft in the safe flight zone zs.
The distension array of empty zone is represented as follows:

Z = { x , y, z } (11.7)

As shown in Figure 11.6, four FFT vertices are identified, two along dt and two
along Dt. An imaginary quadrilateral simulation is then established by extending
lines through vertices and aligning the rest dimensions. Post the establishment of the
empty zone ze, safe zone zs is computed using the following formula:

zs = ( ze ) (11.8)

Zs and Ze denote safe and empty zone, respectively.

where f ( ze ) = k ze (11.9).

Here, k denotes the diminishing factor.

Finally, the distension array of safe zone is obtained as follows:

S = { x n , yn , zn } (11.10)

The safe trajectory ts is confined in the safe flight zone zs. The safe trajectory is a
function of time period T0+J where T0+J signifies the time period that guarantees that

FIGURE 11.6  Estimation of empty zone in the rendered frame.

IoT-Based Spacecraft Anti-Collision HUD Design Formulation 173

FIGURE 11.7  Heads-up display design within the spacecraft’s windscreen.

the spacecraft would maneuver uninterruptedly through the safe flight zone path
until the clearance of all the associated risks.

11.4.5 Heads-Up Display System

HUD system highlights the transparent display feature mechanism that is visually
layered within the windscreen of the spacecraft and its function is to display critical
information and data directly in the vision feed of pilot without distracting the pilot
from the usual vision angle onto the screen. As shown in Figure 11.7, HUD is used
to display a warning message in the red highlighted screen zone and the computer
suggestions as a precautionary move in the green highlighted zone. The HUD is
also shown displaying the dynamic density plot and the distance between the closest
approaching element and the spacecraft being piloted.
In this model, the HUD system receives direct feed from RAM, post which a
precaution message pops up providing recommendation strategies for the pilot to
choose. The ultimate decision rests with the pilot. This has been formulated deliber-
ately to reduce complete machine interventions so that the pilot gets complete free-
dom to operate.

11.5  CONCLUSION AND FUTURE SCOPE

An IoT-based spacecraft anti-collision HUD design formulation is incepted for its
successful implementation in a real-space environment to help mitigate threats aris-
ing out of space traffic accidents. The design constituted five subsystems for the
174 Intelligent Analytics for Industry 4.0 Applications

effective management of space traffic. The model focuses on safe trajectory planning
by computing a safe zone from the previously constructed empty zone present inside
the three-dimensionally distended frame. The model also presented a risk assess-
ment technique for maximum risk evaluation and for computing risk probability
based on spacecraft combinations. The model proves trustworthy for assessing risks
and planning trajectories with multiple possibilities of integration with IoT-based
equipment and systems without engineering conditionalities. Work is underway to
upgrade the performance scenario of the anti-collision system design and to also
incorporate additional mathematical models under a laboratory test environment.
The model incepted in this research could be used for modeling and simulations
of space traffic management in real-world scenarios. The research delves into the
AI-rested mathematical models for assessing the dangers and risks of space traf-
fic crashes and extends the research into the field of aerospace engineering. The
research would serve as a foundation for future research on preventing and con-
trolling space traffic collisions and accidents. This research also provides new and
futuristic insights into space traffic management highlighting the all-inclusive model.

REFERENCES
1. Jones, H. (2018). The recent large reduction in space launch cost. In 48th International
Conference on Environmental Systems, pp. 1–10, 8–12 July 2018, Albuquerque, New
Mexico.
2. Patera, R. P. (2001). General method for calculating satellite collision probability.
Journal of Guidance, Control, and Dynamics, 24(4), 716–722.
3. Muelhaupt, T. J., Sorge, M. E., Morin, J., & Wilson, R. S. (2019). Space traffic manage-
ment in the new space era. Journal of Space Safety Engineering, 6(2), 80–87.
4. Bonnal, C., Francillout, L., Moury, M., Aniakou, U., Dolado Perez, J.-C., Mariez, J., &
Michel, S. (2020). CNES technical considerations on space traffic management. Acta
Astronautica, 167, 296–301.
5. Johnson, N. L. (2004). Airspace complexity and its application in air traffic manage-
ment. Space traffic management concepts and practices. Acta Astronautica, 55(3–9),
803–809.
6. Ailor, W. H. (2006). Space traffic management: Implementations and implications. Acta
Astronautica, 58(5), 279–286.
7. Cukurtepe, H., & Akgun, I. (2009). Towards space traffic management system. Acta
Astronautica, 65(5–6), 870–878.
8. Min, H., Guoqiang, Z., & Junling, S. (2010). Collision avoidance control for formation
flying satellites. In AIAA Guidance, Navigation, and Control Conference, 8–11 August
2011, Portland, Oregon.
9. Omar, S. R., & Bevilacqua, R. (2019). Spacecraft collision avoidance using aerody-
namic drag. Journal of Guidance, Control, and Dynamics, 43, 1–7.
10. Lee, K., Park, C., & Park, S.-Y. (2014). Near-optimal guidance and control for spacecraft
collision avoidance maneuvers. In AIAA/AAS Astrodynamics Specialist Conference,
4–7 August 2014, San Diego, CA.
11. Wang, S., & Schaub, H. (2008). Spacecraft collision avoidance using coulomb forces
with separation distance and rate feedback. Journal of Guidance, Control, and
Dynamics, 31(3), 740–750.
12. Hobbs, K. L., & Feron, E. M. (2020). A taxonomy for aerospace collision avoidance
with implications for automation in space traffic management. In AIAA Scitech 2020
Forum. https://doi.org/10.2514/6.2020-0877
 12 Coverage of LoRaWAN
in Vijayawada
A Practical Approach
P. Saleem Akram and M. Lakshmana
Koneru Lakshmaiah Education Foundation

CONTENTS
12.1 Introduction................................................................................................... 175
12.1.1  Hardware Setup................................................................................. 176
12.1.1.1  LoRaWAN Architecture...................................................... 177
12.2 LoRaWAN Coverage in Real-Time Test Scenarios....................................... 177
12.2.1  Test Scenario 1: Setting of LoRaWAN in the Center of
Vijayawada (India) City..................................................................... 177
12.2.1.1  Assessment of Findings, Practical Observations, and
Suggested Optimization Approaches.................................. 179
12.2.2  Scenario 2: Setting of LoRaWAN in the Suburban Region............... 179
12.2.2.1  Test Results – Observations................................................ 180
12.2.3  Scenario 3: In a Basement, LoRaWAN Is Used to Read the
Water Meter, Gas Meter, and Electricity Meter................................. 180
12.2.3.1  Comprehensive Model 1: Indoor Coverage of
LoRaWAN........................................................................... 180
12.2.3.2  Building Losses via the Exterior Walls............................... 182
12.2.3.3  Losses Via Constructing Interior Walls............................... 182
12.2.3.4  Power Boost Received........................................................ 182
12.3 Conclusions.................................................................................................... 183
References............................................................................................................... 183
Notes���������������������������������������������������������������������������������������������������������������������� 184

12.1 INTRODUCTION
LoRaWAN is a low-power, broad-area network standard (LPWAN). Concerning the
OSI levels, level 2 will be preferred as the MAC (media access control) layer. In other
words, LoRaWAN is responsible for combining the channels and connection settings
of many LORA devices with respect to channel, bandwidth, encryption of data, etc.
(Bourennane et al., 2020). This technology enables point-to-point information to be
sent and received. A LoRa device is characterized by a long range with a minimal
component. It employs the spread spectrum approach, which enables many signals to
be received at the same time while at the same time having varying speeds.

DOI: 10.1201/9781003321149-12 175

176 Intelligent Analytics for Industry 4.0 Applications

It delivers an appealing mix of properties such as ultra-long range, low-power

consumption, and minimum costs, as well as secured data transfer. LoRa can give far
more coverage than cellular networks, both for public and private networks.
Therefore, for its implementation in the Industrial Internet of Things (IIoT), the
long-range and low-power character of LoRa is essential to enhance our standards
of life swiftly. It answers many new issues, including climate change, pollution, and
natural catastrophes, both in rural and urban locations worldwide (Sinha et al., 2017).
LoRa is extensively used for transport, production, facilities, home equipment, and
even wearable products due to its comfort and lucrative attribute, like LPWAN.
IoT generally offers several technology options for data transfer across the net-
work, such as LoRa, NB-IoT, WiFi, Bluetooth, Zigbee, and Sub-1 GHz. However,
extended transmission lengths and energy conservation in wireless network systems
cannot be guaranteed at the same time because more energy is always required for a
greater transmission distance. Therefore, such LoRa is the ideal solution for sending
a comparably small quantity of data over a long distance so that both long range and
low power are available (Butun et al., 2018).

12.1.1 Hardware Setup
Here, we took the help of hardware setup of SX1272 chipset, which has been
­developed by Semtech company through which SX1272 module was developed. Its
transmission power is 14 dBm and sensitivity −134 dBm.1 In our practical approach,
we have taken help of the SX1272 LoRa module with 4.5 dBi Antenna.2 This module
relates to Socket 0, which is placed in SX1272 Waspmote board.
This chart, which was in Table 12.1 (see note 1), illustrates the comparison
between LoRa and many other commonly used wireless network technologies to

TABLE 12.1
LoRa Comparison to Other Wireless Technology
Wireless
Network Energy Communication Construction
Technology Velocity Distance Consumption Application Cost Cost
LoRa Slowest Longest Low Outdoor Free Median
sensors
NB-IoT Median Short High Indoor Data flow charge Low
sensors
3G/4G Fastest Long High Call and Data flow charge High
internet
WiFi Fast Short Highest Home Free Low
network
Bluetooth Median Shortest Low Phone Free Low
accessories
ZigBee Slow Shorter Low Indoor Free Low
equipment
Coverage of LoRaWAN in Vijayawada 177

build a straight and basic overview of the LoRa. It has a key role in establishing and
maintaining costs. In comparison to other comparable technologies, the cost of com-
munication is very negligible. Therefore, the cost from LoRa nodes to the gateway
and the nube may be called free of charge.

12.1.1.1  LoRaWAN Architecture

LoRa is a methodology of spectrum spread modeling (CSS) that derives from the
technology of chirp spread spectrum. LoRaWAN is the architecture of the LoRa
Alliance’s network system and communication protocol standard (Saari et al., 2018).
LoRaWAN is a layered protocol for MAC, which allows LoRa to operate in larger
applications. The LoRa Alliance specifies the network layer and network system
architecture (see note 1). The various components connect end nodes to the cloud
and the application servers in a LoRaWAN communication flow (Mekki et al., 2019).

12.2  LORAWAN COVERAGE IN REAL-TIME TEST SCENARIOS

The unique qualities of LoRa, particularly its low sensitivity to interference, minimal
path loss, and excellent building penetration, allow for extremely long ranges even in
densely populated areas. Even with modern software-assisted modeling approaches,
however, the total range and coverage of a radio network can only be anticipated to a
limited extent – and this is true for any radio technology, not only LoRa (Kanakaraja
and Kotamraju, 2021). As a result, range and penetration testing are critical for better
understanding radio wave behavior in real-world environments, particularly in the
first phase.
Three documented tests in various settings are discussed in this chapter, as
follows:

• Five distinct test situations in an urban environment.

• LoRaWAN in the suburban areas.
• Signal strength calculations inside basement rooms.

Only a few experimental range tests are adequately recorded in the literature in
general.

12.2.1 Test Scenario 1: Setting of LoRaWAN in the

Center of Vijayawada (India) City
Many of the cities have similar kind of architectural characteristics like Vijayawada
(India) city in Andhra Pradesh, which we selected for our real-time testing for
LoRaWAN coverage. There are just a few very tall structures, towering structures,
and many that are only a few storeys high. These are not megacities with virtually
entirely high-rise structures in the city core (Akram et al., 2021).
We could get coverage of well over 10 km with exposed gateway sites in over
20 range tests in metropolitan situations. Individual measures will be covered in
the ­sections that follow. Description of the real-time test setup (Ashok and Saleem
Akram, 2020):
178 Intelligent Analytics for Industry 4.0 Applications

• These were non-line-of-sight (NLOS) connections in metropolitan

environments.
• 12th spreading factor (maximum range).
• Maximum output power: 14 dBm.
• Frequency of the channel: 868 MHz band.
• The receiving gateway Rx was placed on the roof at the height of about
12 m, and the transmitter unit (Tx) was moved between five control stations.
• About 50 transmission attempts were made at each site.

In the following list, we have explain the features of the control points and the
­measurement series results:
Test Point 1: the signal passes across four buildings. A line of sight is not attain-
able along this path due to three high-rise structures and one low-rise structure.
However, a direct channel to the receiver is also accessible. The distance between
the two sites is 830 m. Ninety-six percent of the packets transmitted were received
(Gopal and Akram, 2019).
Test Point 2: the signal passes across 14 buildings. Test Point 2 in this scenario is
home to a massive concentration of low-rise residential complexes. There is also a
block of apartments on the path to the receiver. The range is 960 m. Ninety-two per-
cent of the data packets sent were received (Komalapati and Yarra, 2021).
Test Point 3: the signal passes across six buildings. This point is in the neighbor-
hood on the side of a large plaza. The path passes through a large residential com-
plex, followed by various industrial structures. The maximum range is 1,070 m. Once
again, 98% of the packets sent were received.
Test Point 4: the signal passes through 14 buildings. It is the most significant way.
This spans the apartment’s four high-rise buildings. Then an open area with no obsta-
cles follows until a grouping of high-rise skyscrapers reaches again. Finally, several
industrial structures are located before the road to the recipient is built. The distance
between us is 1,530 m. Ninety-eight percent of the packets transmitted were received.
Test point 5: the signal passes across six buildings. This test spot is divided into
several industrial buildings that do not have open space between them. It is 863 m
high. Received 100% of the transmission of data packets.

TABLE 12.2
Comparison of Various Test Points and Their Success Rates
The Number of
Buildings between
Source to Distance to the
Test Point Destination Gateway (Rx) (m) Success Rate (%)
1 4 830 96
2 14 960 92
3 6 1,070 98
4 14 1,530 98
5 6 863 100
Coverage of LoRaWAN in Vijayawada 179

Table 12.2 provides the (see note 1) locations of the test sites for the results of the
successful tests in real time. All probable scenarios, such as overcrowded buildings
at different distances, are covered by test points. The word “success rate” refers to
the rate at which the data from a LoRa node to a gate and vice versa are successfully
transmitted/received.

12.2.1.1 Assessment of Findings, Practical Observations,

and Suggested Optimization Approaches
The LoRaWAN standard is the fact that not all packets are always received 100%,
which were sent: the communication can be set, if required, as the recipient’s mes-
sage (so-called confirmed connections up and down), e.g., the sender who waits
for a c­onfirmation reception and who returns the data should not be given such
an acknowledgment after a certain period. Therefore, all achieved levels may be
regarded as noncritical, which is explained as follows:

• Three basic solutions for reach and coverage optimization are available.
• In typical LoRaWAN applications, broad coverage and ranges may be
accomplished by allowing lower success rates, meaning accepting single
failed signals and then retransmitting the ranges in the test might have been
expanded.
• Redundant network topologies from several directions enable additional
network coverage while preserving or enhancing success rates.
• The lighting on the surface might be substantially improved with a higher
location of the doorway. Unfortunately, the door was only installed at 12 m
in this test.
• The Hata-propagation model (Hata-propagation model) is a basic radio
propagation model for calculating the loss of paths of radio waves outside.
The initial range roughly doubles when the gateway’s position is extended
from 12 to 30 m, say.

12.2.2 Scenario 2: Setting of LoRaWAN in the Suburban Region

The entrance was located on the second floor of a building, directly before the
­window.  At an external temperature of 15°C, the moisture content was 55% on
the test day. The distance between the door and the five separate test stations was
650–3,400 m.
The terminal system was in a vehicle during the testing. The terminal device’s
transmission power is set at 14 dBm by default. The packet confirmation and retrans-
mission were switched off to evaluate the performance of different spreading factors.
However, the connection test was not disabled so that the propagation factor does not
change even if a packet loss occurs. This is because the LoRaWAN network defaults
to the connection quality in the distribution factor.
For these experiments were used the spreading factors 7, 9, and 12. Around
100 packets were broadcast to the network server with a sequence number for each
test.
180 Intelligent Analytics for Industry 4.0 Applications

12.2.2.1  Test Results – Observations

The larger propagation factors have greater coverage than previously mentioned –
more than 80% of the packets received at an SF value of 12 at 2,800 m and no packet
at an SF of 7. At around 5 m above ground level, the gate was on the second storey.
The test point was 2,800 m away, just behind a seven-storey building. SF12 alone at
3,400 m provided covermount. The price of the high rate using the high dispersion
factor is a substantially lower bit rate. In contrast, network coverage with low disper-
sion factors is greatly reduced.
If the gateway is to be installed permanently, it should be placed in a more advan-
tageous area. Unfortunately, the location is a building on the second storey, which is
undoubtedly unsuitable for providing the widest coverage.

12.2.3 Scenario 3: In a Basement, LoRaWAN Is Used to Read

the Water Meter, Gas Meter, and Electricity Meter

A scenario including cellar coverage is also stated in the previously referenced

Libelium source. However, the basic criteria are only loosely described: In this mea-
surement, the transmitter was positioned at an altitude of roughly 3 m outside the first
floor of the office structure. The receiver was situated near the garage entrance below
the ground floor. The point of view was not visible, and the range of LoRaWAN was
206 m. There are many more which are not documented.
The findings we achieved with consumers in penetration tests are far more illumi-
nating. So we would like to give three more instances this time.
Data were provided to network a water meter during the first test of 998 m from a
road ship to an entrance at the height of 35 m. Thus, a connection may be established
despite the metal cover of the shaft. A delivery car was placed on the shaft to imitate
the conditions in the real world, but signals were still in good form.
Penetration tests were performed at another client over many days in five different
cellar rooms within a 4 km radius (Table 12.3).

12.2.3.1  Comprehensive Model 1: Indoor Coverage of LoRaWAN

Indoor and outdoor planning parameters are often taken into consideration while
developing wireless network coverage. Statistical propagation models have therefore
been built to assist the development of areas. For example, this is the Davide Magrin
model (see network-level performances of a LoRa system). In addition, LPWAN
communication networks, such as LoRaWAN, are essentially PMPs with sensors on
different floors such as cellars. These requirements encourage network planners to
build digital 3D maps and to employ deterministic distribution models.
Different approaches can be used for modeling losses generated from the outside
and inside walls of buildings. For example, in one of the models offered, the follow-
ing units evaluate the impacts of buildings on the route loss of LoRaWAN systems:

• Losses via building through exterior walls.

• Losses via building through inner walls.
• Higher energy use since one unit is above the first floor.
TABLE 12.3
Table of Success Rates Based on Distance and the Gateway to Which They Were Linked
Coverage of LoRaWAN in Vijayawada

1 2 3 4 5
Distance Success Distance Success Distance Success Distance Success Distance Success
(m) rate (%) (m) Rate (%) (m) Rate (%) (m) Rate (%) (m) rate (%)
Gateway 1: 5 m 3,500 0 2,600 0 1,500 0 5,200 0 1,200 0
Gateway 2: 10 m 4,100 0 3,500 0 3,700 2 2,700 0 4,200 26
Gateway 3: 45 m 700 81 1,700 5 2,000 87 2,500 78 3,800 56
181
182 Intelligent Analytics for Industry 4.0 Applications

The following formula may be used to determine route losses through buildings
using this model:

12.2.3.2  Building Losses via the Exterior Walls

The variability of materials and external wall thickness in many different buildings
should be considered because the loss via external walls of both types of equipment
is not usually the same. Table 12.4 illustrates three distinct ranges of values and the
probability of a node losing this kind of value.

12.2.3.3  Losses Via Constructing Interior Walls

The effect of the inner walls is the highest value between the two values. Therefore,
the number of internal barriers causing the loss is calculated as follows:

Gate 1 = Wi – p (12.1)

where Wi is distributed in the [3, 10] dB range evenly and p is the internal wall count
splitting the receiver from the transmitter.
It is assumed that 15% of the units are evenly distributed among the values of
p = {0, 1, 2} with a value of p = 3 for those units.
The second variable required to describe the loss of trajectory through inner bar-
riers is calculated as follows:

Tor3 = α d (12.2)

where

α = 0.5 dB/m – the penetration depth and the coefficient.

d = a value that is equally scattered across the range of [0, 14] m.

12.2.3.4  Power Boost Received

The power signal contribution finally describes how improved the antenna height
reception is achieved. The following expression defines this parameter:

Power signal = n * Gn (12.3)

TABLE 12.4
A Probable Energy Signal Value Range and Likelihood
Probability Ranger
0.24 [3, 10] dB
0.64 [10, 18] dB
0.1 [18, 22] dB
Coverage of LoRaWAN in Vijayawada 183

where

Gn = 1.5 dB – the increase owing to the height increase on one floor.

n = the number of buildings, divided equally throughout the values of n = {0–4}.

The three elements describe the overall indoor building loss are provided by the
­following equation:

L ( dB) = path loss + max [ Tor1,Tor ] − GFH (12.4)

In the virtual environment, certain tests with different scenarios suggest. However,
no penetration could be ensured theoretically, calculating the complete loss on the
direct route over multiple levels and walls. A gateway to the multistorey structure
on the same roof with the basement or cellar to be connected may also have a sense.
There is rarely any importance to the straight passage through the building and
reflection on nearby structures. The attenuation is inadequate to one to two walls,
whereas the pathway is longer, and the structure reflexes.

12.3 CONCLUSIONS
This article presented the real-time test scenarios of the LoRaWAN nodes ­covered
with different distances and various line-of-sight cases with practical demonstra-
tion and comparison of results in suburban and urban areas. Three types of cases
are ­ demonstrated as follows: (i) five various test cases – urban environment,
(ii) LoRaWAN – suburban area, and (iii) the penetration of the basement rooms for
reading data of the meter. Finally, we conclude that LoRaWAN nodes have a better
coverage capacity with fewer losses, and the same is proven with test results and sim-
ulation plots. This is a practical demonstration verified in Vijayawada city in India.

REFERENCES
Akram, P.S., Ramesha, M., Ramana, T.V, Simulation and analysis of smart home architecture,
ARPN Journal of Engineering and Applied Sciences, 2021, vol. 16, issue 6, pp. 280–286.
Ashok, G.L.P., Saleem Akram, P., Implementation of smart home by using packet tracer,
International Journal of Scientific and Technology Research, 2020, vol. 9, issue 2, pp.
678–685.
Bourennane, E.-B., Boucherkha, S., Chikhi, S., A study of LoRaWAN protocol performance
for IoT applications in smart agriculture, Computer Communications, 2020, vol. 164,
pp. 148–157.
Butun, I., Pereira, N., Gidlund, M., Analysis of LoRaWAN v1.1 Security: Research Paper. 1–6,
2018. doi: 10.1145/3213299.3213304
de Carvalho Silva, J., Rodrigues, J.J.P.C., LoRaWAN —A low power WAN protocol for
Internet of Things: A review and opportunities, 2017 2nd International Multidisciplinary
Conference on Computer and Energy Science (SpliTech), 2017, pp. 1–6.
Gopal, S.R., Akram, P.S., Design, and analysis of heterogeneous hybrid topology for VLAN
configuration, International Journal of Emerging Trends in Engineering Research, 2019,
vol. 7, issue 11, pp. 487–491.
184 Intelligent Analytics for Industry 4.0 Applications

Kanakaraja, P., Kotamraju, S.K., IoT enabled ble and lora based indoor localization w
­ ithout
GPS, Turkish Journal of Physiotherapy and Rehabilitation, 2021, vol. 32, issue 2,
pp.642–653.
Komalapati, N., Yarra, V.C., Smart Fire Detection and Surveillance System Using IOT,
Proceedings - International Conference on Artificial Intelligence and Smart Systems,
ICAIS 2021, 2021, pp. 1386–1390.
Mekki, K., Bajic, E., Chaxel, F., Meyer, F., A comparative study of LPWAN technologies
for large-scale IoT deployment, ICT Express, 2019, vol. 5, pp. 1–7. doi: 10.1016/j.
icte.2017.12.005
Butun, I., Nuno Pereira, Mikael Gidlund. “Analysis of LoRaWAN v1. 1 security.” In
Proceedings of the 4th ACM MobiHoc Workshop on Experiences with the Design and
Implementation of Smart Objects, pp. 1–6, 2018.
Sharan Sinha, R., Wei, Y., Hwang, S.-H., A survey on LPWA technology: LoRa and NB-IoT,
ICT Express, 2017, vol. 3, issue 1, pp. 14–21.

NOTES
1 LoRa end nodes are not linked to a particular gateway, unlike cellular technology.
Instead, data are delivered to numerous gateways through a node. Then each gateway
passes the packet to a cloud-based LoRa network server (LNS) from the end node (cel-
lular, ethernet, satellite, and WiFi). LNS is complicated and intelligent for network
managing, filtering redundant packages, doing security checks, and allowing adaptive
data rates (de Carvalho Silva and Rodrigues, 2017). Consequently, there is no transfer
for a mobile LoRaWAN device from one gateway to another.
2 Proportional length to the gateway and the corresponding percentage of success. There
have been thousands of tests of transmission spanning several hours or days at every
site. Even at enormous distances, subterranean rooms with their impediments (rein-
forced cement, 2nd cellar, steel cellar doors, etc.). In this scenario, the height of the door
is critical. In addition, this is based on situations requiring the retransmission of value if
the success rate does not always guarantee direct transfer.
 13 Intelligent Health
Care Industry for
Disease Detection
Ishaan Deep, Payaswini Singh, Aditi Trishna,
Sanskar, Biswa Ranjan Senapati,
and Rakesh Ranjan Swain
Siksha ‘O’ Anusandhan University

CONTENTS
13.1 Introduction................................................................................................... 185
13.2 Literature Study............................................................................................. 188
13.3 Proposed Work.............................................................................................. 188
13.4 Simulation Results......................................................................................... 189
13.5 Confusion Matrix.......................................................................................... 190
13.6 Conclusion and Future Scope........................................................................ 192
References............................................................................................................... 192

13.1 INTRODUCTION
The state of one’s health is crucial. “Health is Wealth,” as the saying goes, because a
healthy individual is eligible for a variety of benefits, including a decent job life and
personal food habits. However, an ill individual is bound by numerous limitations.
People’s lifestyles are rapidly changing nowadays. This has an impact on their health.
Life expectancy is dwindling, and health problems are on the rise.
Researchers are still looking into how to detect these diseases at an early age so
that they can be cured easily. One of the research methods is the machine learn-
ing (ML) approach [1]. ML is a type of artificial learning (AL) in which various
algorithms are used to determine the accuracy and precision values of an algorithm
in order to determine whether it is producing the desired output [2]. ML is widely
used in many fields to enhance performance. Various application domains in which
ML models are widely used are healthcare [3–5], vehicular networking [6–11], cyber
security [12], industry sector [13], etc. Figure 13.1 represents different types of ML.
This chapter focuses on the detection of diseases by means of ML algorithms, and
so the authors have used supervised ML (support vector machine (SVM) and linear
regression).
Supervised learning has one of the most popular algorithms known as SVM,
which is used for classification as well as regression problems. In the SVM algorithm,

DOI: 10.1201/9781003321149-13 185

186 Intelligent Analytics for Industry 4.0 Applications

FIGURE 13.1  Different types of machine learning.

FIGURE 13.2  Support vector machine algorithm.

the data records are plotted as a position in a multidimensional space where every
dimension has some features being the value of a particular coordinate. The SVM
kernel is a functional approach for processing and converting low-dimensional input
into high-dimensional space [14]. Figure 13.2 shows how SVM divides the dataset
into two groups.
Intelligent Health Care Industry for Disease Detection 187

FIGURE 13.3  Logistic regression algorithm.

One of the most often used ML methods is logistic regression (LR), which is used
to predict a categorical dependent variable using a set of independent variables. One
of the most common models is a binary outcome. It is a useful method for analyz-
ing problems and to fit the data into the best category [15]. The logistic function is
depicted in Figure 13.3.
In this chapter, the authors have used datasets of three common diseases that the
people are suffering from such as obesity, heart problem, and cancer. Nowadays,
most of the people are suffering from these diseases at an early age, and thus the life
span is decreasing. Using the ML algorithms, these diseases can be rectified at an
early age and people can take the required precautions to live a healthy and happy
life.
Death rates are also rising, with obesity, heart attacks, and cancer accounting for
the majority of deaths. Because these diseases are hazardous, they should be rec-
ognized at a young age so that necessary safeguards can be taken and a happy and
healthy life can be lived. The goal of this model is to figure out how to detect these
diseases at a young age.
The major contribution of this chapter is presented as follows:

a. To detect the diseases at an early age using SVM and LR.

b. To compare the accuracy obtained for SVM and LR for the proposed work.
c. The performance is measured in terms of performance metrics like
­precision, recall, F1 score, and confusion matrix.

The following is the chapter’s structure. The second section covers the literature
study. The third section examines the proposed work. The fourth section presents
the  simulation results. The conclusion and future scope are discussed in the fifth
section.
188 Intelligent Analytics for Industry 4.0 Applications

13.2  LITERATURE STUDY

When it comes to health informatics, ML has a very broad application. Its goal is to
create algorithms that can make precise predictions. Primary health care is a vast
field for ML in which to experiment with new domains and techniques. It offers
a variety of tools to improve patient safety and risk management techniques. ML
technology can aid in data analysis and risk identification, thereby improving diag-
nosis and treatment. They are useful for tracing complex patterns in large datas-
ets. Cardiovascular disease refers to conditions affecting the heart or blood vessels.
Many studies have been conducted to predict the prevalence of CVD using various
factors such as age, family history, hypertension, cholesterol, and smoking. Usually,
it is found that methods such as Naïve Bayesian, neural network, and decision tree
algorithms were used. This chapter employs supervised learning models such as LR,
decision tree, k-nearest neighbors, and random forest. Sex, type of chest pain, and
number of major vessels are among the variables considered. The disease was cor-
rectly predicted 90.16% of the time [16].
Obesity is measured using BMI (or body mass index). The four ML techniques
used in this chapter are binary LR, improved decision tree, weighted k-nearest neigh-
bor, and artificial neural network. This study includes data from Tennessee high
school students. Energy intake, physical activity, and sedentary behavior were among
the variables studied. It is extremely useful because it defines the application of these
ML techniques in great detail. It predicts obesity using two parameters: accuracy and
specificity. The KNN model had the highest accuracy (88.82%), and the ANN model
had the highest specificity (i.e., 99.46%) [17].
This chapter is based on the prediction of obesity using genetic variant analytics,
where SNPs were used as variables in the prediction of obesity. Following feature
selection, only 13 important variables were chosen. This chapter employs techniques
such as k-nearest neighbor, SVM, and random forest. This chapter was extremely
useful because it employed similar ML techniques. The authors made a significant
observation: when the number of variables increases significantly, the performance
of these ML models decreases. The chapter concludes that the SVM has the highest
accuracy, 90.5% [18].
Cancer is one of the most dangerous diseases, and predicting it in its early stages
is critical. Various methods have been used to detect this disease so that proper treat-
ment can be provided before a patient’s condition becomes too severe. There are
numerous datasets available in the research community. As a result, accurately pre-
dicting the disease is both interesting and difficult. The prediction in this study is
based on microarray datasets. It contains datasets on colon, leukemia, and lymphoma
cancer. This study suggests that information gain and Pearson’s correlation coeffi-
cient were the best feature selection methods for these three datasets, while MLP and
KNN were the best classifiers [19].

13.3  PROPOSED WORK

In this model, the authors used three datasets—the obesity dataset [20], the heart
attack dataset [21], and the cancer dataset [22]. The datasets were trained and tested
Intelligent Health Care Industry for Disease Detection 189

FIGURE 13.4  Working principle of machine learning model.

TABLE 13.1
Details of Baseline Methods of Machine Learning
State Abbreviation
Artificial Learning Al
Machine Learning Ml
Support Vector Machine SVM
Logistic Regression LR
Cardiovascular Disease CVD
Body Mass Index BMI
K-Nearest Neighbor KNN
Artificial Neural Network ANN
Single Nucleotide Polymorphism SNPs
Multilayer Perceptron MLP

properly by using LR and SVM algorithms, where 90% of the dataset has been
trained and 10% of the dataset has been tested to find its accuracy. The recall value,
precision value, and the F1 score were also calculated. For three datasets, the confu-
sion matrix has also been calculated using LR and SVM algorithms.
Figure 13.4 shows the working information of ML algorithms. The considered
algorithm use for analysis is tabulated in Table 13.1
The abbreviation used here are given below:

13.4  SIMULATION RESULTS

The simulation results for the three datasets—obesity, heart attack, and cancer
­dataset—to find the accuracy, recall, precision, F1 score, and confusion matrix using
SVM and LR are provided below, where C and Gamma are hypermeters. The train-
ing model is set using C and Gamma, where C regulates the error and Gamma deter-
mines the curvature weight of the decision border.
Performance metrics is based on four performance values, as follows:

Accuracy: it is the proportion between the number of accurate forecasts and

a.
all predictions combined [23].

Number of correct predictions

Accuracy =
Total number of predictions
190 Intelligent Analytics for Industry 4.0 Applications

TABLE 13.2
Performance Analysis of SVM with Different Values of the Hyperparameters
SVM LR
C Gamma Accuracy Accuracy
Obesity 2 0.001 0.92 0.82
Heart attack 2 0.001 0.82 0.82
Cancer 2 0.0001 0.96 1

TABLE 13.3
The Comparative Analysis of Considered Performance Parameters
SVM/LR
Recall Precision F1 Score
Obesity 0.9218/0.8224 0.9262/0.8252 0.9242/0.8228
Heart attack 0.8285/0.8280 0.8473/0.8218 0.8239/0.8244
Cancer 0.9588/1 0.9621/1 0.9597/1

Precision: it is calculated by dividing the total number of accurate forecasts

b.
by the number of true positives [24].

True positive
Precision =
True positive + False Negative

Recall: it is the proportion of true positives to the sum of true positives and
c.
false negatives [24].

True positive
Precision =
True positive + False negative

F1 Score: it refers to how many correct hits were also discovered, or how
d.
many true positives were recalled [24].

2*Recall*Precision
F1Score =
Recall*Precision

13.5  CONFUSION MATRIX

To analyse the performance, the confusion matrix for Obesity using SVM and Logistic
Regression has been drawn in Figure 13.5 and Confusion matrix for Heart Attack
using SVM and Logistic Regression has been created in Figure 13.6, respectively.
Intelligent Health Care Industry for Disease Detection 191

FIGURE 13.5  Confusion matrix for obesity using SVM and logistic regression.

FIGURE 13.6  Confusion matrix for heart attack using SVM and logistic regression.

FIGURE 13.7  Confusion matrix for cancer using SVM and logistic regression.
192 Intelligent Analytics for Industry 4.0 Applications

FIGURE 13.8  Accuracy graph for SVM and logistic regression.

Similar information has also been tabulated in Table 13.2. Further, the performance
metrics returned by the SVM have been tabulated in Table 13.2. The same is also
presented in Figure 13.7. The accuracy results of the three datasets using SVM and
LR algorithms are shown in Figure 13.8. The same is also tabulated in Table 13.3.

13.6  CONCLUSION AND FUTURE SCOPE

The Intelligent Health Care Industry Model for Disease Detection was created with
the goal of detecting diseases at an early stage and taking the necessary safeguards.
The authors have taken into account three important diseases: obesity, heart attack,
and cancer. The accuracy, precision, recall score, F1 score, and confusion matrix
are determined using LR and SVM. The authors will aim to incorporate other ML
techniques into the proposed work in the future.

REFERENCES
1. M. I. Jordan and T. M. Mitchell, “Machine learning: Trends, perspectives, and pros-
pects,” Science, vol. 349, no. 6245, pp. 255–260, 2015.
2. P. Prajapati and A. Thakkar, “Extreme multi-label learning: A large scale classification
approach in machine learning,” Journal of Information and Optimization Sciences, vol.
40, no. 4, pp. 983–1001, 2019.
Intelligent Health Care Industry for Disease Detection 193

3. A. Kumar, R. Kumar, and S. S. Sodhi, “Intelligent privacy preservation electronic health

record framework using soft computing,” Journal of Information and Optimization
Sciences, vol. 41, no. 7, pp. 1615–1632, 2020.
4. A. Vellido, “The importance of interpretability and visualization in machine learning
for applications in medicine and health care,” Neural Computing and Applications, vol.
32, no. 24, pp. 18069–18083, 2020.
5. J. Wiens and E. S. Shenoy, “Machine learning for healthcare: On the verge of a
major  shift in healthcare epidemiology,” Clinical Infectious Diseases, vol. 66, no. 1,
pp. 149–153, 2018.
6. B. R. Senapati, P. M. Khilar, and R. R. Swain, “Fire controlling under uncertainty in urban
region using smart vehicular ad hoc network,” Wireless Personal Communications, vol.
116, no. 3, pp. 2049–2069, 2021.
7. B. R. Senapati, P. M. Khilar, and R. R. Swain, “Composite fault diagnosis methodology for
urban vehicular ad hoc network,” Vehicular Communications, vol. 29, p. 100337, 2021.
8. B. R. Senapati, P. M. Khilar, and R. R. Swain, “Environmental monitoring through
vehicular ad hoc network: A productive application for smart cities,” International
Journal of Communication Systems, vol. 34, no. 18, p. e4988, 2021.
9. B. R. Senapati and P. M. Khilar, “Optimization of performance parameter for vehicular
ad-hoc network (vanet) using swarm intelligence,” Nature Inspired Computing for Data
Science, 2020, pp. 83–107.
10. B. R. Senapati, P. M. Khilar, T. Dash, and R. R. Swain, “Automatic parking service
through VNET: A convenience application,” in Progress in Computing, Analytics and
Networking: Proceedings of ICCAN, 2019 (pp. 151–159), Springer Singapore, 2020.
11. B. R. Senapati and P. M. Khilar, “Vehicular network based emergency data transmis-
sion and classification for health care system using support vector machine”, 2022.
12. C. Virmani, T. Choudhary, A. Pillai, and M. Rani, “Applications of machine learning
in cyber security,” in P. Ganapathi and D. Shanmugapriya (Eds), Handbook of Research
on Machine and Deep Learning Applications for Cyber Security. IGI Global, 2020,
pp. 83–103.
13. D. P. Penumuru, S. Muthuswamy, and P. Karumbu, “Identification and classification of
materials using machine vision and machine learning in the context of industry 4.0,”
Journal of Intelligent Manufacturing, vol. 31, no. 5, pp. 1229–1241, 2020.
14. S. Suthaharan, “Support vector machine,” in S. Suthaharan (Ed.) Machine Learning
Models and Algorithms for Big Data Classification: Thinking with Examples for
Effective Learning. Springer, 2016, pp. 207–235.
15. D. W. Hosmer Jr, S. Lemeshow, and R. X. Sturdivant, Applied Logistic Regression. John
Wiley & Sons, 2013, vol. 398.
16. M. Nabeel, M. J. Awan, M. Raza, H. Muslih-Ud-Din, and S. Majeed (2021, November).
Heart attack disease data analytics and machine learning. In 2021 International
Conference on Innovative Computing (ICIC) (pp. 1–6). IEEE.
17. Z. Zheng, and K. Ruggiero (2017, November). Using machine learning to predict obe-
sity in high school students. In 2017 IEEE International Conference on Bioinformatics
and Biomedicine (BIBM) (pp. 2132–2138). IEEE.
18. C. A. C. Montañez, P. Fergus, A. Hussain, D. Al-Jumeily, B. Abdulaimma, J. Hind, and
N. Radi (2017, May). Machine learning approaches for the prediction of obesity using
publicly available genetic profiles. In 2017 International Joint Conference on Neural
Networks (IJCNN) (pp. 2743–2750). IEEE.
19. S. B. Cho, and H. H. Won (2003, January). Machine learning in DNA microarray analy-
sis for cancer classification. In Proceedings of the First Asia-Pacific Bioinformatics
Conference on Bioinformatics 2003-Volume 19 (pp. 189–198).
20. https://www.kaggle.com/datasets/ankurbajaj9/obesity-levels?select=ObesityDataSet_
raw_and_data_sinthetic.csv
194 Intelligent Analytics for Industry 4.0 Applications

21. https://www.kaggle.com/datasets/johnsmith88/heart-disease-dataset
22. https://www.kaggle.com/datasets/rishidamarla/cancer-patients-data
23. https://developers.google.com /machine-learning/crash-course/classification/
accuracy#:~:text=Accuracy%20is%20one%20metric%20for, predictions%20Total%20
number%20of%20predictions
24. https://towardsdatascience.com/essential-things-you-need-to-know-about-f1-score-
dbd973bf1a3#:~:text=1.-, Introduction, competing%20metrics%20%E2%80%94%20
precision%20and%20recall
 14 Challenges with
Industry 4.0 Security
Mohit Sajwan and Simranjit Singh
Bennett University Greater Noida

CONTENTS
14.1 Introduction................................................................................................... 195
14.2 Security Threats and Vulnerabilities............................................................. 197
14.2.1 Cyber-Physical Systems..................................................................... 197
14.2.1.1 Cyber-Physical System Security Threats............................ 197
14.2.1.2 Cyber Threats..................................................................... 197
14.2.1.3 Physical Threats.................................................................. 198
14.2.1.4 Vulnerabilities..................................................................... 199
14.2.2 Internet of Things.............................................................................. 199
14.2.2.1 Threats................................................................................ 199
14.2.2.2 Vulnerabilities.....................................................................200
14.2.3 Cloud Computing............................................................................... 201
14.2.3.1 Threats................................................................................ 201
14.2.3.2 Vulnerabilities.....................................................................202
14.3 Security Challenges of the Industry 4.0........................................................ 203
14.3.1 Inadequate IT/OT Security Expertise, Knowledge, or
Understanding����������������������������������������������������������������������������������� 203
14.3.2 Inadequate Security Policies and Funds............................................ 203
14.3.3 The Legal Obligation Regarding Products........................................204
14.3.4 Uneven and Insufficient Standardization...........................................204
14.3.5 The Devices’ Capabilities and Limitations from a Technical
Perspective���������������������������������������������������������������������������������������� 204
14.4 Conclusion..................................................................................................... 205
References...............................................................................................................205

14.1 INTRODUCTION
The most significant turning points in human history have been the industrial revolu-
tions. Many researchers believe the industrial revolution has greatly impacted peo-
ple’s daily lives more than the scientific revolutions. A substantial shift has occurred
in the sector over the past few decades as new developments have been integrated
into the system. Industry 4.0 (also known as the “fourth industrial revolution”) is
a significant outcome of the digitalization of a sector characterized by inflexible
standards and outdated fashions [1]. Manufacturing facilities have made significant

DOI: 10.1201/9781003321149-14 195

196 Intelligent Analytics for Industry 4.0 Applications

investments in developing intelligent production systems, which is a vital step toward

Industry 4.0. Industries are now looking to integrate technology and techniques such
as IoT, cloud solutions, and machine learning (ML) in order to improve productivity
and reduce costs [2].
Throughout history, revolutions have developed rapidly in response to the chang-
ing requirements of each generation. Electronic automation in industrial environ-
ments began in the 1970s, and the Internet of Things (IoT) has recently evolved
with data-driven production systems based on the cyber-physical system (CPS)
[3]. An attempt to make German industry more competitive by modernizing it was
the driving force behind the coining of the term “Industry 4.0.” The word “fourth
industrial revolution” began to be used interchangeably with this phenomenon as it
spread over the world. This industry’s rapid growth means that businesses will need
to implement a new process management strategy. The procedure itself needs to be
carried out as the primary objective. So, networked machines can change their pro-
duction processes on their own based on criteria that have been set up in advance and
without making any mistakes [4].
Security and privacy concerns are a result of the fourth industrial revolution and
some of its major characteristics. Our current manufacturing and production indus-
try is undergoing a digital change known as “Industry 4.0,” as previously stated.
Artificial intelligence, CPSs, cloud computing, and the IoT are just a few of the vital
factors of Industry 4.0. Because factory systems are designed to run on all 7 days of
the week, there can be no extended outages due to an attack. Intelligent manufactur-
ing firms are beginning to see the value in knowing and defending their systems, as
well as interpreting the data they generate [5].
In the context of Industry 4.0, it is absolutely necessary to devise an innovative
cyber security strategy and approach. The security operations center (SOC) is in
charge of combating these dangers, and it does so by constantly monitoring and con-
trolling the equipment under its control. Threat detection and response, along with
the speedy discovery of security concerns, are the primary functions of a SOC. We
can lessen the damage of an attack if we can find threats early and let local teams
know so they can fix security problems quickly without stopping production [6]. If
IoT grows quickly by 2022, attackers will target it more [7].
Many corporations and organizations predict an exponential rise in Internet-
connected devices in the future [8]. Gartner estimates 50 billion IoT connections
for Cisco by 2020, 20.8 billion for Gartner, and 100 billion for Huawei by 2025.
The most important lesson is to expect strong growth despite wide estimates. The
number of Internet-connected devices will expand rapidly, requiring a comprehen-
sive ­security solution [9]. Figure 14.1 shows Internet-connected device adoption from
2012 to 2025.
In recent years, cyber attacks have become more widespread, with victims rang-
ing from people to entire countries. The year 2014 was dubbed the Year of the
Breach, whereas 2015 was dubbed the Year of the Breach 2.0 by several industry
experts. It is clear that cyber attacks have wreaked havoc over the world from a global
­perspective. Customers who use the IoT systems should be taught how to protect
themselves against cyber attacks by their employers [10].
Challenges with Industry 4.0 Security 197

FIGURE 14.1  Connected IoT devices worldwide from 2012 to 2025.

Some thoughts on the problems of Industry 4.0, focusing on security, are ­presented
in this paper in order to increase awareness of the need for security in Industry 4.0.
The following is the outline for this document: After providing an introduction of
manufacturing industry procedures in Section 14.1, Section 14.2 provides an over-
view of the security and vulnerability concerns over industry 4.0, after which Section
14.3 provides security challenges of the industry 4.0, and Section 14.4 concludes with
conclusions and future developments.

14.2  SECURITY THREATS AND VULNERABILITIES

14.2.1  Cyber-Physical Systems
14.2.1.1  Cyber-Physical System Security Threats
When a cyber-physical threat is paired with one of the other risks that are listed
below, it is a CPS security threat.

14.2.1.2  Cyber Threats

To properly center information, you must secure the flow of data not only while it is
being stored but also while it is being sent and even while it is being processed [11].
Oriented Function: In order to do this, the cyber-physical components of the
whole CPS need to be integrated.
Oriented Threat: A threat that has an influence on the data’s availability, integ-
rity, confidentiality, and accountability.
198 Intelligent Analytics for Industry 4.0 Applications

Due to the aforementioned problems, CPSs are susceptible to

1. System wireless capabilities can be used to gain remote access or control

over a system or interrupt its operations if the system’s structure is known,
and hence wireless exploitation is necessary. Wired Exploitation Accidents
and/or loss of control are likely outcomes.
2. Jamming: the attacker’s purpose in this scenario is usually to modify the state
of the device and the planned actions to inflict harm [12]. Deauthentication
or wireless jamming signals are sent out in waves, and this prevents devices
and systems from receiving services. An illustration of this type of risk is
when intelligence services consistently carry out operations aimed at the
computational intelligence (CI) of a nation, mostly by means of the trans-
mission of malware. Therefore, the secrecy of the data is compromised as a
result of the limitations of traditional protections.
3. An attack on a network, either a logical or physical one, can result in unau-
thorized access, which can be used to get access to critical data and so
breach privacy. Physical or logical intrusions into the network can cause
this.
4. Interception: hackers can listen in on private conversations by exploiting
weaknesses that already exist or creating new ones, which results in yet
another breach of privacy and confidentiality [13].
5. Hackers can monitor a device or even a vehicle by taking advantage of
global positioning system (GPS) navigation systems, which violates users’
expectations of having their privacy regarding their location protected.
6. Information Gathering: software makers stealthily collect audit logs and
files that are stored on each device to get information they can sell for mar-
keting and business purposes. They do this so they can sell a lot of personal
information in an unethical way.

14.2.1.3  Physical Threats

Industrial neighborhood area networks (NANs), advanced metering infrastructure
(AMI), and data meter management systems are being introduced to industrial CPSs
to preserve their resilience. Physical risks can be categorized into three elements [12].
Damage from Physical Threats: power-generating stations (i.e., the base stations,
power plants, and power grid) are exceptionally well guarded since different types
of facilities employ diverse degrees of security. Authentication, authorization, and
access control methods, including video surveillance, biometrics, access cards, and
usernames and passwords, have been implemented at these stations, so they have
enough staff and security. As a result of their high levels of security and staffing,
these facilities have been able to remain open. However, the primary source of con-
cern is the power generation substations, which have fewer robust physical safeguards
in place [13].
The possibility of a hostile adversary bringing down more than one substation
at the same time is the most worrisome prospect of the loss. In the event that there
is significant damage to the smart grid, large metropolitan regions may experience
Challenges with Industry 4.0 Security 199

a blackout that lasts for a number of hours. The People’s Liberation Army (PLA), a
Chinese organization with political motivations, was responsible for the cascading
blackout that occurred in the United States on August 14, 2003 [14]. This incident is
an example of a real-life worst-case scenario.
Repair: if a problem is detected, it may either be isolated and fixed manually, or
an alarm may be sent to the associated control system to automatically rearrange the
backup resources so that the service can be provided at all times. The self-healing
method, which relies on the system’s ability to recognize and isolate errors and inter-
ruptions, can be used to cure the problem, or it can be sent to a professional for repair.
The goal here is to expedite the patient’s recuperation in the shortest amount of time
possible. However, critical components sometimes have no backup capacity at all, or
only a limited amount of it. Therefore, self-healing is better able to deal with more
serious injuries in a shorter period of time [15].

14.2.1.4 Vulnerabilities
There are three major categories of CPS vulnerabilities [16], as follows:

1. Platform: hardware, software, configuration, and database faults all fall

under the umbrella of “platform vulnerabilities” (PV). There are many
types of network vulnerabilities, but the most common are packet manipu-
lation attacks, denial-of-service (DoS) and distributed denial-of-service
(DDoS) attacks, back doors, communication-stack attacks, spoofing (on the
application, transport, and network layer), sniffing, replay, eavesdropping,
man-in-the-middle attacks, and others that compromise open wireless or
wired communication and connections [17].
2. Platform: hardware, software, configuration, and database faults all fall
under the umbrella of “PV”.
3. Management: there are security rules, processes, and policies missing
from the management structure [18].

14.2.2 Internet of Things
14.2.2.1 Threats
An assault that is also known as a man-in-the-middle attack occurs when an adver-
sary completely or partially modifies the data stream of an IoT device in order to
impersonate the device or system from which the data originated. They are able to
see data that is being transmitted, manipulate equipment, and intercept shared sensi-
tive information [19].
Dangers associated with the sharing of information in order to gain information
without proper authority, attackers, may listen in on broadcasts and eavesdrop, jam
the signal to prevent information from being distributed, or partly override the broad-
cast in order to substitute it with fraudulent information [20]. They then threaten to
make the information public or sell it.
Attackers have the ability to obtain access to the firmware or operating system of
a device that is running an IoT application and subsequently changes it, either par-
tially or entirely, on the device. Then, in order to access the network and any other
200 Intelligent Analytics for Industry 4.0 Applications

associated services, they will utilize the authentic IDs of the device and a­ pplication.
Assaults such as SQL or XML injection, as well as distributed denial-of-service
attacks, are examples of tampering concerns for the IoT applications [21].
Threats posed by the elevation of privilege attackers create damage by modifying
the access control rules of an application by using unprotected IoT applications. For
instance, in a manufacturing or industrial setting, an attacker may do harm to the
production system or personnel by forcing a valve to open all the way when it should
only open halfway. This would be contrary to the intended behavior of the valve.
Malware is still capable of infecting the majority of devices that are connected to
the IoT [22], despite the fact that these devices have little processing capability. In the
most recent few years, fraudsters have been very successful when making use of this
tactic. IoT botnet malware is one of the most common types since it is both adapt-
able and profitable for cybercriminals. It is also one of the most often encountered
forms. The most prominent cyberattack occurred in 2016 when Mirai brought major
websites and online services to a halt by enlisting the help of a legion of common
IoT devices. Malware that mines cryptocurrencies and ransomware are two more
categories of malicious software.
An increase in the number of cyberattacks. DDoS attacks, often known as DDoS
assaults, frequently make use of infected devices [23]. Devices that have been com-
promised can also be used as an attack base to infect further machines and dis-
guise malicious behavior. They can also serve as an entry point for lateral movement
within a corporate network. Even while businesses would appear to be the more
lucrative targets, even smart homes are subject to a startling amount of cyberattacks
that were not anticipated.
Theft of information and possible disclosure to third parties. Connected gadgets,
like anything else that has to do with the internet, enhance the likelihood of exposure
to threats on the internet. Important information, both technical and personal, might
be mistakenly kept in these devices and used to target individuals [22].
Poor management and incorrect setting of the device. Inadequate management of
devices, carelessness with passwords, and other security oversights can all contribute
to the success of these assaults. There is also the possibility that users just lack the
knowledge and competence to adopt appropriate security measures, in which case
service providers and manufacturers may need to assist their clients in achieving
greater protection.

14.2.2.2 Vulnerabilities
IoT applications suffer from various vulnerabilities that put them at risk of being
compromised [24] are as follows:

1. Passwords that are easy to guess or are hard coded. There are a lot of pass-
words that are easy to guess, exposed to the public, or cannot be changed.
Some members of the IT team do not bother changing the device’s or soft-
ware’s factory-default password once they install it.
2. The absence of a system or procedure for updating. Many IoT applications
and devices do not receive updates because the administrators of IT systems
cannot see them on the network. Additionally, IoT devices may not even
Challenges with Industry 4.0 Security 201

have an update mechanism built into them owing to their age or function, so
administrators cannot frequently update the software on these devices.
3. Network services and ecosystem interfaces that are not properly s­ afeguarded.
Every connection made by an IoT app has the potential to be hacked, either
as a result of a flaw that is built into the components themselves or as a result
of the fact that they are not protected from intrusion. This encompasses
any gateway, router, modem, external web app, application programming
­interface (API), or cloud service connected to an IoT app.
4. IoT app components are either obsolete or not secure. When being
­constructed, many applications for the IoT make use of frameworks and
libraries provided by third parties. They may be a threat to network security
if they are out of date, have been found to be vulnerable in the past, and their
installation in a network is not confirmed [23, 24].
5. Storing and transmitting data in an unsecured environment. Between the
IoT apps and other connected devices and systems, it is possible to store and
transfer many different sorts of data. Everything needs to be encrypted and
securely protected using transport layer security or one of the other proto-
cols, depending on the situation.

14.2.3  Cloud Computing

14.2.3.1 Threats
Common threats to cloud data security are as follows:

1. Identity, Authentication, and Access Management: this includes the

absence of multifactor authentication, incorrectly set access points, weak
passwords, scalable identity management systems, and continual automated
rotation of cryptographic keys, passwords, and certificates. There is also a
lack of continual automatic rotation of cryptographic keys and passwords.
2. Vulnerable public APIs: The development of application programming
interfaces is necessary to prevent sensitive data from being accessed mali-
ciously as well as accidentally. In addition to authentication and access con-
trol, it is recommended that encryption be used, as should the ability to keep
track of user activities.
3. The attackers may attempt to spy on user activities and transactions, change
data, return fraudulent information, and redirect users to sites that are not
legitimate.
4. Malicious insiders. It is possible for a current or former employee or con-
tractor of an organization with authorized access to the company’s network,
systems, or data to willfully misuse the access in a manner that leads to a
data breach or negatively impacts the organization’s information systems.
Insider threats include malicious insiders.
5. Sharing data is a kind of cooperation. Data exchange between businesses
are made simple by several cloud providers. As a result, fraudsters now have
a larger target pool from which to launch attacks and steal critical data.
202 Intelligent Analytics for Industry 4.0 Applications

Denial-of-service attacks: As a result of cloud infrastructure disruption,

6.
hackers can inflict damage on enterprises without having access to their
cloud service accounts or internal network.

14.2.3.2 Vulnerabilities
1. Cloud storage with incorrect configuration: data theft is made easier
because of the widespread use of cloud storage. Businesses continue to con-
figure cloud storage inappropriately despite the severe dangers, resulting
in huge financial losses for many [24]. Inadequate constraints or protec-
tions: unauthorized access to your cloud infrastructure might put your orga-
nization at risk if you do not have enough controls or protections in place.
Insecure cloud storage buckets can lead to attackers gaining access to your
company’s secret data and downloading it, which can have serious conse-
quences. The default setup of Amazon Web Services (AWS) led to a huge
number of data breaches since the S3 buckets were accessible by default
[25].
2. Theft or loss of an organization’s intellectual property: intellectual prop-
erty (IP) is unquestionably one of the most precious assets that a business
possesses; but it is also extremely susceptible to security risks, particularly
when the data in question is kept online [26]. Cloud-based file-sharing sys-
tems have been found to include sensitive information such as intellectual
property in roughly 21% of the files uploaded. When cloud services are
breached, attackers may be able to access sensitive information stored in
them.
Changes made to the data: if the data are changed in a way that pre-
vents them from being returned to their original condition, this might lead
to a loss of the data’s entire integrity and could render the data unusable.
Deletion of Data: An adversary’s ability to remove critical data from a
cloud service, which constitutes a clear and present danger to an organiza-
tion’s operations, is a major concern when it comes to data security.
Loss of Access: attackers can lock up information and demand a pay-
ment in order to release it (this is known as a ransomware attack), or they
can encrypt data using robust encryption keys until they carry out their
nefarious operations. As a result, operating in the cloud necessitates putting
in place precautionary measures to secure your intellectual property and
data [17].
3. Unsatisfactory Management of Access: incorrect management of user
access is perhaps the most prevalent security concern associated with cloud
computing. In security breaches involving online applications, credentials
that have been stolen or lost have been the instrument that has been utilized
by attackers the most frequently for the past many years.
Access management guarantees that persons are only able to carry out
the activities for which they are responsible. Authorization refers to the pro-
cess of determining what a person is permitted to access and validating
that permission. In addition to the standard access management issues that
plague organizations in the modern day, such as user password fatigue and
Challenges with Industry 4.0 Security 203

managing a distributed workforce, there are also a number of cloud-specific

challenges that organizations must contend with. These include the follow-
ing issues and more [17, 23]:
a. Users who are allocated but are not active.
b. Multiple administrator accounts.
c. Users avoiding business access management controls.
Devious Actors on the Inside: even if you take precautions to protect your-
4.
self from the various potential threats posed by cloud computing, you may
still be susceptible to assaults by malevolent insiders, including both current
and past employees.
If your adversaries are able to get illegal access to your systems, they
have the capability of stealing information, destroying data, and sabotaging
your IT infrastructure. Research published by Ponemon in 2020 states that
the number of insider attacks has grown by 47% since 2018. The cost of
insider attacks has increased by 31% since 2018. Only 23% of insider threats
are intentional; the majority of insider attacks are due to incompetence.
Insecure APIs: the usage of application programming interfaces, or APIs,
5.
is a common way to make cloud computing more efficient. APIs, or appli-
cation programming interfaces, simplify the process of exchanging data
across many software programs and are therefore widely utilized in work-
places. APIs, which are well known for their practicality and the capacity to
improve efficiency, can also be a source of vulnerabilities in the cloud [13].
Attackers can readily access company data and conduct DDoS assaults
by exploiting weak APIs. When they initiate attacks against APIs, sophisti-
cated attackers might employ a number of techniques to avoid being discov-
ered by security systems.
As a result of the rising reliance that businesses have on APIs, there has
been a rise in the number of attacks that target APIs. Abuse of application
programming interfaces is projected to become the most prevalent type of
attack vector by the year 2022, as stated by Gartner.

14.3  SECURITY CHALLENGES OF THE INDUSTRY 4.0

14.3.1 Inadequate IT/OT Security Expertise,
Knowledge, or Understanding
As mentioned above, this difficulty is caused by personnel participating in changing
industrial processes or digital security. Both lack information security skills. People
involved in the manufacturing process do not know what security measures to take,
and those in charge of security do not fully comprehend the process to secure it from
outside threats. Cybersecurity training are getting fewer and more expensive [27].

14.3.2 Inadequate Security Policies and Funds

The 4th industrial revolution sector is still unwilling to focus on security. The lack
sees this of policies and organizational standards that assure product privacy and
204 Intelligent Analytics for Industry 4.0 Applications

security. No matter how much we stress it, the industry has not embraced privacy and
security. Most operators consider it an add-on or luxury function, which leads to our
following argument [28].
For the same reason that industries and operators fail to prioritize security, they
have cut back on money and are reluctant to engage in R&D for current security
concerns. This creates problems.
A corporation that migrates its information systems to the cloud instead of storing
them locally isn’t spending enough on security. Now, when a corporation discovers
how much it is wasting on saving money by migrating to cloud services is a major
motivator for storing data locally. How much would a corporation spend on cloud
security if it only wanted to save money [29]? Cybersecurity will not improve as it
should.

14.3.3 The Legal Obligation Regarding Products

Next, industry 4.0 operators have a security challenge with procedures. The problem
is Industry 4.0 cybersecurity liability. Many stakeholders are engaged in making a
smart device that can be connected to the internet via IoT. These stakeholders include
the manufacturing brand, companies who supply one of the numerous parts used to
make that smart device, security teams working on the security measures included
in that smart device, software teams that program that device, and many more in the
supply chain. This large number of stakeholders apportions accountability in case a
smart device security event occurs, which might be a big obstacle in implementing
industry 4.0 [30].

14.3.4 Uneven and Insufficient Standardization

Over the ages, science has advanced greatly. Thousands of studies, publications, and
standards make these technologies easy to deal with, but not Industry 4.0 security.
Unlike IoT or other new technologies with comprehensive standardization, Industry
4.0 security has little or no market standards. Only fragmented or unagreed-upon
standards are used by existing Industry 4.0 operators [31].

14.3.5 The Devices’ Capabilities and Limitations

from a Technical Perspective

The digitalization of existing industrial processes is a key component of industry

4.0. Thus it is not a brand-new system. Instead, the identical tasks previously per-
formed by hand will now be carried out digitally. This means that digital platforms
will be incorporated into existing platforms, which is the only way to avoid starting
from scratch. This easy technique to adopt Industry 4.0 has several drawbacks. Most
gadgets now in use or made using obsolete approaches have technological limita-
tions that pose security risks. Most of these devices have low power and can’t do
simple tasks. One of the main reasons is to keep pricing stable. All these already-
operating devices lack the foundation to implement any protective measures. All this
Challenges with Industry 4.0 Security 205

equipment was intended for one major operation, yet their protection was ignored.
Many new devices lack this basic protection, too [6, 18].

14.4 CONCLUSION
In today’s manufacturing industry, moving toward Industry 4.0 presents a wide range
of technological difficulties, many of which have a significant impact on the security
domain. An incident that has substantial and dramatic ramifications for the organization
is normally handled after the development process is complete. Organizational security
is commonly defined as a financial investment with no return on investment, which is
why this is the case Return on Investment (ROI). Although this procedure is expensive, it
often does not provide a long-term solution to the problem at hand. This may also impair
the source of differentiation among rivals and diminish the competitive advantages and
organizational trust in their commercial activities. Prior to beginning the implementa-
tion of Industry 4.0’s technological problems, it is vital to design a strategy that includes
all stakeholders and to agree on security issues as well as the appropriate architecture.
Security concerns are already being brought up. Concerns about security and putting
in place suitable organizational safeguards to protect processing activities have been
reinforced by European Union (EU) directive 2016/679. Considering how important it is
to secure the IT systems in industrial equipment, these talks will make the next stage of
industrial development safer and more secure for everyone.
The experts in the working groups will play a crucial role in highlighting the vari-
ous options that organizations will have in the near future. Even so, it is important to
find and compare solutions that have worked well in the past and can help all stake-
holders in the different business sectors.

REFERENCES
1. Mohamed Abomhara and Geir M. Køien. Cyber security and the internet of things:
Vulnerabilities, threats, intruders and attacks. Journal of Cyber Security and Mobility,
4: 65–88, 2015.
2. Nilufer Tuptuk and Stephen Hailes. Security of smart manufacturing systems. Journal
of Manufacturing Systems, 47: 93–106, 2018.
3. Beyzanur Cayir Ervural and Bilal Ervural. Overview of cyber security in the indus-
try 4.0 era. In Industry 4.0: Managing the Digital Transformation, pages 267–284.
Springer, 2018.
4. T. Pereira, L. Barreto, and A. Amaral. Network and information security challenges
within industry 4.0 paradigm. Procedia Manufacturing, 13: 1253–1260, 2017.
5. Katalin Ferencz, József Domokos, and Levente Kovács. Review of industry 4.0 security
challenges. In 2021 IEEE 15th International Symposium on Applied Computational
Intelligence and Informatics (SACI), pages 245–248. IEEE, 2021.
6. Mohammed M. Alani and Mohamed Alloghani. Security challenges in the industry 4.0
era. In Industry 4.0 and Engineering for a Sustainable Future, pages 117–136. Springer,
2019.
7. Capgemini (2015). Securing the internet of things opportunity: Putting cybersecurity
at the heart of the IoT. Capgemini Worldwide. Retrieved March 10, 2021, from https://
www.capgemini.com/resources/securing-the-internet-of-things-opportunity-putting-
cybersecurity-at-the-heart-of-the-iot/.
206 Intelligent Analytics for Industry 4.0 Applications

8. Rolf H. Weber and Evelyne Studer. Cybersecurity in the internet of things: Legal
aspects. Computer Law & Security Review, 32(5): 715–728, 2016.
9. Abid Haleem, Mohd Javaid, Ravi Pratap Singh, Shanay Rab, and Rajiv Suman.
Perspectives of cybersecurity for ameliorative industry 4.0 era: A review-based
framework. Industrial Robot: The International Journal of Robotics Research and
Application, 49: 582–597, 2022.
10. Jean-Paul A. Yaacoub, Ola Salman, Hassan N. Noura, Nesrine Kaaniche, Ali Chehab,
and Mohamad Malli. Cyber-physical systems security: Limitations, issues and future
trends. Microprocessors and Microsystems, 77: 103201, 2020.
11. Michael Rushanan, Aviel D. Rubin, Denis Foo Kune, and Colleen M. Swanson. Sok:
Security and privacy in implantable medical devices and body area networks. In 2014
IEEE Symposium on Security and Privacy, pages 524–539. IEEE, 2014.
12. Stephen Checkoway, Damon McCoy, Brian Kantor, Danny Anderson, Hovav Shacham,
Stefan Savage, Karl Koscher, Alexei Czeskis, Franziska Roesner, and Tadayoshi Kohno.
Comprehensive experimental analyses of automotive attack surfaces. In 20th USENIX
Security Symposium (USENIX Security 11), 2011.
13. Hossein Zeynal, Mostafa Eidiani, and Dariush Yazdanpanah. Intelligent substation
automation systems for robust operation of smart grids. 2014 IEEE Innovative Smart
Grid Technologies-Asia (ISGT ASIA), pages 786–790, 2014.
14. Thomas M. Chen, Juan Carlos Sanchez-Aarnoutse, and John Buford. Petri net model-
ing of cyber-physical attacks on smart grid. IEEE Transactions on Smart Grid, 2(4):
741–749, 2011.
15. A. Muir and J. Lopatto. Final report on the August 14, 2003 blackout in the United
States and Canada: Causes and recommendations. 2004.
16. Hang Zhang, Bo Liu, and Hongyu Wu. Smart grid cyber-physical attack and defense: A
review. IEEE Access, 9: 29641–29659, 2021.
17. Narendra Mishra, R.K. Singh, and Sumit Kumar Yadav. Design a new protocol for
vulnerability detection in cloud computing security improvement. In Proceedings of the
International Conference on Innovative Computing & Communication (ICICC), 2021.
18. Jacob Morgan. A simple explanation of the internet of things. Retrieved November, 20:
2015, 2014.
19. Derui Ding, Qing-Long Han, Yang Xiang, Xiaohua Ge, and Xian-Ming Zhang. A
survey on security control and attack detection for industrial cyber-physical systems.
Neurocomputing, 275: 1674–1683, 2018.
20. Naser Hossein Motlagh, Mahsa Mohammadrezaei, Julian Hunt, and Behnam Zakeri.
Internet of things (IoT) and the energy sector. Energies, 13(2): 494, 2020.
21. Zhiyan Chen, Jinxin Liu, Yu Shen, Murat Simsek, Burak Kantarci, Hussein T. Mouftah,
and Petar Djukic. Machine learning-enabled IoT security: Open issues and challenges
under advanced persistent threats. ACM Computing Surveys (CSUR), 55(5): 1–37, 2022.
22. Arup Barua, Md Abdullah Al Alamin, Md Shohrab Hossain, and Ekram Hossain.
Security and privacy threats for bluetooth low energy in IoT and wearable devices: A
comprehensive survey. IEEE Open Journal of the Communications Society, 3: 251–281,
2022.
23. Ritika Raj Krishna, Aanchal Priyadarshini, Amitkumar V. Jha, Bhargav Appasani,
Avireni Srinivasulu, and Nicu Bizon. State-of-the-art review on IoT threats and attacks:
Taxonomy, challenges and solutions. Sustainability, 13(16): 9463, 2021.
24. Omar Alrawi, Charles Lever, Kevin Valakuzhy, Kevin Snow, Fabian Monrose, Manos
Antonakakis, et al. The circle of life: A {Large-Scale} study of the {IoT} malware life-
cycle. In 30th USENIX Security Symposium (USENIX Security 21), pages 3505–3522,
2021.
Challenges with Industry 4.0 Security 207

25. Manish Snehi and Abhinav Bhandari. Vulnerability retrospection of security ­solutions
for software-defined cyber–physical system against DDoS and IoT-DDoS attacks.
Computer Science Review, 40: 100371, 2021.
26. Kurdistan Ali and Shavan Askar. Security issues and vulnerability of IoT devices.
International Journal of Science and Business, 5(3): 101–115, 2021.
27. Jan Svacina, Jackson Raffety, Connor Woodahl, Brooklynn Stone, Tomas Cerny,
Miroslav Bures, Dongwan Shin, Karel Frajtak, and Pavel Tisnovsky. On vulner-
ability and security log analysis: A systematic literature review on recent trends. In
Proceedings of the International Conference on Research in Adaptive and Convergent
Systems, pages 175–180, 2020.
28. David Talbot. Vulnerability seen in amazon’s cloud computing. Technology Review.
Retrieved from http://www.cs.sunysb.edu/˜sion/research/sion, 2009.
29. Sarvesh Kumar, Himanshu Gautam, Shivendra Singh, and Mohammad Shafeeq. Top
vulnerabilities in cloud computing. ECS Transactions, 107(1): 16887, 2022.
30. Nnamdi Johnson Ogbuke, Yahaya Y. Yusuf, Kovvuri Dharma, and Burcu A. Mercangoz.
Big data supply chain analytics: Ethical, privacy and security challenges posed to busi-
ness, industries and society. Production Planning & Control, 33(2–3): 123–137, 2022.
31. Adrien Bécue, Isabel Praça, and João Gama. Artificial intelligence, cyber-threats and
industry 4.0: Challenges and opportunities. Artificial Intelligence Review, 54(5): 3849–
3886, 2021.
 15 Dodging Security Attacks
and Data Leakage
Prevention for Cloud
and IoT Environments
Ishu Gupta
National Sun Yat-Sen University

Ankit Tiwari
TechMatrix IT Consulting Pvt. Ltd.

Priya Agarwal
Amdocs Development Center India LLP

Sloni Mittal
Hewlett-Packard (HP)

Ashutosh Kumar Singh

National Institute of Technology Kurukshetra

CONTENTS
15.1 Introduction................................................................................................... 210
15.2 Literature Survey........................................................................................... 211
15.2.1 Historical Notes................................................................................. 211
15.2.2 State of Art........................................................................................ 213
15.2.3 Research Gaps and Contributions..................................................... 214
15.3 Proposed Method........................................................................................... 215
15.3.1 Operational Flow............................................................................... 215
15.3.1.1 Message Encryption............................................................ 216
15.3.1.2 Message Authentication Code............................................ 217
15.3.1.3 Hash Function..................................................................... 218
15.3.1.4 Textual Documents Clustering........................................... 219
15.3.1.5 Textual Documents Graph Representation......................... 220
15.3.1.6 Depletion of Redundant Information.................................. 220
15.3.2 Data Leakage Prevention Model....................................................... 220
15.3.3 Architecture of Secure Data Transmission........................................ 221
15.4 Performance Evaluation................................................................................. 223

DOI: 10.1201/9781003321149-15 209

210 Intelligent Analytics for Industry 4.0 Applications

15.4.1 Experimental Set-Up......................................................................... 223

15.4.2 Data Security Outcome and Comparison.......................................... 223
15.4.3 DLP Outcome and Comparisons....................................................... 226
15.5 Conclusions and Future Work........................................................................ 229
Acknowledgment.................................................................................................... 230
References............................................................................................................... 230

15.1 INTRODUCTION
Cloud computing acts as a backbone for the emerging techniques such as the Internet
of Things, big data, cyber-physical system (CPS) in the computer science field Wei
et al. (2016) where the interconnection links between workstations are formed either
by means of wired network connections or remote connections Gupta et al. (2022).
With its rapid advancement, there is a need for data security as well as maintain-
ability over the network in the cloud and IoT environments Gupta et al. (2020). The
confidential data which is being revealed to malicious users can outcome in a major
setback in organizations as well as for an individual Zaghloul et al. (2019). The ille-
gitimate users carry out the attack with the help of unlicensed tools in the system
to endanger system privacy and integrity Zhang et al. (2019). Protecting the system
from undesired hostilities, illegitimate activities, and loss is called Security Gupta et
al. (2019). Data security within wireless networks is provided with the help of cryp-
tographic techniques and tools. Security over data ensures confidentiality, authentic-
ity, and integrity for the documents sent across the network which further raises the
availability and classification issues for the documents that compromise the security
to a further extent Gupta and Singh (2019). This can further lead to trade secrets,
bank details, privacy of patients, health records, and security of accounts, and the list
goes on due to inappropriate classification of documents in categories different from
the one they belong which violates the confidentiality of the users. These breaches
grow in accordance with their size along with action Gupta and Singh (2018). For
instance, eBay, an eminent online shopping website experienced one of the major
setback leaks in history when more than 145 million customers’ personal details such
as phone numbers, names, and email IDs were theft. As an outcome, a massive reset
account password was carried out from the customer’s end Singh and Gupta (2020).
To eliminate these problems executants and scholars have illuminated techniques
and methodology to safeguard the confidential data which is mainly acknowledged
by the term data leakage prevention systems (DLPSs) Kaur et al. (2017a). DLPSs
mark the continuous investigation of confidential data which lacked standard secu-
rity mechanisms such as intrusion detection and firewalls. The confidential data is
endlessly supervised by DLPSs irrespective of data “in transit” or “in use”. The sub-
ject matter and context of the private data are mainly analyzed by the DLPSs for such
kind of detection. Studying components such as sender, receiver, format, time, and
data size fall under context analysis Kaur et al. (2017b). The regular expression, fin-
gerprinting, and statistical analysis of content fall under content analysis. Thus, the
use of DLPSs is increasing for curative actions such as caution, chunks, encryptions,
and audits for safeguarding these tasks.
Dodging Security Attacks and Data Leakage Prevention 211

For all confidential and nonconfidential data, comprehensive indexing is

required. Additional storage is a major drawback, as well as processing compe-
tence, is needed. To search for acceptable and resemble documents, binary codes
as a memory address (BCMA) are used which was capable of getting resembling
documents within the insignificant time taken by locality sensitive hashing (LSH).
The main constraint is that it is only capable of getting documents determined from
specific topics like government borrowing and disaster. The perception is made
when duplicate or contents of the intimate data is being retrieved, broadcasted, or
preowned without the consent of the user. Identification takes place when duplicates
or parts of the intimate data are monitored, retrieved, or transferred without valida-
tion. Statistical analysis, regular expressions, and fingerprinting are used to recog-
nize duplicates of confidential data. The drawback of the regular expression is that it
has a restricted extent so only rule-based items are in the ease of satisfaction Baker
and McCallum (1998). Advanced fingerprinting is the fingerprints produced by the
data hashing having some faults. To overcome the limitations of the advanced fin-
gerprinting technique, a modified full-data fingerprinting was presented Shu et al.
(2015). However, in the fingerprints, the major culpability is that they are unsafe and
can be dodged with smaller changes to the authentic one: that is why to generate
altered fingerprints, k-skip-n-gram came into existence. Even after the altercation,
k-skip-n-gram is a vigorous way of detecting the authentic data (such as subtraction,
addition, and synonyms). In this method, both intimidated and nonintimidated docs
are managed to create fingerprints, in which intimidated docs create nonintimi-
dated k-skip n-grams that will help in removing unwanted n-grams in the confiden-
tial documents. In almost all scenarios this suggested procedure outperformed the
nominal fingerprinting method Alneyadi et al. (2016). Still, the major drawback of
fingerprinting is the vulnerability of being different from the one suspected. This
chapter deals with and scrutinizes the success achieved by the use of statistical
analysis to identify intimate connotations of data. Term frequency-inverse docu-
ment frequency (TF–IDF) is the weighting function for the term on which the DLP
model is based. The proposed method is shown to be bewildered when compared
to various stages.
The rest of the chapter is organized as follows: Section 15.2 entails the historical
journey, state of artworks along with the research gaps identified and contributions
made in the domain. Section 15.3 presents the proposed solution by explaining the
operational flow along with its architecture. The details of the implementation envi-
ronment achieved the outcome of the proposed scheme, and the comparisons with
the existing techniques are described in Section 15.4 followed by the conclusions and
future directions in Section 15.5.

15.2  LITERATURE SURVEY

15.2.1 Historical Notes
The central feasible networking exposure of security was given by Marin con-
sisting of traffic analysis, computer intrusion detection, and network monitor-
ing aspects of network security. Information security has been categorized into
212 Intelligent Analytics for Industry 4.0 Applications

three sections by Wu Kehe – network system security, data security, and network
business security. Flauzac has represented a similar proposal for the execution of
dispersed security solutions in a supervised collective way, often known by the
name grid of security, ensured by section of device, which is reliable and inter-
connections among workstations may be executed under system guidelines power
Shabtai et al. (2012). It is also accepted as a speculative basis for enterprise auto-
matic production systems as security defense. Wuzheng has established a scheme
over security for mobile networks on the basis of public key infrastructure (PKI).
Cryptographic tools and network security tools have also been established mode
(counter with cipher block chaining message authentication code (CBC-MAC)),
Advanced Encryption Standard (AES), and cipher-based message authentication
code (CMAC) are the various feasible applications for the hurdle related to network
security Chandra et al. (2014).
The crucial challenge for an organization is to transfer the message and docu-
ments over the network in a confidential and protected manner. According to a
survey, internal data leakage is the major reason for the threats to information
security which consist of an estimated 29% confidential or responsive accidental
data leakage, an estimated 16% theft of intellectual property, and an estimated 15%
other thefts consisting of financial data, and customer information. Boley et al.
(1999) estimated that internal threats cause more destruction than external threats,
with an overall estimate of 67%. The other major challenge in an organization is
the classification of data outcomes. An article entitled “Automatic Indexing: An
Experimental Inquiry” was published which suggested a “technique for classifi-
cation of documents automatically in accordance with their index/contents”. The
report was broadcasted by Gartner, in 2013 on “Enterprise content-aware DLP”
in contrast to the report broadcasted by Forrester Wave in 2008. Many machine
learning approaches like support vector machine (SVM) in which data belonging
to a major amount is categorized. The high FPR was achieved by the Naive-based
approach of data classifier Lewis (1998).
The security and data leakage threats are increasing due to the following
reasons:
(a) Tremendous increase in the wireless area network and rapid growth for the
­in-motion device network usage. (b) Enterprise networking and security are being
transformed by mobile and cloud devices. (c) Tremendous growth in the malware prac-
tice is adversely affecting the security. (d) There is a vital role in active attacks such
as ­modification, ­fabrication, and interruption of the messages. (e) The documents con-
taining ­sensitive data are difficult to identify. (f) Data are distributed throughout the
networks.
(g) The contribution is toward the digital data across the network. The significance
and importance of security and leakage prevention are as follows: (a) it gradually
increases the digital transactions over the network. (b)  Confidentiality, authentic-
ity, and integrity of the message are maintained. (c) Unauthorized users are unable
to access  the message sent across the network. (d) Sensitive data movement and
usage  can  be ­monitored. (e)  The exfiltration of data by the outsider hackers can be
identified.
Dodging Security Attacks and Data Leakage Prevention 213

15.2.2 State of Art
Network security is something that tends to become a new social circumstance in
today’s world. It can often lead to message leaks, message alteration, confidentiality,
and authenticity issues over the network. The crime recently committed by Kevin
Mitnick is one of the biggest crimes in the history of the world. Losses were source
code from different companies and 80 million dollars in US intellectual property.
This is the reason network security has taken a great spot in the world. To embrace
the central idea and focus behind network security, Komal Gandhi has conducted a
survey related to network security attacks Gandhi (2016). A slow internet connec-
tion can cause various attacks on networks. The attacks are classified mainly into
two classes active attacks and passive attacks. Former are caused by manipulating
the documents or information sent over the network. The attacks classified under
active attacks are Denial-of-Service (DOS) attacks, Domain Name Service (DNS)
spoofing, smurf attacks, etc. Passive attacks are the attacks in which the third party
tries to gain access to confidential documents sent over the network. The attacks
classified under the passive attacks are wiretapping port scanner and isle scan.
Security protocols were not accomplished within the Transmission Control Protocol/
Internet Protocol (TCP/IP) communication profusion. Communication is required
to make it more secure due to the expansion in the internet architecture Singh et al.
(2011); Uchnár and Hurtuk (2017). Yan et al. performed an exploration of network
security and stated some of the network security technologies based on authentica-
tion and data encryption technologies Yan et al. (2015). Some of the authentication
technology is categorized as message authentication, identity authentication, etc.
Data encryption technology in the paper included cryptography using symmetric
and asymmetric keys individually. The former case makes use of a similar key for
encrypting and decrypting the data at both sender’s as well as receiver’s sides. The
problem identified in this symmetric key encryption is the secure transfer of the
private key to both parties. While the latter used both public and private keys for
securing the data. The message encrypted using a public key can be decrypted with
the use of a private key conversely. Asymmetric key encryption is demonstrated as
more secured and intricate. Unconventional feature linked to message authentication
has been presented by Meena et al. in Meena and Jha (2016). The basic lookout of
message authentication and requirements has been discussed. It is interpreted that
data integrity and source authentication are the foremost presumptions of message
authentication. Additionally, the different inferences and methodologies that exist
for message authentication have been discussed. Afterward, the differences between
public and private key-based message authentication have been depicted. Various
problems related to lack of compatibility and key sharing are also described. Bellare
et al. have presented two related schemes named hash-based message authentication
code (HMAC) and nested construction-based message authentication code (NMAC).
The security and efficiency are provisioned by using the cryptographic hash function
Bellare et al. (1996). Their analysis consists of the consideration of generic Message
Authentication Code (MAC) schemes. However, the studies show that if notable
weaknesses are found in the MAC then the hash function needs to be relinquished
214 Intelligent Analytics for Industry 4.0 Applications

TABLE 15.1
Comparison of Network Security Outlooks
Various Outlooks
Metrics Basics Emphasis On Definition
Security attack Degrade information Influencing available An attack which endangers
system data, replaying the security over
message, analyzing information
data traffic, etc.
Process for security Formulated Data consistency, Techniques to find, restore
implementation to digital signature, or avoid any attacks over
safeguard security access control etc. security
Security service Includes confidentiality, Security endangered, The services that upgrade
authenticity, integrity processing data, and the transfer process of
and other security transmission of information and data
parameters information processing system. One or
more security techniques
can be employed

TABLE 15.2
Major Algorithms and Their Operations
Algorithms
Operations Hashing Algorithm Symmetric Asymmetric
Electronic signature SHA family RC4 DSA
Encryption – DES RSA
Key exchange MD5 AES DH

from significant as well as a huge range of other standards and known usage. The
detailed examination is done on various cryptography techniques (security algo-
rithms) such as hash function, symmetric and asymmetric technique, and different
approaches used for detecting data semantics. Table 15.1 summarizes the compari-
son of network security outlooks, and Table 15.2 lists the state-of-the-art algorithms
along with their operations that are used for securing the data.

15.2.3 Research Gaps and Contributions

The objective of this research work is to build a model which is capable of effectively
and securely transmitting the message or information over the network and categoriz-
ing the transferred data for “in-use” or “in transit” state individually by overcoming the
shortcomings of the major existing methods for securing the data that are demonstrated
in Table 15.3 Semwal and Sharma (2017). The aim is to prevent the message from trans-
mitting from one node to another in the communication channel without having any
intervention from the third or unauthorized party in the network and safeguarding the
confidentiality which might occur due to the wrong clustering of documents and leads
Dodging Security Attacks and Data Leakage Prevention 215

TABLE 15.3
Drawbacks of Existing Techniques
Operations Technique Limitations
Message Symmetric key The key can be accessed by any intruder.
encryption Asymmetric key Public key of sender which is used for decryption on
receiver’s site is vulnerable to access by another
person on the network.
Message Authentication without No encryption technique performed over the message.
authentication cryptography
code Authentication tied to a Encryption of data is not conducted. The identical keys
plain text are used for both the purposes, enciphering and
deciphering the data.
Authentication tied to Enciphering and deciphering the data is done using
cipher text symmetric key.
Hash function Authentication with Plain data is in unencrypted form.
plain text
Authentication with Message is decrypted by employing public key and that
encrypted text is accessible to any intruder on network.

to revealing of the personal information to the outside source other than the intended
user. The main goal of this work is to ensure the secure transmission of confidential
messages or information over the network, without any alteration or modification, while
maintaining confidentiality, authenticity, and integrity throughout the transmission and
use. Only authorized parties should have access to the information, and the classification
should align with predefined classes for the documents. Moreover, the documents that
are newly generated in which confidential data might be contained can be recognized if
they are classified as a confined category. Based on the semantic approach, a statistical
data leakage prevention (DLP) model is presented to classify the data. It uses neighbor-
ing context and content of personal/private data to identify malicious access and secure
the confidential data despite the data being used by the intended users or while data are
transferred internally or externally between communication channels.

15.3  PROPOSED METHOD

The proposed solution allows the safe and secure transmission of a message from
one network to another without the interruption from the third or unauthorized party
where the known and authorized party only is permissible to access the message. The
MITM (Man In The Middle) is incapable of performing the MITM attacks over the
message as the method encrypts the messages twice and it becomes very difficult to
have access to the keys or codes for the decryption.

15.3.1 Operational Flow
The proposed work achieves confidentiality, authenticity, integrity, reliability, and
availability by applying the following techniques Kumari (2017); Michail et al.
(2007).
216 Intelligent Analytics for Industry 4.0 Applications

15.3.1.1  Message Encryption

Using Symmetric Key: symmetric key encryption comes with the problem of key
transmission. Secret keys have to be sent to the receiver workstation before trans-
mitting the actual message. The channels being used for transmission are insecure
as it is not possible to promise that the channel will not be accessed by anyone
else. Therefore, exchanging the keys individually is the only secure and intact way.
The entire functioning of encryption by employing a symmetric key is depicted in
Figure 15.1.
Public Key Encryption Using Asymmetric Key: in this technique, data attains its
privacy but loses its legitimacy. For instance, the receiving station B is supposed to
get messages from station A only but all users in the network are having the public
key of B so messages may be dropped to B from any other user in the network other
than A. Therefore, receiving station B cannot be sure of the message being received
from supposed user A or any other users in the network. Figure 15.2 presents the
entire functioning of this encryption technique.

FIGURE 15.1 Encryption using symmetric key.

FIGURE 15.2 Asymmetric message encryption using public key.

FIGURE 15.3  Asymmetric data encryption through private key.

Dodging Security Attacks and Data Leakage Prevention 217

Private Key Encryption Using Asymmetric Key: in this technique, legitimacy is

attained but privacy is not achieved. Since the public key of the sender is known to
everyone in the network therefore data transmitted by the sender station could be
deciphered by employing its private key by everyone on the network other than the
supposed receiver. Figure 15.3 represents the entire functioning of the asymmetric
key encryption through private key Trnka et al. (2018).

15.3.1.2  Message Authentication Code

Here, no cryptography technique is applied to the data therefore confidentiality is
unachievable. The data can be accessed by any illegitimate user on the network.
Figure 15.4 represents the functioning of this scheme.

FIGURE 15.4  Process of message authentication code.

FIGURE 15.5  MAC without encryption.

218 Intelligent Analytics for Industry 4.0 Applications

FIGURE 15.6  MAC with encryption.

Authentication link associated with plain data: the main issue associated with
this technique is the identical key is used for both operations enciphering and deci-
phering the data. Therefore, this technique is not safe to use. If any illegitimate party
gets access to the key, then they can easily encipher and decipher the data. Figure
15.5 presents the entire scenario.
Authentication link to the cipher text: the identical symmetric key is used for
both enciphering and deciphering data due to which confidentiality is not actualized
completely. Figure 15.6 presents the entire scenario.

15.3.1.3  Hash Function

The function which is used to map data of whimsical size into a determined size is
called a hash function. Hash values, digests, and hash codes are the values that are
given back by the hash function. Fixed length output and efficiency of operations are
the features of the hash function. Hash code is built on the basis of a mathematical
function that operates on two blocks of data of fixed size. The portion of the hash-
ing algorithm is built by the hash function, as shown in Figure 15.7 Alkandari et al.
(2013).
The rounds of the above-mentioned hash function in Figure 15.7 are indulged
by the hashing algorithm like a block cipher. The determined size is taken at every
round, typically a mixture of results of the last round and the most current message
block. To hash the complete message, the required rounds are repeated as depicted
in Figure 15.8.

FIGURE 15.7  Block diagram of hash function.

Dodging Security Attacks and Data Leakage Prevention 219

FIGURE 15.8  Schematic representation of hashing algorithm.

FIGURE 15.9  Authentication without encrypted plain text.

Authentication of hash code without ciphered plain data: the plain data are not
in encrypted form; therefore, confidentiality is endangered and when this plain data
is adjoined with hash code as shown in Figure 15.9, it is vulnerable to access by an
intruder on the network. Also, the public key used for the decryption of data is acces-
sible to everyone on the network, therefore, authenticity is also endangered and we
cannot ensure the decryption of data by only the supposed receivers or anyone else.
Authentication with encrypted plain text: as decryption is done by employing a
public key, authenticity is not realized as depicted in Figure 15.10. The sent informa-
tion is accessible to everyone on the network and it is difficult to ensure if it is the
supposed recipient or anyone else.

15.3.1.4  Textual Documents Clustering

The clustering textual documents problem is related to high dimensional clustering.
Every document is assumed to be a vector containing 1,000 terms and every term
of the textual document in an individual dimension. In order to classify each and
every document related to resemblance angle, a cosine measure between two docu-
ments is calculated. Support vector machine (SVM) that uses the SVM algorithm
is used to categorize three types of data: enterprise private, enterprise public, and
nonenterprise. Detection of data leakage is 97% with 3% false negative. But this
method only identifies private and public categories and failed to detect more flexible
classification levels like top secret, secret, and confidential. Thus, this method can
220 Intelligent Analytics for Industry 4.0 Applications

FIGURE 15.10  Authentication with encrypted plain text.

no longer be used and makes the enforcement of security policies difficult Van and
Thuc (2015).

15.3.1.5  Textual Documents Graph Representation

In various text similar, tasks graphs have been used, due to which text disclosure
graphs are used as a model despite various existing techniques. The advantage of
the graph-based model is that it constitutes a term simultaneous with their context in
addition to monitoring the context along with document structure.

15.3.1.6  Depletion of Redundant Information

Various methods for the depletion of duplicate information exist such as latent
semantic indexing (LSI), principal component analysis (PCA), and singular value
decomposition (SVD). The task operating efficiency is mainly reduced by the
duplicate information which is incompetent and the worst case is in text-related
tasks.

15.3.2 Data Leakage Prevention Model

Data leakage is defined as a strategy that guarantees sensitive and confidential data
will not go outside the organization. To detect confidential data semantics, statisti-
cal data analysis is used for DLP. The most-popular term weighting function named
TF–IDF is employed to measure the tested document and the amount of information
contained in it. The predefined categories cluster the related topic together which is
basically considered to be the main aim of clustering. Considering that every cat-
egory has a confidential level, documents with restricting secrecy levels can be easily
detected, and accordingly, actions like blocks, alerts, and quarantines are taken. The
document D can be easily identified with the use of fingerprinting and regular expres-
sion if a user U is attempting malicious access, use, and alter D. However, the user
U alter the document D by inserting, deleting, or exchanging words, lines, or para-
graph. Even in the case of robust data fingerprinting, identifying altered document
Dodging Security Attacks and Data Leakage Prevention 221

FIGURE 15.11  Deployment of data leakage avoidance model.

D is a challenging event. Therefore, statistical analysis is employed to calculate and

approximate the private data with the help of the DLP model. The deployment model
showcases the various data states, analyzing the various DLP context and various
remedial actions, which have been taken is represented in diagrammatic form in
Figure 15.11 for a better understanding of the prevention model.

15.3.3 Architecture of Secure Data Transmission

In the proposed solution, plain data are first encrypted by employing the sender’s pri-
vate key to generate the ciphertext. This ciphertext produces hash code from the hash
function. Besides the copy of generated ciphertext that is adjoined together with the
hash code is being used and it is encrypted again with the public key of the receiver
for the stronger security of the document. On the receiver side, the private key of the
receiver is used to decrypt the ciphertext, and then the decryption is applied again
222 Intelligent Analytics for Industry 4.0 Applications

with the use of the sender’s public key on the obtained decrypted data to generate
the final data. At the end of the receiver, a hash code is generated from the message
which is produced after decryption with the use of a hash function. The generated
hash code is cross-checked with the decrypted hash code to validate the authenticity
of the received message. Figure 15.12 shows the workflow and functioning of the pro-
posed solution that achieves authenticity and confidentiality simultaneously unlike
the existing schemes.
Further, the documents are classified by utilizing the centroid-based classifier. For
this purpose, the documents are showcased with the help of vector space. Searching
is performed with the aim of finding some suitable documents for the query. In the
proposed centroid-based model, every text in the document is assumed to be the
vector in term space where every document is presented by the TF. TF is used in
finding connections with information retrieval and display. It determines how fre-
quently a term is occurring in a document. TF shows the count of a particular word
within the whole document. This value is often mentioned in inverse document fre-
quency (IDF). IDF is defined as a number of documents in the corpus divided by
the document frequency of a term. It is basically used for data mining and informa-
tion retrieval Cohen and Hirsh (1998). IDF plays an important role in measuring
how important a term is. Since TF considers all the terms equally important, that’s
why TF only is not sufficient to calculate the weight of a term in a document. It is
known that the common words appear more and multiple times in the document but
they have less importance. Due to this reason, the proposed models weigh down the
important terms while considering the rare ones, and to solve this purpose, both TF

FIGURE 15.12  Proposed architecture.

Dodging Security Attacks and Data Leakage Prevention 223

and IDF are combined together and named TF–IDF in the proposed solution to get
an ultimate score of term t in a document d. Further, the stop list is utilized to remove
similar words.

15.4  PERFORMANCE EVALUATION

15.4.1 Experimental Set-Up
The implementations are performed on a machine assembled with two Intel® Xeon®
Silver 4114 [email protected] GHz 40 core processor, 64-bit Ubuntu 16.04 LTS, and 128
GB RAM with the following required installation: i Text Editor ii Java Development
Kit 17.0.2. The implementation scenario is considered for the data security perfor-
mance evaluation where the two users named Parle and Britania have interchanged
their public keys. A confidential file named “conf.txt” needs to be sent from Parle to
Britania. The comprehensive features of numerous documents are gathered for the
experiments of DLP. For the performance evaluation of the DLP phase, the data sets
Badminton (Badm), Soccer (Socc), Cricket (Crick), and Tennis (Tenn) are used in the
experimental results. The considered data sets are collected from Badminton (2022);
Cricket (2022); Soccer (2022); Tennis (2022), respectively. To prepare data for analy-
sis, they are first converted into text format. The categorization of data from various
algorithms on these data sets is performed. Approximately, 80% documents are ran-
domly selected as the training set while the remaining 20% documents are taken as a
test set. The average precision outcome from 10 runs has been presented in this chap-
ter. k-Nearest neighbor (KNN), C4.5, Naive Bayesian (NB), and proposed centroid-
based categorization have been performed on the same platform. Furthermore, the
centroid-based proposed solution interpretation has been assessed and validated by
contrasting it against numerous numbers of document-gathering techniques named
KNN, NB, and C4.5.

15.4.2 Data Security Outcome and Comparison

Figure 15.13 shows the sample of the confidential message which needs to be pro-
tected during transmission in a cloud environment. For the protection of this mes-
sage, encryption is performed. Public and private keys are generated for Parle, as
shown in Figures 15.14 and 15.15, respectively. Similarly, the public and private keys
for Britania are also generated as depicted in Figures 15.16 and 15.17, respectively.
The following are the steps performed for encryption: (i) a new symmetric
or secret key is generated as displayed in Figure 15.18. (ii) The file “conf.txt” is
encrypted with the help of the obtained key in the first step. (iii) The key obtained in
step 1 is encrypted

FIGURE 15.13  Sample of confidential message.

224 Intelligent Analytics for Industry 4.0 Applications

FIGURE 15.14  Parle private key.

FIGURE 15.15  Parle public key.

by using Britania public key for the stronger protection where the encrypted secret
key is portrayed in Figure 15.19. (iv) Parle transfers both the attained encrypted key
and encrypted message to Britania that are portrayed in Figures 15.19 and 15.20,
respectively.
Britania obtains two files and performs the decryption over these files to generate
the original message. The following steps are performed by Britania for decryption:
Dodging Security Attacks and Data Leakage Prevention 225

FIGURE 15.16  Britania private key.

FIGURE 15.17  Britania public key.

1. It performs the decryption with the help of its own private key over the
obtained encrypted secret key and generates the decrypted secret key as
shown in Figure 15.21.
2. The attained encrypted message is decrypted using the secret key obtained
in step 1 and the plain text is obtained as depicted in Figure 15.22.
226 Intelligent Analytics for Industry 4.0 Applications

FIGURE 15.18  Secret key.

FIGURE 15.19  Encrypted secret key.

FIGURE 15.20  Encrypted data.

FIGURE 15.21  Decrypted secret key.

Furthermore, Table 15.4 lists the comparison of the proposed scheme with the state-
of-the-art techniques. It can be seen that no other models except the proposed solution
provide both authentication and confidentiality simultaneously.

15.4.3 DLP Outcome and Comparisons

To evaluate the DLP outcome, the data classification is performed over the four datas-
ets named badminton Badminton (2022), cricket Cricket (2022), soccer Soccer (2022),
and tennis Tennis (2022) and compared with the existing KNN and NB approaches.
The original and predicted outcomes obtained from the existing classifiers named
KNN, Naive Bayesian, and centroid-based proposed solution are reflected in Figures
Dodging Security Attacks and Data Leakage Prevention 227

FIGURE 15.22  Generated text.

TABLE 15.4
Comparison of Proposed and Other Considered Methods
Technique Key Mode Authentication Confidentiality
Message Symmetric key – Yes No
encryption Asymmetric key Public key No Yes
encryption
Private key Yes No
encryption
Message Symmetric key Authentication link Yes No
authentication to plain text
code Authentication link Yes No
to ciphertext
Hash function Asymmetric key Hash function No No
without encrypted
text
Hash function with No Yes
encrypted text
Proposed solution Asymmetric key – Yes Yes

15.23 and 15.24 and Table 15.5, respectively where all the three schemes are imple-
mented at the same platform. It can be seen that the centroid-based proposed scheme
categorizes the data correctly unlike the existing classifiers because of maintaining
the resemblance between the specific class and a test document. Furthermore, these
outcomes are used to compute the accuracy for measuring the performance of the
model, which is presented in Table 15.6.
The categorization of data from various algorithms on multiple data sets is per-
formed. Table 15.6 depicts the accuracy of data categorization achieved by the pro-
posed scheme and comparison with the existing classifiers KNN and NB. It can be
seen that the proposed scheme which is based on a centroid
classifier outperforms both the Naive Bayes and KNN for all the data sets,
i.e., Cricket (Crick), Tennis (Tenn), and Soccer (Socc) except Badminton (Badm).
Centroid-based classifier outperforms various algorithms and performs well in all
the cases due to regulating the resemblance between the specific class and a test
document.
Differentiation between various classification algorithms using some sample
pairs tested values is listed in Table 15.6. Here, the analytical notable results are
228 Intelligent Analytics for Industry 4.0 Applications

FIGURE 15.23  Outcome of k-means algorithm.

FIGURE 15.24  Outcome of Naive Bayesian algorithm.

summarized using the sample paired test in which various classification algorithms
are taken. It compares the performance of the two classifiers. It is shown whether a
classifier is performing best or worse than another classifier. It shows that a better
performance is given by the row classifier than the classifier represented in the col-
umn. It is found that the centroid-based proposed scheme is three times better than
NB and worse in one data set. Alike, the centroid-based proposed scheme is three
times better than C4.5 and one time worse in all four data sets. Overall, the cen-
troid classifier-based proposed scheme is unpredictably good than the other existing
Dodging Security Attacks and Data Leakage Prevention 229

TABLE 15.5
Comparison of Actual and Predicted Labels
Sr. No. Original Predicted Sr. No. Original Predicted
0 Badminton Badminton 18 Soccer Soccer
1 Badminton Badminton 19 Soccer Soccer
2 Badminton Badminton 20 Soccer Soccer
3 Badminton Badminton 21 Soccer Soccer
4 Badminton Badminton 22 Soccer Soccer
5 Badminton Badminton 23 Soccer Soccer
6 Badminton Badminton 24 Soccer Soccer
7 Badminton Badminton 25 Soccer Soccer
8 Badminton Badminton 26 Soccer Soccer
9 Cricket data Cricket data 27 Soccer Soccer
10 Cricket data Cricket data 28 Tennis Tennis
11 Cricket data Cricket data 29 Tennis Tennis
12 Cricket data Cricket data 30 Tennis Tennis
13 Cricket data Cricket data 31 Tennis Tennis
14 Cricket data Cricket data 32 Tennis Tennis
15 Cricket data Cricket data 33 Tennis Tennis
16 Cricket data Cricket data 34 Tennis Tennis
17 Soccer Soccer 35 Tennis Tennis

TABLE 15.6
Comparative Analysis of the Proposed and Other Considered Methods
Categories
Algorithms Cricket Tennis Badminton Soccer
Naive Bayesian 89.3 91.2 84.3 72.3
KNN 85.8 87.5 77.5 84.6
Proposed scheme 91.8 93.9 82.7 94.2

classifiers such as KNN, NB, and C4.5 since it regulates the resemblance between the
specific class and a test document. The mean resemblance is calculated between the
test document and all the additional documents in the current class. If the magnifica-
tion is higher, it corresponds to a small level of resemblance between the documents,
whereas if the magnification is lower, it corresponds to a high level of resemblance
among the documents.

15.5  CONCLUSIONS AND FUTURE WORK

The malicious affairs are rising expeditiously during transmission on the network,
therefore increased security is required on the network for the data being transferred.
230 Intelligent Analytics for Industry 4.0 Applications

For this, authenticity and confidentiality must be maintained to forward the mes-
sages from sender to receiver in the most consistent and safe form. This chapter has
reviewed some of the common existing techniques, highlighted the limitations, and
provided an integrated solution to overcome the existing problems. The proposed
solution achieved both confidentiality and authenticity simultaneously unlike the
existing scheme with the use of asymmetric key and applying double encryption and
decryption to secure the message from intruder’s undesirable access. Now, the data
can be securely transferred; however, still, data leakage can take place. Data leakage
can happen because of using bad clustering approaches. To prevent data leakage,
an approach based on a centroid document classifier is proposed in which the data
is clustered appropriately which helps in DLP. Centroid-based document classifica-
tion is steady and maintainable, and it outperforms other classification algorithms on
various datasets that prove the validity of the proposed model.
In the future, the emerging quantum machine learning algorithms could be incor-
porated with the work for improved accuracy in identifying and preventing data leak-
age incidents. Furthermore, the malicious entity could be predicted in advance using
a quantum computing mechanism to strongly protect the data against data leakage
occurrences.

ACKNOWLEDGMENT
This work is supported by the University Grants Commission (UGC), Ministry of
Human Resource Development (MHRD), and Government of India under the scheme
of National Eligibility Test-Junior Research Fellowship (NET-JRF) with grant no.
F.15-9(JUNE 2015)/2015(NET).

REFERENCES
Alkandari, A. A., Al-Shaikhli, I. F., & Alahmad, M. A. (2013, September). Cryptographic
hash function: A high level view. In 2013 International Conference on Informatics and
Creative Multimedia (pp. 128–134). IEEE.
Alneyadi, S., Sithirasenan, E., & Muthukkumarasamy, V. (2016). A survey on data leakage
prevention systems. Journal of Network and Computer Applications, 62, 137–152.
Badminton (2022). Badminton Dataset. https:badminton-information.com
Baker, L. D., & McCallum, A. K. (1998, August). Distributional clustering of words for text
classification. In Proceedings of the 21st Annual International ACM SIGIR Conference
on Research and Development in Information Retrieval (pp. 96–103). ACM.
Bellare, M., Canetti, R., & Krawczyk, H. (1996, August). Keying hash functions for message
authentication. In Annual International Cryptology Conference (pp. 1–15). Springer.
Boley, D., Gini, M., Gross, R., Han, E. H. S., Hastings, K., Karypis, G., … & Moore, J. (1999).
Document categorization and query generation on the world wide web using webace.
Artificial Intelligence Review, 13(5), 365–391.
Chandra, S., Paira, S., Alam, S. S., & Sanyal, G. (2014, November). A comparative survey
of symmetric and asymmetric key cryptography. In 2014 International Conference on
Electronics, Communication and Computational Engineering (ICECCE) (pp. 83–93).
IEEE.
Cohen, W. W., & Hirsh, H. (1998, August). Joins that generalize: Text classification using
WHIRL. In KDD (pp. 169–173). AAAI Press.
Dodging Security Attacks and Data Leakage Prevention 231

Cricket (2022). Cricket Dataset. https:cricbuzz.com

Gandhi, K. (2016, March). Network security problems and security attacks. In 2016
3rd International Conference on Computing for Sustainable Global Development
(INDIACom) (pp. 3855–3857). IEEE.
Gupta, I., & Singh, A. K. (2018). A probabilistic approach for guilty agent detection using
bigraph after distribution of sample data. Procedia Computer Science, 125, 662–668.
Gupta, I., & Singh, A. K. (2019). Dynamic threshold based information leaker identification
scheme. Information Processing Letters, 147, 69–73.
Gupta, I., Singh, N., & Singh, A. K. (2019). Layer-based privacy and security architecture for
cloud data sharing. Journal of Communications Software and Systems, 15(2), 173–185.
Gupta, I., Gupta, R., Singh, A. K., & Buyya, R. (2020). MLPAM: A machine learning and
probabilistic analysis based model for preserving security and privacy in cloud environ-
ment. IEEE Systems Journal, 15(3), 4248–4259.
Gupta, I., Singh, A. K., Lee, C. N., & Buyya, R. (2022). Secure data storage and sharing
techniques for data protection in cloud environments: A systematic review, analysis, and
future directions. IEEE Access, 10, 71247–71277.
Kaur, K., Gupta, I., & Singh, A. K. (2017a, August). A comparative evaluation of data leakage/
loss prevention systems (DLPS). In Proc. 4th Int. Conf. Computer Science & Information
Technology (CS & IT-CSCP) (pp. 87–95).
Kaur, K., Gupta, I., & Singh, A. K. (2017b, December). Data leakage prevention: e-mail
­protection via gateway. In Journal of Physics: Conference Series (Vol. 933, No. 1,
p. 012013). IOP Publishing.
Kumari, S. (2017). A research paper on cryptography encryption and compression techniques.
International Journal of Engineering and Computer Science, 6(4), 20915–20919.
Lewis, D. D. (1998, April). Naive (Bayes) at forty: The independence assumption in informa-
tion retrieval. In European Conference on Machine Learning (pp. 4–15). Springer.
Meena, U., & Jha, M. K. (2016, March). A retrospective investigation of message authentica-
tion in wireless sensor networks: A review. In 2016 3rd International Conference on
Computing for Sustainable Global Development (INDIACom) (pp. 613–616). IEEE.
Michail, H. E., Kakarountas, A. P., Selimis, G., & Goutis, C. E. (2007, July). Throughput
optimization of the cipher message authentication code. In 2007 15th International
Conference on Digital Signal Processing (pp. 495–498). IEEE.
Semwal, P., & Sharma, M. K. (2017, September). Comparative study of different cryptographic
algorithms for data security in cloud computing. In 2017 3rd International Conference
on Advances in Computing, Communication & Automation (ICACCA)(Fall) (pp. 1–7).
IEEE.
Shabtai, A., Elovici, Y., & Rokach, L. (2012). A survey of data leakage detection and preven-
tion solutions. Springer Science & Business Media.
Shu, X., Yao, D., & Bertino, E. (2015). Privacy-preserving detection of sensitive data exposure.
IEEE Transactions on Information Forensics and Security, 10(5), 1092–1103.
Singh, A. K., & Gupta, I. (2020). Online information leaker identification scheme for secure
data sharing. Multimedia Tools and Applications, 79(41), 31165–31182.
Singh, S. K., Singh, M. P., & Singh, D. K. (2011). A survey on network security and attack
defense mechanism for wireless sensor networks. International Journal of Computer
Trends and Technology, 1(2), 9–17.
Soccer (2022). Soccer Dataset. https:eurosoccerfan.com
Tennis (2022). Tennis Dataset. tennis-ontheline.com
Trnka, M., Cerny, T., & Stickney, N. (2018). Survey of authentication and authorization for the
internet of things. Security and Communication Networks, 2018, 1–18.
Uchnár, M., & Hurtuk, J. (2017, January). Safe user authentication in a network environment.
In 2017 IEEE 15th International Symposium on Applied Machine Intelligence and
Informatics (SAMI) (pp. 000451–000454). IEEE.
232 Intelligent Analytics for Industry 4.0 Applications

Van, D. H., & Thuc, N. D. (2015, August). A privacy preserving message authentication
code. In 2015 5th International Conference on IT Convergence and Security (ICITCS)
(pp. 1–4). IEEE.
Wei, J., Liu, W., & Hu, X. (2016). Secure data sharing in cloud computing using revoca-
ble- storage identity-based encryption. IEEE Transactions on Cloud Computing, 6(4),
1136–1148.
Yan, F., Jian-Wen, Y., & Lin, C. (2015, June). Computer network security and technology
research. In 2015 Seventh International Conference on Measuring Technology and
Mechatronics Automation (pp. 293–296). IEEE.
Zaghloul, E., Zhou, K., & Ren, J. (2019). P-mod: Secure privilege-based multilevel organiza-
tional data-sharing in cloud computing. IEEE Transactions on Big Data, 6(4), 804–815.
Zhang, L., Cui, Y., & Mu, Y. (2019). Improving security and privacy attribute based data
­sharing in cloud computing. IEEE Systems Journal, 14(1), 387–397.
 16 Role of Blockchain
in Industry 4.0
Keshav Kaushik and Bhavana Kaushik
University of Petroleum & Energy Studies

CONTENTS
16.1 Introduction to Blockchain Technology: Characteristics and Protocols....... 233
16.2 Challenges and Solutions of Blockchain Technology in Industry 4.0........... 236
16.3 Future Research Scope of Blockchain in Industry 4.0.................................. 238
16.4 Conclusion..................................................................................................... 238
References............................................................................................................... 239

16.1 INTRODUCTION TO BLOCKCHAIN TECHNOLOGY:

CHARACTERISTICS AND PROTOCOLS
A blockchain is a distributed, decentralized database of records that allows for quick,
secure transactions without being monitored by a centralized administration. The
future of cryptocurrencies is anyone’s guess, but it is important to recognize the
importance of blockchain technology in a number of financial and nonfinancial
areas. Blockchain is a development of linked data structures, or blocks, that store or
trace everything that happens in distributed environments in a distributed organiza-
tion. Every block is joined to the preceding block by a unique pointer known as a
hash pointer, culminating in a chain and a structure made up of annexes: a permanent
and unchangeable history that may be used as a reviewing trail by any member to
continuously verify the accuracy of the records by simply analyzing the data itself.
On a blockchain (Zheng et al., 2017), a digital signature that attests to its legiti-
macy protects every transaction. The information saved on the blockchain is tamper
proof and cannot be altered thanks to the use of encryption and digital signatures.
Consensus, which is the process of all network members coming to an accord, is
made possible by blockchain technology. Every piece of information kept on a block-
chain is digitally recorded and has a shared history that is accessible to everyone on
the network. By doing this, any possibility of fraudulent or transaction repetition is
avoided without the use of a third party.
Imagine a situation in which you are trying to find a way to transmit some funds to
a buddy who lives somewhere in order to better grasp blockchain. A bank or a money
transfer service like PayPal or Paytm are two options you can often employ. With this
choice, third parties are needed to complete the transaction, which results in a cost
that is added to your money as a transfer fee. Additionally, in situations like these,
it is impossible to guarantee the security of your funds because it is very likely that

DOI: 10.1201/9781003321149-16 233

234 Intelligent Analytics for Industry 4.0 Applications

a hacker may interrupt the network and take your money. The victim in both situa-
tions is the client. Blockchain is useful in this situation. If we utilize a blockchain in
these situations rather than a bank to transfer money, the procedure is significantly
simpler and more reliable. As the money is processed directly by you, there is no
additional cost as a third party is not required. Additionally (Gamage et al., 2020),
the blockchain system is decentralized and not restricted to a particular place, mak-
ing all the data and records maintained to be public and decentralized. No danger of
data tampering by a hacker exists since the data is not kept in a single location. There
are several characteristics of blockchain that makes it a promising technology. The
blockchain’s characteristics are shown in Figure 16.1.

• Immutable: any data that have been added to a blockchain cannot be modi-
fied after it has been done so, which is known as the immutability feature
of a blockchain. Attempt sending an email as an illustration to better grasp
immutability. An email that has been sent to a large group of recipients

FIGURE 16.1  Characteristics of blockchain.

Role of Blockchain in Industry 4.0 235

cannot be canceled. You will have to ask each receiver of your email to
erase it, which is a laborious workaround. This is the operation of immu-
tability. Data cannot be modified or amended once it has been processed.
Because each block in a blockchain retains the hash of the one before it, if
you try to modify the data of one block, you will have to update the whole
blockchain that follows it. Any changes to one hash will affect the remain-
ing hashes as well. Since it takes a lot of processing power to modify all the
hashes, it is quite difficult for somebody to do so. Data stored in a block-
chain is unaffected by modifications or hacker attacks due to irreversibility.
• Peer to Peer: the usage of blockchain makes it simple to engage between
two parties using a peer-to-peer architecture without the need for a mid-
dleman. Blockchain (Kaushik et al., 2020) is a peer-to-peer protocol that
enables each member of the network to have an exact copy of each trans-
action, allowing machine consensus for authorization. For instance, with
blockchain, you may complete any transactions from one region of the
world to the other in a matter of seconds. Any additional costs or delays
won’t be deducted from the transaction either.
• Tamper Proof: because blockchains have the immutability characteristic
built in, it is simpler to spot data manipulation. Because any alteration to
even a single block can be easily discovered and corrected, blockchains are
thought to be tamper proof. Hashes and blocks are the two main methods
for spotting tampering. Each block-related hash function is distinct, as was
previously mentioned. It is comparable to the fingerprint of a block. Any
alteration to the information will cause the hashing algorithm to change as
well. A hacker would have to modify the hashes of all the blocks following
that one in addition to making any modifications because the hash function
of one block is connected to the next, which is a challenging task.
• Decentralized: Since blockchains are decentralized, no one individual or
organization has control over the whole network. The distributed ledger is
accessible to everybody on the internet, but nobody is able to independently
edit it. This particular aspect of blockchain technology provides safety and
transparency while offering people control.
• Consensus: Each blockchain has an agreement to assist the network in
reaching choices quickly and impartially. Consensus (Dhar Dwivedi et
al., 2021) is a decision-making technique that helps the network’s active
nodes swiftly come to a consensus and ensures the system runs smoothly.
Although nodes might not have much confidence in one another, they might
have confidence in the network’s central algorithm. There are numerous
accessible consensus techniques, each having advantages and disadvan-
tages. A consensus method is necessary for any blockchain or else it will
start to lose value.

It is crucial to realize that there are hundreds of procedures in use, making it impos-
sible to research all of the available possibilities in a reasonable length of time. The
most crucial protocols, nevertheless, are the following four: a summary of the most
common protocols used in blockchain development solutions is provided below.
236 Intelligent Analytics for Industry 4.0 Applications

• Multichain: in order to promote more effective transactions and to

­generate new implementations for the proof-of-work methods that block-
chain ­ networks depend on, multichain was founded to assist for-profit
businesses in building private blockchains. As a privately held business,
multichain is able to provide an Application Programming Interface (API)
that blockchain development services may utilize to simplify collaboration
and ­hasten implementation. The way that multichain is built to function
with fiat ­currencies and actual repositories of value sets it distinct from its
rivals. On  the other hand, the majority of cryptocurrency initiatives are
focused on the eventual replacement of traditional currency with digital
forms of trade.
• Corda: Corda is a rival of multichain that provides a protocol tailored for
businesses. The financial and banking industries have seen the most Corda
application development. Nevertheless, Corda’s technology may be applied
to a wide range of unique blockchain applications. Corda is a suitable option
for blockchain development alternatives in the finance sector because it is
certified by the R3 banking consortium.
• Hyperledger: An open-source initiative called Hyperledger (Kaushik
et al., 2022) intends to develop a set of tools that businesses may use to
install blockchain technology rapidly and efficiently. The protocol’s librar-
ies, which speed up implementation, make it a popular choice for block-
chain software applications. Strongly supporting Hyperledger, the Linux
Foundation has contributed considerable expertise to hasten the devel-
opment of the ­standard. Due to the fact that Linux and Hyperledger are
extremely interoperable, these servers are often employed in the modern
corporate environment.
• Ethereum: A variant of Ethereum’s software geared for commercial use
cases is available. The objective of Ethereum Enterprise is to expand the
commercial use cases for blockchain technologies. Businesses may quickly
create extensive apps to trade value with Ethereum Enterprise. The main
benefit of Ethereum Enterprise is that it enables companies to develop their
own private variations of Ethereum while still utilizing the most recent
Ethereum technology. Ethereum’s license makes it challenging to create
proprietary versions of the program under normal conditions, but the com-
mercial edition offers enterprises a workaround.

16.2 CHALLENGES AND SOLUTIONS OF BLOCKCHAIN

TECHNOLOGY IN INDUSTRY 4.0
One of the upcoming mobile technologies that Industry 4.0 will develop with is
blockchain. Blockchain technology has the potential to improve security, confiden-
tiality, and data transparency both for small and large organizations. Modern manu-
facturing practices that accelerate corporate goals are combined to form Industry
4.0. Research has recently focused on the potential impact of a number of Industry
4.0 technologies, including artificial intelligence (AI), IoT, big data, and blockchain.
Role of Blockchain in Industry 4.0 237

These technologies give the industrial and supply chain sectors a wide choice of
possibilities. Blockchain is a digital ledger that has attracted a lot of attention and
can improve the environment for manufacturing and supply chains. The benefits
of blockchain now are fascinatingly understood in a variety of areas. The signifi-
cant potential of blockchain in Industry 4.0 is discussed in this article (Javaid et al.,
2021). For perspectives, a number of blockchain technology’s drivers, facilitators,
and related capabilities are examined.
Blockchain is a tamper-proof solution with several uses. The secret of block-
chain (BC) is its tamper-proof design, which gives data authentication reliability.
The principles of blockchain and how it functions in numerous different applica-
tions and technology are the main topics of this paper (Kaushik & Dahiya, 2022).
The readership will also learn more about problems with blockchain technology
integration in this study. The paper will outline the role that blockchain technol-
ogy plays in solving problems related to cutting-edge technologies. This analysis
finds peer-reviewed material that attempts to leverage blockchain for cybersecu-
rity goals and provides a rigorous study of the most often used blockchain security
features. IoT, security, wireless communications, 5G, and even beyond networks
are just a few of the industries where the revolutionary blockchain has a wide
range of applications. This article discusses how blockchain technology will
be used in 5G and other future networks. The many difficulties that arise when
using the blockchain in 5G and future networks are also highlighted in this article
(Kaushik, 2022). The report also covers how blockchain will be used in 5G net-
works in the future.
To increase the levels of security of the smart applications, numerous security
benchmarks and remedies have been suggested over the years; however, the current
solutions are either based on centralized architectures with single points of failure
or have high communication and computation costs. Additionally, the majority of
the security solutions now available have a narrow focus and do not address scal-
ability, resilience, storage systems, network latency, traceability, data integrity, and
provenance. Blockchain technology may provide a remedy for the aforementioned
problems. These facts served as the impetus for the rigorous assessment of various
blockchain-based technologies and their suitability for various Industry 4.0-based
applications that we offer in this article (Bodkhe et al., 2020).
The fourth industrial revolution, also referred as Industry 4.0, is on the horizon,
and as a result, new disruptive innovations are being taken into account for integra-
tion in the manufacturing setting. A few of these options is the blockchain, which
attempts to handle business transactions, connect disparate systems, and promote
asset traceability. As a result, this innovation helps to build an efficient supply chain
that might have an influence on the worldwide market. The intersection between
blockchain and Industry 4.0 is revealed in this article (Silva et al., 2020). In order to
provide instances of what is now being suggested, promoted, and produced, scien-
tific, corporate, and governmental efforts are investigated in this manner.
Innovative upcoming technologies like the 5G mobile network and blockchain
can fill these demands. In contrast to blockchain’s novel data-sharing method of
operation, which improves the high degree of security, visibility, and trustworthiness
238 Intelligent Analytics for Industry 4.0 Applications

of stored data, 5G will enable extraordinarily large channel capacity and lower data
latency. As a result (Jovović et al., 2019), this study provides a broad overview of
the prospective use cases for the 5G network and blockchain technology in Industry
4.0. The findings might help early adopter businesses acquire market technical
dominance.
The growth (Lypnytskyi, 2019) of blockchain’s decentralized versions is most
consistent with the idea of Industry 4.0, according to research on the technology’s
progress. Nevertheless, as demonstrated in the article, internal conflicts caused by
the blockchain “trilemma” and ineffective energy utilization make it difficult to
apply such variants in the industrial setting. There are ways to break the “blockchain
trilemma” deadlock, along with some cutting-edge, contemporary alternatives like
the Level-2 standard. Blockchain technology (BCT) can significantly contribute to
industrial sustainability by safeguarding resources, protecting the environment, and
improving citizen quality of life. In this paper, a four-class taxonomy of the most
significant cyberattacks in Industry 4.0 during the past ten years is offered. Industry
4.0’s most significant BCT-based works are compared in terms of their secrecy,
integrity, scalability, security, and multifactor identification capabilities. Our analy-
sis demonstrates that BCT integration in the industry can guarantee integrity and
confidentiality and should be mandated to maintain data availability and confiden-
tiality. Prospective (ElMamy et al., 2020) research topics are outlined with a focus
on implementing BCT in the industrial setting while taking into account machine
learning, 5G systems, and other evolving technologies.
Distribution and decentralization are seen as common architectural elements.
Aoun et al. (2021) aim to broaden people’s perspectives on blockchain technology’s
potential as a liberating instrument for the fourth industrial revolution. We are first
examining the foundations of Industry 4.0, its difficulties, constraints, and opportu-
nities, and then we are looking at ways that blockchain technology might offer new
capabilities and worth to the implementation of Industry 4.0.

16.3 FUTURE RESEARCH SCOPE OF

BLOCKCHAIN IN INDUSTRY 4.0
Blockchain technology plays a crucial role in the Industry 4.0 and has its wide range
of applications. This section of the chapter highlights the future scope of block-
chain in industry 4.0 by comparing the major findings of the latest research work
done in this domain. Table 16.1 shows the major findings of scope of blockchain in
industry 4.0.

16.4 CONCLUSION
The book chapter highlights the protocols and characteristics of blockchain technol-
ogy. The chapter will also enlighten the readers about the challenges and solutions
of blockchain technology in Industry 4.0. Moreover, the chapter also discusses the
future research scope of blockchain technology in Industry 4.0. The chapter will be
helpful for blockchain enthusiasts, blockchain developers, financial market experts,
Ph.D. scholars, researchers, and students.
Role of Blockchain in Industry 4.0 239

TABLE 16.1
Major Findings of Scope of Blockchain in Industry 4.0
Authors Year Major Findings
Alladi et al. 2019 In addition to major commercial blockchain networks in various pertinent
(2019) industries, this report offered the most recent research findings in each of
the connected industrial sectors. The writers talked about the difficulties
that each sector will face in using blockchain.
Swami et al. 2021 The authors of this study made an effort to analyze some of the most recent
(2021) IoT applications in a variety of industries, including sports, smart gadgets,
location monitoring, and blockchain-based security. IoT is still a subject
that is fast expanding and has a lot of room for multidisciplinary study
and research. IoT will continue to grow as sensors are integrated into
more systems.
Hassan Onik 2018 Both a blockchain-based human resource management system algorithm
et al. (2018) and a recruitment system management have been presented. It is clear
from the examination of the case study data that the suggested
methodology has distinct benefits over the current recruitment methods.
Zuo (2020) 2020 In-depth information about blockchain systems, structures, methodologies,
and research problems is presented in this study. Our talks on utilizing
blockchain technology for different implementations of the smart factory
and smart supply chain are guided by the authors’ suggested blockchain
architecture for smart manufacturing.
Leng et al. 2021 This research shows how the literature has investigated various
(2021) cybersecurity-related challenges. Future research paths for blockchain-
secured smart factory are offered based on the insights gleaned from this
analysis, which may serve as a roadmap for research on pressing
cybersecurity issues for reaching intelligence in Industry 4.0.

REFERENCES
Alladi, T., Chamola, V., Parizi, R. M., & Choo, K. K. R. (2019). Blockchain applications for
Industry 4.0 and industrial IoT: A review. IEEE Access, 7, 176935–176951. https://doi.
org/10.1109/ACCESS.2019.2956748
Aoun, A., Ilinca, A., Ghandour, M., & Ibrahim, H. (2021). A review of Industry 4.0 char-
acteristics and challenges, with potential improvements using blockchain technol-
ogy. Computers & Industrial Engineering, 162, 107746. https://doi.org/10.1016/J.
CIE.2021.107746
Bodkhe, U., Tanwar, S., Parekh, K., Khanpara, P., Tyagi, S., Kumar, N., & Alazab, M. (2020).
Blockchain for Industry 4.0: A comprehensive review. IEEE Access, 8, 79764–79800.
https://doi.org/10.1109/ACCESS.2020.2988579
DharDwivedi, A., Singh, R., Kaushik, K., Rao Mukkamala, R., & Alnumay, W. S. (2021).
Blockchain and artificial intelligence for 5G-enabled Internet of Things: Challenges,
opportunities, and solutions. Transactions on Emerging Telecommunications
Technologies, e4329. https://doi.org/10.1002/ETT.4329
ElMamy, S. B., Mrabet, H., Gharbi, H., Jemai, A., & Trentesaux, D. (2020). A survey on
the usage of blockchain technology for cyber-threats in the context of Industry 4.0.
Sustainability 2020, 12(21), 9179. https://doi.org/10.3390/SU12219179
240 Intelligent Analytics for Industry 4.0 Applications

Gamage, H. T. M., Weerasinghe, H. D., & Dias, N. G. J. (2020). A survey on blockchain

technology concepts, applications, and issues. SN Computer Science 2020, 1(2), 1–15.
https://doi.org/10.1007/S42979-020-00123-0
HassanOnik, M. M., Miraz, M. H., & Kim, C. S. (2018). A recruitment and human resource
management technique using blockchain technology for industry 4.0. IET Conference
Publications, 2018(CP747). https://doi.org/10.1049/CP.2018.1371
Javaid, M., Haleem, A., Pratap Singh, R., Khan, S., & Suman, R. (2021). Blockchain technol-
ogy applications for Industry 4.0: A literature-based review. Blockchain: Research and
Applications, 2(4), 100027. https://doi.org/10.1016/J.BCRA.2021.100027
Jovović, I., Husnjak, S., Forenbacher, I., & Maček, S. (2019). Innovative application of 5G and
blockchain technology in Industry 4.0. EAI Endorsed Transactions on Industrial Networks
and Intelligent Systems, 6(18), e4–e4. https://doi.org/10.4108/EAI.28-3-2019.157122
Kaushik, K. (2022). Demystifying blockchain in 5G and beyond technologies. Journal of
Mobile Multimedia, 18(5), 1379–1398. https://doi.org/10.13052/JMM1550-4646.18513
Kaushik, K., & Dahiya, S. (2022). Scope and Challenges of Blockchain Technology. Springer
Nature, 461–473. https://doi.org/10.1007/978-981-16-8248-3_38
Kaushik, K., Dahiya, S., & Sharma, R. (2022). Role of blockchain technology in digital foren-
sics. Blockchain Technology, 235–246. https://doi.org/10.1201/9781003138082-14
Kaushik, K., Dahiya, S., Singh, R., & Dwivedi, A. D. (2020). Role of blockchain in fore-
stalling pandemics. Proceedings -2020 IEEE 17th International Conference on
Mobile Ad Hoc and Smart Systems, MASS 2020, 32–37. https://doi.org/10.1109/
MASS50613.2020.00014
Leng, J., Ye, S., Zhou, M., Zhao, J. L., Liu, Q., Guo, W., Cao, W., & Fu, L. (2021).
Blockchain-secured smart manufacturing in Industry 4.0: A survey. IEEE Transactions
on Systems, Man, and Cybernetics: Systems, 51(1), 237–252. https://doi.org/10.1109/
TSMC.2020.3040789
Lypnytskyi, D. V. (2019). Opportunities and challenges of blockchain in industry 4.0. Economy
of Industry, 1(85), 82–100. https://doi.org/10.15407/ECONINDUSTRY2019.01.082
daSilva, T. B., de Morais, E. S., de Almeida, L. F. F., da Rosa Righi, R., & Alberti, A. M.
(2020). Blockchain and Industry 4.0: Overview, Convergence, and Analysis. Springer,
27–58. https://doi.org/10.1007/978-981-15-1137-0_2
Swami, M., Verma, D., & Vishwakarma, V. P. (2021). Blockchain and industrial Internet of
Things: Applications for Industry 4.0. Advances in Intelligent Systems and Computing,
1164, 279–290. https://doi.org/10.1007/978-981-15-4992-2_27/COVER/
Zheng, Z., Xie, S., Dai, H., Chen, X., & Wang, H. (2017). An overview of blockchain tech-
nology: Architecture, consensus, and future trends. Proceedings - 2017 IEEE 6th
International Congress on Big Data, BigData Congress 2017, 557–564. https://doi.
org/10.1109/BIGDATACONGRESS.2017.85
Zuo, Y. (2020). Making smart manufacturing smarter – A survey on blockchain technology in
Industry 4.0. Enterprise Information Systems, 15(10), 1323–1353. https://doi.org/10.10
80/17517575.2020.1856425
 17 Blockchain and Bitcoin
Security in Industry 4.0
Yadunath Pathak and Praveen Pawar
Visvesvaraya National Institute of Technology

CONTENTS
17.1 Introduction................................................................................................... 241
17.2 Blockchain Architecture................................................................................ 243
17.3 How Bitcoin Works?......................................................................................244
17.4 Challenges in the Bitcoin Implementation: Security Threats........................244
17.4.1 Bitcoin Protocol................................................................................. 245
17.4.2 Bitcoin Services................................................................................. 245
17.4.3 Other Factors..................................................................................... 245
17.5 Rise of Bitcoin Value..................................................................................... 245
17.6 Consensus Algorithms in Blockchain............................................................246
17.6.1 Proof of Work (PoW).........................................................................246
17.6.2 Proof of Stake (PoS).......................................................................... 247
17.6.3 Pure Stake-Based Consensus............................................................. 247
17.6.4 Proof of Elapsed Time (PoET)..........................................................248
17.6.5 Proof of Activity (PoA)......................................................................248
17.6.6 Proof of Burn (PoB)...........................................................................248
17.7 Various Social Blockchain Applications....................................................... 249
17.8 Conclusion..................................................................................................... 250
References............................................................................................................... 250

17.1 INTRODUCTION
Blockchain, which Satoshi Nakamoto created to serve as Bitcoin’s public transaction
log, has gained much attention [1]. Blockchain technology allows for decentralized
transaction processing. It is currently employed in a number of industries, including
financial services, Internet of things (IoT) applications, reputation management, and
others. The security and scalability of blockchain are just two of the problems that
still need to be resolved.
Cryptocurrency is the buzzword in business and academia these days. Bitcoin, a
very popular cryptocurrency, has a market value of over $10 billion [2]. Bitcoin has
a distinctive data storage structure, and transactions are carried out alone. Bitcoin’s
core technology is blockchain. In 2008, the concept of blockchain was developed,
and in 2009, it was realized [1]. Blocks are used to store all transactions on the block-
chain. It is referred to as the public ledger. With the addition of new blocks, the chain

DOI: 10.1201/9781003321149-17 241

242 Intelligent Analytics for Industry 4.0 Applications

grows longer. A distributed consensus mechanism and ­asymmetric ­cryptography

algorithms are used to keep the ledger consistent and ensure user ­security. Blockchain
technology has important aspects of decentralization, auditability, consistency,
and anonymity. By combining all of these features, blockchain reduces costs and
increases efficiency.
Blockchain applications may be utilized in various financial services such as remit-
tance, digital assets, and online payment because blockchain payments can be made
without a middleman [3]. It can be utilized in a variety of domains, ­including security
services [4], public services, smart contracts [5], IoT [6], and reputation ­systems [7].
These fields favor the blockchain in a variety of ways. Ciaian et al. [8] attempted to
make it worldwide money by evaluating Bitcoin’s properties. Kleineberg and Helbing
[9] demonstrated how digital variety could be maintained. In a blockchain, trans-
action information cannot tamper with the original data. Businesses that require a
high level of privacy and security using blockchain to attract customers. It employs
a distributed computing model to avoid the occurrence of a single point of failure.
When a smart contract is activated on the blockchain, it is carried out automati-
cally. Regarding the internet platform, the blockchain has a slew of t­echnological
Zchallenges. Bitcoin is not suited for high-frequency trading because its network can
only handle seven transactions per second. As a result, the larger the block, the more
­storage space is required, resulting in slower network diffusion. Hence, decentraliza-
tion in the blockchain is required. The trade-off between security and size is the most
challenging task. Second, miners can increase their profits using a selfish mining
technique [10]. Miners disguise their blocks to increase their profits.
Blockchain development hinders because it allows for frequent branching. If users
utilize a private key [11] and a public key, privacy leakage might occur in blockchain.
In addition, the current proof-of-stake or proof-of-work consensus algorithm has
several flaws. For example, proof-of-work wastes much electrical energy, whereas
proof-of-stake consensus allows you to get wealthy faster. Wikis, journal articles,
blogs, forum postings, and codes are all examples of blockchain literature. A tech-
nological analysis was conducted for the decentralized digital currencies known as
Bitcoin [12].
Real-time information is required in the Industry 4.0 environment to create a
seamless service and production system. Processing time is the limiting compo-
nent, hence practical applications must be properly taken into account. Blockchain
is the ideal technology to keep records and information because it can overcome
several obstacles. Therefore, the purpose of this study is to determine how block-
chain might fit into Industry 4.0. In order to improve transaction efficiency in the
company, blockchain replenishment is now necessary. Industry 4.0’s sustainable
product lifecycle management is made possible by blockchain technology. This
will thrive and give processes a higher level of safety in production environments
with lower risks. Before organizations adopt blockchain for implementation,
though, there is still a lot of work to be done with the platform. Although the risk
can be reduced, organization ­executives must recognize the positive effects of this
technology on their business because it is always developing. Industry 4.0 involves
a higher level of privacy and faith. The next section presented the architecture of
blockchain.
Blockchain and Bitcoin Security in Industry 4.0 243

17.2  BLOCKCHAIN ARCHITECTURE

In contrast to a single central server, a blockchain is a network of interconnected
computers. Computers in a network must agree with the state of their shared data
and abide by its restrictions. Every blockchain’s block has some effect on the shared
state. A blockchain is a set of interconnected blocks that serves as a public ledger
and records all transactions. Blockchain’s blocks have only one parent block, and the
block header contains the hash of the preceding block. The Ethereum blockchain is
made up of uncle blocks, or the offspring of a block’s ancestors. The first block in
a blockchain without any parents is known as the genesis block [13]. As shown in
Figure 17.1, a block comprises a block body and a block header. The block header is
made up of the following elements:

• Block Version: it specifies the block validation processes to be followed.

• Merkle Tree Root Hash: this is made up of the hash values for all transac-
tions in a block.
• Timestamp: this is a number that represents the current time in seconds.
• n Bits: this is the minimum size of a valid hash block.
• Nonce: this field is four bytes long. It starts at zero and grows in value with
each calculation.
• Parent Block Hash: with a 256-bit hash value, this field points to the previ-
ous block.

Transactions and the transaction counter are found in the block’s body. The size of
each transaction and the size of the block determine the number of transactions in
a block. Asymmetric cryptography is used to validate the transaction authenticity
[14]. Asymmetric cryptography-based digital signatures are employed in unreliable

FIGURE 17.1  Block diagram of a blockchain.

244 Intelligent Analytics for Industry 4.0 Applications

environments. It is important to know how Bitcoin works, which is explained in the

next section.

17.3  HOW BITCOIN WORKS?

Bitcoin is built on the peer-to-peer transaction concept, which means that the owner
only has access to a public address and a private key. It indicates that the transaction
is peer to peer and does not involve a financial institution, and that the timestamp
of transactions is hashed using the network. The hashes are then connected to cre-
ate the blockchain, which is a chain of hashes. The Bitcoin network, which is its
foundation, acts as a permanent record or ledger of all transactions observed there.
Cryptographic data are generated using the SHA-256 method. The private key of the
Bitcoin’s owner is used to sign each transaction. Thus, unauthorized interference is
avoided.
Two items are required for Bitcoin transactions: (i) A Bitcoin address and (ii) A
private key that only the owner knows about it. An asymmetric key pair is formed
by these two things. The public key that makes up the Bitcoin address is made up of
a random string of numbers and letters. User A’s private key is used to sign the input
(transaction source and amount) before using user B’s public key to sign the output,
which is then transmitted to user B as a private message. The Bitcoins are then
transferred to user B’s wallet. To complete a sequence of independent verification,
it takes roughly 10 minutes. The Bitcoin miners will provide these confirmations.
The network’s shared consensus structure provides confidence to Bitcoin transac-
tions through the mining process. Miners are at the heart of the Bitcoin system,
and they’re in charge of ensuring that all transactions on the network are genuine.
A block is added to the blockchain for each new Bitcoin production using the hash
function.
Bitcoin cryptocurrency is the first decentralized, and independent of any central-
ized control (such as a monetary authority). Users perceive decentralization when
the verification of the code used to provide services is transparent. However, its
­decentralized nature has the following disadvantages:

• Every user maintains a public ledger.

• There is no central authority to certify distributed transactions.
• Bitcoin production is anonymous.
• Bitcoin exchange values are dynamic.

With all these noticeable points, the challenges in the Bitcoin implementation are
discussed in the next section.

17.4 CHALLENGES IN THE BITCOIN

IMPLEMENTATION: SECURITY THREATS
The working concept of Bitcoin must first be grasped to better understand the s­ ecurity
threats to Bitcoin. As a result, it has become the focus of modern study and the topic of
in-depth studies [15]. Government legislation and the power of traditional c­ urrencies
Blockchain and Bitcoin Security in Industry 4.0 245

are two more nontechnical elements that influence Bitcoin’s value. This chapter aims
to figure out which of the causes described above will produce s­ ignificant variations
in the Bitcoin currency value.
In 2017, Fraser and Bouridane compiled a list of potential security concerns about
Bitcoin [16] and an analysis of the factors that contributed to the price drop. The
­following are the details:

17.4.1 Bitcoin Protocol
There are always risks as Bitcoin is purely an online commodity. More than 50% of
the Bitcoin network can be taken over by a single person or group of users, thanks to
a security flaw that Bradbury nicknamed the “51 percent attack” [17]. According to
Bonneau et al. [18], the problem of the size of the Bitcoin network (millions of nodes)
is extremely challenging to handle. Even if it is not a major threat, Bitcoin is nonethe-
less subject to it. Moreover, the value of Bitcoin is impacted by public apprehension.

17.4.2 Bitcoin Services
The number of services using Bitcoin increased in tandem with its popularity.
B¨ohme et al. [19] suggested the two public sectors that hold and trade Bitcoins:
digital wallet services and currency exchanges. A digital wallet, according to their
definition, is a computerized representation of the accounts, transactions, and private
keys needed to send or use Bitcoins. In addition, rather than going through the time-
consuming process of mining, people can buy Bitcoin through a Bitcoin exchange
[20]. The Bitcoin system is less susceptible to cyberattacks of any kind than the
aforementioned Bitcoin services.

17.4.3 Other Factors
Other factors that influence the value of Bitcoin include financial decisions and
­government regulations, in addition to security breaches. Because Bitcoin is a form
of currency, any issues with traditional currencies will also apply to Bitcoin. It is
noted that the most influential element was security threats to the Bitcoin protocol
and their research was regarded to be the most effective method for determining the
most important element influencing Bitcoin’s value.

17.5  RISE OF BITCOIN VALUE

Despite the fact that Bitcoin’s value fluctuates frequently, it remains the most ­valuable
cryptocurrency. Bitcoin had a fantastic year in 2019, with more individuals starting
to trust the technology and spend money on it. By the middle of 2019, its worth had
risen to $20,000. Only 21 million Bitcoins are available for use around the world.
With 4 million still to be mined, just 17 million Bitcoins have already been created,
further increasing the value of the cryptocurrency. Bitcoin reached a peak in 2021
of roughly $47,300. Bitcoin and other cryptocurrencies have piqued the curiosity of
countries such as the United States, Japan, and South Korea.
246 Intelligent Analytics for Industry 4.0 Applications

We are aware that blockchain is a decentralized network, that offers security,

anonymity, transparency, and integrity. Despite the absence of a central authority to
confirm and validate the transactions, every blockchain transaction is regarded as
being 100% safe and validated. The consensus protocol, a crucial component of any
blockchain network, is the fundamental mechanism that makes this feasible. In the
next section, consensus algorithms are discussed in detail.

17.6  CONSENSUS ALGORITHMS IN BLOCKCHAIN

The Byzantine Generals Problem transition is brought up in relation to the b­ lockchain.
A party of command generals and some of the Byzantine army encircled the city in
the event of the Byzantine Generals Problem [21]. There were two points of view
among the command generals. Some wanted to flee, while others wanted to assault.
It would, however, collapse if only a few of the generals attacked. As a result, agree-
ment on whether to attack or withdraw must be established. Similarly, obtaining a
consensus in a distributed setting is difficult. As a result, because blockchain is a dis-
tributed environment, it faces the same difficulty. In blockchain, there is no central
node that determines the similarity of ledgers on dispersed nodes. Some protocols
must be followed to ensure the consistency of ledgers among nodes [22]. This section
goes through the many strategies for reaching consensus in blockchain. Proof-based
consensus algorithms are described in this section.

17.6.1  Proof of Work (PoW)

The Bitcoin network employs PoW as a consensus method [1]. In a decentralized
­network, someone must be chosen to keep track of transactions. Although the
­random selection method is the easiest, it is also the most vulnerable. Therefore, a
time-consuming procedure is needed to validate a node for publishing transactions
in order to make sure that it won’t attack the network. The block header’s hash value
is determined by each node. Nonce and miners comprise the head of a block. The
nonce is frequently changed, giving the miners a variety of hash values. That con-
sensus insures the resulting value must be less than or equal to the provided value.
One node will reach the goal value once. A block is broadcast to the other nodes to
validate the hash value. When a new block is verified, the miners upload it to their
own blockchain. The nodes that calculate the hash values are known as miners, and
Its PoW method is referred to as mining in Bitcoin.
When numerous nodes and the right nonce are chosen at almost the same moment,
a valid block is formed in a decentralized network. As seen in Figure 17.2, branches
can be generated in a variety of ways.
It is, however, impossible for two forks to develop branches at the same time.
The chain is legitimate in the PoW protocol if it becomes long after that. The forks
generated by blocks V4 and A4 that were both validated at the same time were con-
sidered. Miners execute the mining until a longer chain is discovered. A4, A5, and
A6 make up the lengthier chain. As a result, the V4 miners move to the longer chain.
In PoW, miners conduct large computations, yet they squander several resources.
Specific PoW protocols with work-related side applications have been developed to
Blockchain and Bitcoin Security in Industry 4.0 247

FIGURE 17.2  Two scenarios of blockchain branches.

compensate for the loss. Primecoin [22], for example, is used in mathematical studies
to find a particular chain of prime integers.
According to the analysis, one of PoW’s drawbacks is that there may be some security
or usage difficulties. As a result, many remedies to these limits are proposed. The speed
at which the blocks are added to the puzzles is increasing daily, raising the difficulty of
the riddles even more. As a result, miners must spend on the gear to be the first to solve
the mystery. Miners with lesser investment levels are unable to meet market demands.
Tromp [23] presented a way for replacing puzzle labor with the Cuckoo Hash Function,
which requires less effort from miners and allows for more accessible block adding.

17.6.2  Proof of Stake (PoS)

PoS is a low-energy alternative to PoW. In a PoS system, miners must show owner-
ship of the currency amount. The people who have a huge quantity of currencies are
less inclined to assault the network. Choosing a candidate based on account balance
is unfair because the richest individual dominates the network. As a result, numerous
ways combining stake size to determine the next block to be forged have been presented.
BlackCoin [24], in particular, uses randomization to forecast the next generation. The
formula takes the lowest hash value and multiplies it by the stake size. Peercoin prefers
a coin-based selection mechanism. In Peercoin [25], the older and larger set of coins
have a better chance of mining the next block. In comparison to PoW, PoS saves energy
and is more efficient. Because mining is free, an attack could occur as a result. In many
blockchains, PoW is used at first, and afterward, PoS is implemented. Ethereum, for
example, is planning to switch from Ethash (a PoW type) to Casper (a PoS type) [26].

17.6.3  Pure Stake-Based Consensus

PoS in its purest form is NextCoin [27]. The bigger the stake a miner has, the more
chances he or she has of mining a block. Consider, “a” number of coins available
248 Intelligent Analytics for Industry 4.0 Applications

to the miner out of total “n” number of coins, then miner has a/n chance of mining.
Another way for locating the miner based on pure stake is proposed. A “follow-the-
satoshi” rule is also implemented. The lowest unit of Bitcoin, the Satoshi value, is
taken into account. The Satoshi index is used as a parameter in the Satoshi method.
The Satoshi index is a number that ranges from zero to the total number of Satoshi.
The Satoshi block has been discovered. These are the incentives for the block’s
­creators, the miners. The last owner of Satoshi is the one who adds the block to the
chain. Satoshi is based on the hash function. The inputs are chosen based on the
chain’s current state. The first input comes in the form of bits from the comb func-
tion. The second input is provided by the current blocks in the chain, and the third
input is a random integer.
Bentov et al. [28] presented another solution to the problem. The block is not
formed by Satoshi during his opportunity in this case. There are three possibilities
to choose Satoshi. If these opportunities are not taken, the owner will be blacklisted
and will not be able to add to the block. The Bentov principle is applied to the PoS
consensus [29] through the follow-the-satoshi approach. It specifies that the leader
is selected at random using the entropy value. This determined value must be safe-
guarded because it would be difficult to replicate the protocol and forecast the value
in order to influence the leader’s election. On the basis of the Bentov election [30], the
entropy value is determined. Epoch is taken, which is a picture of each stakeholder.
Each epoch determines the amount of stake holders who will participate in the coin-
flipping procedure.

17.6.4  Proof of Elapsed Time (PoET)

On permissioned blockchain networks, PoET is one of the best consensus m ­ ethods
[31]. The permissioned network makes decisions based on concepts of voting or
­mining rights. The network as a whole is transparent. Miners are identified before
joining, so the system login is safe. As a result, the winners are chosen only on the
basis of fairness. In this case, an individual’s waiting time restriction is set at random.
A ledger can be assigned to the player who completes their fair share of waiting time
in order to construct a new block. The Intel SGX system [32] is an example of this.

17.6.5  Proof of Activity (PoA)

PoA is a hybrid of PoW and PoS algorithms [33]. It has the advantages of being
more secure against attacks and not being a power-hungry system. Miners solve a
critical puzzle for a payout, similar to the PoW algorithm. Miners are just i­ nterested
in the block templates. A block’s header specifies a set of random stakeholders.
Stakeholders with higher stakes have a better chance of getting blocks approved.

17.6.6  Proof of Burn (PoB)

To secure cryptocurrency, a percentage of the coins are burned in the PoB process.
The eater’s address receives some coins. The coins supplied to the eater’s address
are nonreturnable. Burned coins are recorded in the ledger, making them really
Blockchain and Bitcoin Security in Industry 4.0 249

unspendable. Although there is a loss when coins are burned, the damage is just tem-
porary because the currencies are protected against hackers in the long run. It also
raises the stakes for other coins. This raises the likelihood of mining the upcoming
blocks and the payments [34].

17.7  VARIOUS SOCIAL BLOCKCHAIN APPLICATIONS

• Business: traditional systems are more error-prone and move more slowly.
In order to settle disputes that are stressful, expensive, and time-­consuming,
intermediaries are required. Asset Management: Trade Processing and
Settlement [35] is an example of a business. For cross-border transactions,
the old method is dangerous and costly. Records are kept inefficiently and
with inaccuracies. The encryption of records in blockchain ledgers reduces
these inaccuracies. Insurance claims are another example. These are more
prone to errors because the processors must manually review bogus claims
and process forms. Blockchain allows for risk-free administration and trans-
parency. Money laundering is possible in the worldwide payment sector,
which is also prone to errors. Blockchain startups such as Align Commerce,
Bitspark, and Abra are providing solutions to this challenge. The blockchain
has been combined with a payment application [36].
• Smart Property: any tangible or intangible asset, including real estate titles,
stock certificates, patents, homes, and automobiles, can be equipped with
smart technology. The registration information, as well as the terms of
the property ownership contract, can be kept in a ledger. Smart property
reduces the danger of fraud while also improving efficiency and trust.
• Blockchain IoT: the IoT refers to a network of interconnected objects. The
Internet of things has a variety of effects on the user. When a printer’s
­cartridges run out, for example, they can be ordered instantly from any
shopping website using the printer. Smart appliances, supply chain sensors,
and other smart devices can all benefit from blockchain and IoT [37].
• Smart Contracts: it includes autonomous digital machines with built-in “if
this, then that” (IFTTT) coding. The intermediate makes sure that protocols
are followed in real life. Blockchain eliminates the use of an intermediary
by making all contract information available to users. It has applications in
financial derivatives, insurance premiums, and other related applications.
Additionally, it might be applied to blockchain-based healthcare. Using a
private key, all individual medical records may be kept and safeguarded
on the blockchain. Only those with specific access privileges would have
access. Drug monitoring, healthcare supply management, and test results
can all be tracked using the ledger [38].
• Blockchain Identity: by encrypting data and safeguarding it from scam-
mers, the blockchain secures the user’s identity. A passport is one of the
examples. In 2014, the digital passport was introduced to assist users in
identifying themselves. The blockchain ensures the accuracy of birth,
death, and anniversary data. It encrypts personal information such as birth
and death dates[39].
250 Intelligent Analytics for Industry 4.0 Applications

• Financial Services: the use of blockchain technology, also known as

­distributed ledger technology, to share data in a safe and effective ­manner
may help the banking sector improve business procedures Distributed
Ledger Technology (DLT). As a major mechanism for digital monetary
transactions, blockchain technologies are altering the financial services
sector. Profit pools and company models that are now inefficient will be
exposed to the possibilities of inefficient blockchain financial platforms.
Using blockchain technology, the distributed ledger increases the security
and transparency of back-office activities. From a regulatory and auditing
standpoint, this is preferred. The current capital market infrastructure is
both costly and slow. Intermediates are required. The new blockchain inno-
vations for capital markets are streamlining processes and lowering costs.
• Government: government services can be also improved with the appli-
cations of blockchain. It promotes the openness of government-citizen
­relations. It improves company processes by allowing for safe data ­sharing.
It has a wide range of applications in the public sector. Blockchain tech-
nology can be used by governments to enhance service delivery, cut costs,
do away with red tape, and stop tax fraud. Health care, social security
benefits, and other fields are significantly impacted by the usage of block-
chain in ­government processes. The processes used by the government are
streamlined.

17.8 CONCLUSION
One of the most important aspects of Industry 4.0 is digitization, which allows
businesses to benefit from effectiveness in all areas, from technology and manage-
ment consulting to supply business model and strategies. According to the findings,
security vulnerabilities in Bitcoin services have the greatest impact on Bitcoin’s
value. Bitcoin’s benefits have exceeded its drawbacks. As a result, Bitcoin has gained
acceptance throughout the world, and many countries are now interested in its tech-
nology. According to a Nasdaq estimate, the stock’s predicted value by the end of
the year will be $23,499. Bitcoin is off to a strong start. People would be able to use
their blockchain identities for a variety of tasks, ranging from basic actions to apps,
software, and digital signatures. Blockchain may be the alternative to facilitating
this step by providing a reliable source for high-quality interactional experience and
understanding to smaller businesses. The volatility of Bitcoin value will be reduced
if security breaches consequently weaken the Bitcoin services, hence the protocols
need to be carefully designed.

REFERENCES
1. Satoshi Nakamoto. Bitcoin: A peer-to-peer electronic cash system. Decentralized
Business Review, page 21260, 2008.
2. Garrick Hileman. State of blockchain q1 2016: Blockchain funding overtakes bitcoin.
CoinDesk, New York, NY, May, 11, 2016.
Blockchain and Bitcoin Security in Industry 4.0 251

3. George Foroglou and Anna-Lali Tsilidou. Further applications of the blockchain. In

12th Student Conference on Managerial Science and Technology, volume 9, 2015.
4. Benjamin W. Akins, Jennifer L. Chapman, and Jason M. Gordon. A whole new world:
Income tax considerations of the bitcoin economy. Pittsburgh Tax Review, 12:25, 2014.
5. Ahmed Kosba, Andrew Miller, Elaine Shi, Zikai Wen, and Charalampos Papamanthou.
Hawk: The blockchain model of cryptography and privacy-preserving smart contracts.
In 2016 IEEE Symposium on Security and Privacy (SP), pages 839–858. IEEE, 2016.
6. Yu Zhang and Jiangtao Wen. An IoT electric business model based on the protocol
of bitcoin. In 2015 18th International Conference on Intelligence in Next Generation
Networks, pages 184–191. IEEE, 2015.
7. Mike Sharples and John Domingue. The blockchain and kudos: A distributed system
for educational record, reputation and reward. In European Conference on Technology
Enhanced Learning, pages 490–496. Springer, 2016.
8. Pavel Ciaian, Miroslava Rajcaniova, and d’Artis Kancs. The economics of bitcoin price
formation. Applied Economics, 48(19): 1799–1815, 2016.
9. Kaj-Kolja Kleineberg and Dirk Helbing. A “social bitcoin” could sustain a democratic
digital world. The European Physical Journal Special Topics, 225(17): 3231–3241,
2016.
10. Ittay Eyal and Emin Gun Sirer. Majority is not enough: Bitcoin mining is vulnerable.
In  International Conference on Financial Cryptography and Data Security, pages
436–454. Springer, 2014.
11. Alex Biryukov, Dmitry Khovratovich, and Ivan Pustogarov. Deanonymisation of
­clients in bitcoin p2p network. In Proceedings of the 2014 ACM SIGSAC Conference on
Computer and Communications Security, pages 15–29, 2014.
12. Florian Tschorsch and Björn Scheuermann. Bitcoin and beyond: A technical survey on
decentralized digital currencies. IEEE Communications Surveys & Tutorials, 18(3):
2084–2123, 2016.
13. Vitalik Buterin et al. A next-generation smart contract and decentralized application
platform. White Paper, 3(37): 2-1, 2014.
14. Zibin Zheng, Shaoan Xie, Hong-Ning Dai, Xiangping Chen, and Huaimin Wang.
Blockchain challenges and opportunities: A survey. International Journal of Web and
Grid Services, 14(4): 352–375, 2018.
15. Feroz Ahmad Ahmad, Prashant Kumar, Gulshan Shrivastava, and Med Salim Bouhlel.
Bitcoin: Digital decentralized cryptocurrency. In Handbook of Research on Network
Forensics and Analysis Techniques, pages 395–415. IGI Global, 2018.
16. John Gregor Fraser and Ahmed Bouridane. Have the security flaws surrounding bitcoin
effected the currency’s value? In 2017 Seventh International Conference on Emerging
Security Technologies (EST), pages 50–55. IEEE, 2017.
17. Danny Bradbury. The problem with bitcoin. Computer Fraud & Security, 2013(11):
5–8, 2013.
18. Joseph Bonneau, Andrew Miller, Jeremy Clark, Arvind Narayanan, Joshua A. Kroll,
and Edward W. Felten. Sok: Research perspectives and challenges for bitcoin and
­cryptocurrencies. In 2015 IEEE Symposium on Security and Privacy, pages 104–121.
IEEE, 2015.
19. Rainer Böhme, Nicolas Christin, Benjamin Edelman, and Tyler Moore. Bitcoin:
Economics, technology, and governance. Journal of Economic Perspectives, 29(2):213–
238, 2015.
20. Tyler Moore and Nicolas Christin. Beware the middleman: Empirical analysis of
­bitcoin-exchange risk. In International Conference on Financial Cryptography and
Data Security, pages 25–33. Springer, 2013.
21. Leslie Lamport, Robert Shostak, and Marshall Pease. The byzantine generals problem.
In Concurrency: The Works of Leslie Lamport, pages 203–226. 2019.
252 Intelligent Analytics for Industry 4.0 Applications

22. Christopher Burks. Bitcoin: Breaking bad or breaking barriers? North Carolina Journal
of Law & Technology, 18(5):244, 2017.
23. John Tromp. Cuckoo cycle: A memory-hard proof-of-work system. IACR Cryptology
ePrint Arch., 2014: 59, 2014.
24. Sunny King. Primecoin: Cryptocurrency with prime number proof-of-work. July 7th,
1(6), 2013.
25. Gavin Wood et al. Ethereum: A secure decentralised generalised transaction ledger.
Ethereum Project Yellow Paper, 151(2014): 1–32, 2014.
26. V. Zamfir. Introducing Casper the friendly ghost, Ethereum blog, 2015.
27. Renato P. dos Santos et al. Consensus algorithms: A matter of complexity. In Melanie
Swan, Jason Potts, Soichiro Takagi, Frank Witte and Paolo Tasca (eds) Between Science
and Economics, pages 147–170. World Scientific, 2019.
28. Iddo Bentov, Ariel Gabizon, and Alex Mizrahi. Cryptocurrencies without proof of
work. In International Conference on Financial Cryptography and Data Security,
pages 142–157. Springer, 2016.
29. Aggelos Kiayias, Alexander Russell, Bernardo David, and Roman Oliynykov.
Ouroboros: A provably secure proof-of-stake blockchain protocol. In Annual
International Cryptology Conference, pages 357–388. Springer, 2017.
30. Anamika Chauhan, Om Prakash Malviya, Madhav Verma, and Tejinder Singh Mor.
Blockchain and scalability. In 2018 IEEE International Conference on Software
Quality, Reliability and Security Companion (QRS-C), pages 122–128. IEEE, 2018.
31. Mitar Milutinovic, Warren He, Howard Wu, and Maxinder Kanwal. Proof of luck: An
efficient blockchain consensus protocol. In Proceedings of the 1st Workshop on System
Software for Trusted Execution, pages 1–6. 2016.
32. Marko Vukolić. The quest for scalable blockchain fabric: Proof-of-work vs. BFT
­replication. In International Workshop on Open Problems in Network Security, pages
112–125. Springer, 2015.
33. Fahad Saleh. Blockchain without waste: Proof-of-stake. The Review of Financial
Studies, 34(3): 1156–1190, 2021.
34. Maher Alharby and Aad Van Moorsel. Blockchain-based smart contracts: A systematic
mapping study. arXiv preprint arXiv:1710.06372, 2017.
35. Harish Sukhwani, José M. Martínez, Xiaolin Chang, Kishor S. Trivedi, and Andy
Rindos. Performance modeling of PBFT consensus process for permissioned blockchain
network (hyperledger fabric). In 2017 IEEE 36th Symposium on Reliable Distributed
Systems (SRDS), pages 253–255. IEEE, 2017.
36. Du Mingxiao, Ma Xiaofeng, Zhang Zhe, Wang Xiangwei, and Chen Qijun. A review
on consensus algorithm of blockchain. In 2017 IEEE International Conference on
Systems, Man, and Cybernetics (SMC), pages 2567–2572. IEEE, 2017.
37. Imran Bashir. Mastering Blockchain: Deeper Insights into Decentralization,
Cryptography, Bitcoin, and Popular Blockchain Frameworks. Packt Publishing
Limited, 2017.
38. Aleksey Novikov, Evgeny Gavrikov, Aleksandr Oleynik, Yuriy Zhirnov, and Nikolay
Pestov. Blockchain technologies in managing socioeconomic systems: A study of legal
practice. Revista Inclusiones, pages 452–461, 2020.
39. Tiana Laurence. Blockchain for Dummies. John Wiley & Sons, 2019.
 18 Technology in
Industry 4.0
Rashmy Moray and Amar Patnaik
Symbiosis Institute of Management Studies (SIMS),
Symbiosis International (Deemed) University

CONTENTS
18.1 Introduction................................................................................................... 253
18.2 Evolution of the Industrial Revolutions-from 1st to the 4th.......................... 254
18.2.1 Industry 1.0 | Mechanization.......................................................... 254
18.2.2 Industry 2.0 | Electrification........................................................... 255
18.2.3 Industry 3.0 | Automation............................................................... 255
18.2.4 Industry 4.0 | Cyber-Physical Systems............................................ 256
18.3 Key Technologies in Industry 4.0.................................................................. 257
18.3.1 Robotics.......................................................................................... 257
18.3.2 Industrial-IoT (IIoT)........................................................................ 258
18.3.3 Artificial Intelligence (AI).............................................................. 258
18.3.4 5G Network (5G)............................................................................. 259
18.3.5 Cloud............................................................................................... 259
18.3.6 Cybersecurity..................................................................................260
18.3.7 Big Data Analytics..........................................................................260
18.3.8 Additive Manufacturing.................................................................. 261
18.3.9 Augmented Reality (AR)................................................................ 262
18.3.10  Digital Twin (DT)............................................................................ 262
18.3.11  Blockchain....................................................................................... 263
18.4 Contemporary Scenario................................................................................. 263
18.5 Need and Significance...................................................................................264
18.6 Conclusion.....................................................................................................264
References............................................................................................................... 265

18.1 INTRODUCTION
From the first to the fourth industrial revolution (IR), we have witnessed a substantial
radical renaissance of technological evolution, from steam power-driven locomo-
tives to electrical and digital mechanized output, making the production processes
increasingly complex, automatic, challenging, and sustainable [1]. However, the
aging population and intereconomic competition are some of the driving factors for
the development of efficient technologies such as industrial Internet of things (IIoT)

DOI: 10.1201/9781003321149-18 253

254 Intelligent Analytics for Industry 4.0 Applications

and the cyber-physical systems (CPSs) to cater to the needs of effective systems and
for improved industrial productivity [2].
Historians who studied the magnitude of radical transformations in the 18th and
19th centuries believed that industrial revolution should comprise and be reflected in
the growth of macroeconomic aspects such as gross national product (GNP), gross
domestic product (GDP), manufacturing produce, wealth creation, and production
efficiency.
The manufacturing sector has been the driver of productivity growth by contribut-
ing 16% to the global GDP and playing a significant role in global economic devel-
opment. The history of manufacturing has transformed dramatically, witnessing
immense growth fueled by four big industrial revolutions.

18.2 EVOLUTION OF THE INDUSTRIAL

REVOLUTIONS-FROM 1ST TO THE 4TH
18.2.1 Industry 1.0 | Mechanization
The first industrial revolution (IR1) was characterized by “mechanization” and
took place in the 18th century with the invention of the steam engine. It started
around 1733 with the first cotton mill. Technological change was dominant in the
years 1700–1830, and that period was deemed to have witnessed the IR1 by Thomas
Ashton [3]. The period between the mid-1700s and mid-1800s represents the IR1,
during which the production moved from being human centric to machine centric. It
began in England and later spread across Europe and North America.
Britain recognized the significance of the inflow of technical know-how, striking
the future of the 19th century in spreading technological knowledge and making
microinventions their competitive advantage [4]. Powered by water and steam, orga-
nized production greatly increased the output of manufactured goods like cotton and
textiles. In the late 19th century, Britain could not sustain the technological competi-
tive advantage and lost to America due to the factors like R&D-based endogenous
growth models, availability of human labor at a cheaper cost, and large markets [5,6].
The invention of the steam power-driven locomotive and the design and develop-
ment of railways spurred industrial growth by way of providing cost-effective trans-
portation of various materials and finished goods.
With these developments and inventions, the total factor of productivity improved
marginally as compared to the pre-IR1 era, leading to substantial per-capita GDP
growth. The pace and direction of technological advancement during the IR1 were
fueled by the spirit of mechanical culture that brought science and technology
together on the shop floor to achieve productivity [7].
From 1760s to 1770s, American colonies under British rule were not happy
with the constant pressure of British imperial tax policies, and the lack of colonial
representation led to repeated protests against and boycotting of British goods by
American colonies, leading to the “American War of Independence” from 1775 to
1783.1 Eventually, America won its independence from Britain in 1783.2 However,
after independence, owing to poor domestic production and shortage of labor force,
increased dependency on Britain for import of goods compelled America to build a
Technology in Industry 4.0 255

strong foundation on the indigenization of production of goods for a better economy

[8]. Thus, with this vision, many technological inventions in the United States such as
electricity, internal combustion (IC) engines, alloys and communication technologies
contributed to the second industrial revolution (IR2) in the 19th century.3

18.2.2 Industry 2.0 | Electrification

The period from 1860 to 1914 is marked as the phase of the IR2, which, with the
invention of electricity, witnessed a shift toward large-scale assembly lines with spe-
cialized labor and machines, powered by oil, gas, and electricity. It is popularly called
the American Industrial Revolution that paved the path from a rural to an urban
society. It began in the United States and later spread throughout many other parts
of the world [8]. Samuel Slater is known as the “Father of the American Industrial
Revolution.”4
Electric motors, electric railways, and electric elevators are some of the key inven-
tions of the IR2. American Industrialist Henry Ford, the founder of the Ford Motor
Company, made a mass-produced Model-T on October 1, 1908, which reduced cost
of production and helped Ford Motor make its automobiles available to the middle-
class people, something that was not possible before. Thus, Ford became the most
popular brand of choice for the Americans during this IR2. Development of fully
interchangeable parts and laying the foundation for aviation industry by Wright
Brothers in the United States, were some of the inventions that took place in this
IR2. Invention of the telegraph and the telephone allowed the world to come closer
with the exchange of information quickly over long distances and brought changes
to communications technology. As the output and consistency of goods like automo-
biles, electric bulbs, and diesel engines increased massively, the real GDP of USA
was more than seven times, with the real per-capita product more than double [8].
With the spread of plant production and output, equivalent growth in the cities was
observed [9].
Increase in world population demanded more goods to be consumed, which led
to the growth of global per-capita levels of industrialization. Factories, automobile
vehicles, etc., burned fossil fuel to produce goods to meet the demand. Rise in fuel
consumption, increased the level of CO2 in the atmosphere which became the pri-
mary factor for global warming. This led to the rise in Earth’s temperature caus-
ing an imbalance in global thermodynamics and thus, global warming and climate
change [8], eventually leading to unforeseen natural disasters, and becoming threat
to various actors involved in the ecosystem.
With this concern in mind, came the vision of moving toward a world with
“Planetary Stewardship of Earth’s Ecosystems” and the “Global Interconnectivity,”
which became the motivation for the Industry 3.0 [10].

18.2.3 Industry 3.0 | Automation

The third industrial revolution (IR3) is based on the convergence of communica-
tion and energy, shifting from using coal, oil, gas, and tar sands to using distributed
energy sources such as the sun, wind, geothermal heat under the ground, biomass
256 Intelligent Analytics for Industry 4.0 Applications

garbage, agricultural and forest waste, small hydro, ocean tides, and waves [11].
In the year 1960, the IR3 commenced with the usage of computers and paved the way
through the expansion and conjunction of information and communications technol-
ogy (ICT). For the first time, industrial robots were introduced by General Motors
in 1961, and the first personal computers were seen during the 1980s. This period
witnessed the “dotcom” boom fueled by investments in cyber-tech companies from
around the globe [12].
Driven by technological advancement and the development of IT and automa-
tion, it transformed the manufacturing industries. Recognized as an era of high-level
automation, PLCs (programmable logic controllers) and Industrial Robots, played
an important role in the manufacturing of increasingly complex products, from
automobiles with onboard computers to handphones. China, South Korea, Taiwan,
Hong Kong, and Singapore—with their growing industrialization and being fueled
by exports—were able to achieve higher economic growth during the IR3. One of
the biggest inventions in this era was “3D printing,” a technology that revolutionized
the way the products were made and reduced the manufacturing time of complex
parts. Thus, robotics and 3D printing are important transformational drivers in the
IR3 [12].
Jeremy Rifkin, the architect of “The Third Industrial Revolution,” described the
five energy pillars as (i) shift to renewable energy, (ii) transformation of the building
stock into green micropower plants to collect renewable energies onsite, (iii) distri-
bution of hydrogen and supplementary storage technologies across infrastructures
to store intermittent energies, (iv) the use of Internet technology to transform the
power grid of every continent into an energy Internet that acts just like the Internet,
and (v)  transition of the transport fleet to electric plug-in and fuel-cell vehicles.
Digitalization was set to mark the disruption of small, medium, and large manufac-
turers in this IR3 [13].
However, despite the technological innovation during these three IRs, emerging
and advanced economies alike have experienced a decline in their manufacturing
sectors, which became insignificant for national incomes and growth for the last
couple of decades [14].
Apparently, the fourth industrial revolution (IR4.0) is the current mechanization
of conventional engineering and industrial practices using contemporary smart tech-
nologies to bring the economy back to the growth trajectory [15]. The contemporary
wave of I4.0 is characterized by “CPSs,” which combines the power of ubiquitous
ICT and physical manufacturing systems together to create a positive impact on pro-
ductivity improvement.5

18.2.4 Industry 4.0 | Cyber-Physical Systems

The previous IRs used electronics and IT to automate the industry, but the Industry
4.0 (I4.0), is taking the shape of a digital resurgence in the industry [16] and is
considered to be the extension of the IR3. It is trademarked by the combination
of ­technologies  that are bridging the gap between physical, digital, and biological
spheres [17].
Technology in Industry 4.0 257

The key concepts of I4.0 were issued in the public domain for the first time as a
part of a strategic initiative by the German government in the year 2011 and were
incorporated in the “High-Tech Strategy 2020 Action Plan.” The first three IRs rep-
resent Mechanization, Electrification, and Automation, respectively, followed by the
I4.0 with the introduction of the IIoT, where the machineries, warehousing systems,
and production facilities are poised to connect seamlessly through a cloud network,
forming CPSs [18]. Thus, CPSs are the basis of IR4.0, with socio-technical interac-
tions between the actors and the manufacturing resources [19].
While physical systems in manufacturing generate massive amounts of data, their
connectivity to the digital world enables data collection and analysis to obtain mean-
ingful insights for better visibility and faster decision-making. Thus, analytics is the
key to manufacturers in the I4.0 [20].
The next level of automation and data exchange in manufacturing technologies,
including CPSs, the IIoT, cloud computing, and cognitive computing, are some of the
few key elements that define the “smart factories” in I4.0. Custom-based modular-
ized production with personalized products, a service-oriented approach, a supply
chain with near-real-time visibility, and a connected-ecosystem enabling customer-
to-consumer interaction are some of the goals of I4.0 for providing value (quality
and cost) and speed to the end customers.6 Apparently, Industry 4.0 is evolving at an
exponential rather than at a linear pace, disrupting each and every industry world-
wide and heralding the transition of entire structures of production, management,
and governance [16].

18.3  KEY TECHNOLOGIES IN INDUSTRY 4.0

As the I4.0 is characterized by the convergence of the physical and digital worlds,
technology plays an important role in the transformation of businesses in the
21st ­century. Some of the key technologies that will drive I4.0 are briefly described
below.

18.3.1 Robotics
Developments in mechatronics, computing, and ICT have given birth to the contem-
porary field of robotics and autonomous systems (RAS) [21]. Known as collaborative
robots (COBOTs), these smart robots interact with the environment through mod-
els and perform various tasks autonomously to optimize production efficiency and
maximize industrial gains/profits [22]. Robots possess clear advantages in terms of
speed, quality, and strength over manual labor. Industrial robots are used for material
handling, which automates tedious, dull, repetitive, and often unsafe tasks at produc-
tion lines. Enriching industrial robots with technologies like artificial intelligence
(AI) and computer vision enables the completion of complex tasks.6 With the help
of AI, computer vision, and haptic-sensing, industrial robots will have capabilities
similar to those of humans to interact, manipulate, train, and identify objects that are
required to separately tailor the products. Thus, the new generation COBOTs would
enhance the impact of intelligent automation by maximizing the abilities of both
humans and machines together [21].
258 Intelligent Analytics for Industry 4.0 Applications

18.3.2 Industrial-IoT (IIoT)
IIoT refers to the technology and application of connecting devices and systems to
exchange data and collaboratively implement services. It has been estimated that
by 2030, over 50 billion IoT devices will be connected to the Internet.7 In IIoT, sen-
sors embedded in the machines continuously generate real-time data that provides
manufacturing companies useful insights about the machines such as condition-
monitoring, efficiency-analysis, and machine-utilization. Using appropriate services,
networking technologies, applications, sensors, software, middleware, and storage
systems, IIoT offers solutions and functions that develop insights to improve the
capability of monitoring and controlling organizations’ procedures and assets [23].
Additionally, through various interconnected devices and more centralized control-
lers, it provides scope for decentralized analytics and decision-making process to
make real-time responses and reactions much faster [24]. This results in better avail-
ability and maintainability of enterprise assets, improved operational efficiency,
faster time-to-market with reduced unplanned downtime, improved production effi-
ciency, and exceptional levels of economic growth [25].

18.3.3 Artificial Intelligence (AI)

AI is the discipline of study in which machines, with the help of data from various
industrial physical systems, IT systems and governed by certain algorithms, mimic
human-like learning such as reasoning and self-correction within an interconnected
digital ecosystem [26]. The subsets of AI comprising machine learning (ML) and
deep learning (DL) are the technologies that help manufacturers with quality opti-
mization, industrial automation, and adaptation capabilities of robots for improving
their precision handling via tactile sensors [27].
ML is the ability of a machine to learn by itself without being explicitly pro-
grammed. Data, being the key requirement in ML, and collected from various
sources of plants and machineries through IIoT technology, undergoes the process of
data engineering to make it suitable for ML. Once the data are ready, with the help
of relevant algorithms such as support vector machines (SVMs), k-nearest neighbor
(KNN), and neural networks (NNs), the machine learns from the data and creates a
model in the form of a relationship between the dependent and independent variables
defined in the dataset. This model is then used for predicting the output from a given
set of input parameters. This type of learning is often called statistical learning.
Depending on the ML problems, there are three types of ML: supervised learn-
ing, unsupervised learning, and reinforcement learning. What a machine actually
learns is derived from these three techniques of ML [28]. Machinery failure predic-
tion, energy consumption optimization, and inventory optimization are some of the
examples of ML use cases in manufacturing.8
DL is the next level of ML, which uses the NN architecture for learning. The sim-
plest form of an artificial neural network (ANN) will have one input layer, two hid-
den layers, and one output layer. An ANN with more than one hidden layer is often
called a deep neural network (DNN). DL involves DNN models with the NN, which
has many layers such as an input layer, hidden layers, and output layers. The learning
Technology in Industry 4.0 259

comes from these layers with built-in forward and backward propagation functions,
which train the model to perform specific tasks as per the requirement [29]. One area
of research in DNN is in the field of natural language processing (NLP), which uses
two main types of architectures: convolutional neural network (CNN) and recur-
rent neural network (RNN) [30]. DL uses manufacturing data for product quality
inspection, fault diagnosis, anomaly detection, etc. and offers insightful information
to various stakeholders in manufacturing organizations, for making better decisions
and manufacturing more productive [31].
Today, AI is considered to be the driving force behind I4.0 across both discrete
and continuous manufacturing industries, thus giving rise to the intelligent factories
where humans and CPSs interact in the cloud9 with telecom technologies such as 2G,
3G, 4G, and the evolving 5G, as the basis for network communication.

18.3.4  5G Network (5G)

5G is the fifth-generation mobile network, set to be the new global wireless standard.
It has the potential to become the future communication platform of choice for many
manufacturing industries, driving the future of I4.0 and smart manufacturing [32].
Ultrareliable low-latency communication, higher bandwidth, and support for higher
device density, network slicing for virtual separation of networks, and high reliabil-
ity are some of the key features that 5G provides the manufacturers to support their
manufacturing mission-critical applications.10 5G is posed to set as the foundational
platform for other I4.0 technologies such as AI, automation, augmented reality (AR),
and IIoT.
The effect of 5G on I4.0 will be distinctive as its features provide solutions that
are essential for manufacturing and helping manufacturers create fully integrated
collaborative smart manufacturing systems that respond in the near-real time to meet
the dynamics and environments in the plant and supply-chain networks and client
requirements.11

18.3.5  Cloud
Cloud is an infrastructure-as-a-service (IaaS) platform with computational capa-
bilities such as storage, databases, servers, networking, software, analytics, and
­intelligence, which are offered as-a-service over-the-Internet or “the cloud.” Cloud
allows companies to tap into computing resources and services as a utility instead
of building and maintaining them in-house [33]. It offers rapid innovation, flexi-
ble resources, and economies-of-scale for various business entities and individuals
involved in the ecosystem of I4.0.12 Today, multinational companies have globally
distributed manufacturing setups and an increased requirement for data sharing for
more production-related undertakings. As cloud technologies evolve over time, they
will provide more opportunities to manufacturers for production system-related data-
driven services by leveraging the functionality and machine data deployed on the
cloud [34].
Cloud-based design and manufacturing (CBDM) is a term used to represent the
process of product realization, which includes design and manufacturing resources
260 Intelligent Analytics for Industry 4.0 Applications

into the model of cloud computing. Networked manufacturing, scalability, agility,

ubiquitous access, multitenancy and virtualization, big data and IoT, IaaS, platform-
as-a-service (PaaS), and software-as-a-service (SaaS) are some of the characteristics
of CBDM [35].
As cloud services evolve, manufacturing companies around the world are looking
to the cloud as a long-term investment in infrastructure and are the biggest adopt-
ers of hybrid cloud solutions (on-premises + cloud) to get a competitive edge in the
market.13
Amazon Web Services, Microsoft Azure Cloud, and Google Cloud are the
key cloud services being offered by the respective cloud players across the globe.
Though cloud has got competitive advantages for the manufacturing industries, it
poses challenges in terms of cyberattacks. This is where “cybersecurity” comes in
as a technology discipline that plays an important role in protecting the plants and
machinery-control systems from unauthorized access.

18.3.6  Cybersecurity
In the interconnected I4.0 realm, ensuring privacy and data security is the key to
prevent cyber crime and cyber terrorism, which could disrupt the system and take it
down when everyone is connected. This is no exception for manufacturing as it gets
more digitized and adopts automation.14 Thus, the existence of IIoT and the demand
for the integration of information technology (IT) and operational technology (OT),
has dramatically changed the entrance of cyber threats. Security threats and vulner-
abilities of IIoT challenges are the vital reasons for cyber attacks [36]. Cybersecurity
ensures to prevent such threats.
Vulnerabilities, threats, and attacks are the three vital concepts that form the
core of cybersecurity problems. Vulnerability refers to any weakness that can be
easily exploited within computing or networking systems that can lead to possible
threats. A threat refers to any such process that can potentially infringe system’s
security ­policies that lead to an attack [37]. An attack is essentially contemplated to
be an active process that intentionally pursues to breach a system’s security policies.
With  the help of cybersecurity, organizations minimize the effect of attacks that
potentially exploit vulnerabilities and reduce threats within systems to the extent
possible [38].
According to the security software provider McAfee, cyber attacks targeting
manufacturers increased sevenfold in 2020 between January and April, with finan-
cial losses caused by data breaches increasing 270% to $8.4 billion from January to
March year on year.15 This is the reason why cybersecurity in most recent times is
considered to be one of the most popular and demanding technologies of research
interest in the I4.0 world.

18.3.7 Big Data Analytics

Big data refers to a technology of processing large volumes of structured, semi-struc-
tured, and unstructured complex data that gets generated during various manufactur-
ing operational stages that traditional data processing methods cannot handle. It is
Technology in Industry 4.0 261

characterized by three Vs, viz., volume, velocity, and variety, which differentiate it
from traditional data [39]. It entails the process of the collection, transfer, storage,
and analysis of data, for example, in predictive modeling, which supports decision-
making.16 Big data is essential in manufacturing to achieve efficiency gains and to
uncover novel acumens to drive innovation. With the help of big data-based analyt-
ics, manufacturers can discover new data and identify patterns that allow them to
enhance procedures, improve effectiveness of the supply chain, and identify vari-
ables that disturb manufacturing.17
Big data analytics are presently used for many industrial applications. This
includes product lifecycle management [40], process redesigning [41], supply-chain
management, and data analysis of the production systems for monitoring, detection
of anomalies, root cause analysis, and additional knowledge [42].
Improved factory operations and production, product quality, supply chain effi-
ciency, customer experience, and reduced machine downtime are some of the exam-
ples of use cases of big data analytics in manufacturing [43].

18.3.8 Additive Manufacturing
Additive manufacturing, in contrast to subtractive manufacturing, describes the pro-
cess of building physical objects by adding materials rather than removing materials.
It works by using 3D computer-aided design (CAD) models to create 3D-printed
parts by depositing materials layer-by-layer with the help of a 3D printing machine
onto a surface. Materials to make these parts can be of a wide variety of types,
including metallic, ceramic, and polymeric materials, along with combinations in the
form of composites, hybrid, or functionally graded materials. 3D printing is emerg-
ing as a valuable digital manufacturing technology of I4.0 [44]. Initially, it was only a
rapid prototyping technology (RPT) that evolved over a period to advanced additive
manufacturing technology, which fundamentally proposes an enormous opportunity
for manufacturing, from tooling to mass customization across virtually all industries
[45]. Additive manufacturing allows components to be stored in virtual inventories
as 3D design files so that they can be manufactured on demand, a model known as
distributed manufacturing. Such a decentralized approach reduces the cost of trans-
port and simplifies inventory management by storing digital files instead of physical
components.18
Some of the potential benefits associated with additive manufacturing are as fol-
lows: (i) direct translation of design to components, (ii) part generation with greater
customization, (iii) ability to produce complex internal features, (iv) reduced weight
of the parts, (v) substantial reduction in overall product development and manufac-
turing lead time, (vi) involvement of smaller manufacturing operations, (vii) high
scalability, (viii) maximum material utilization with minimal waste, and (ix) capabil-
ity to produce fully functional parts [46].
Although the complete adoption of this technology in an industrial setting requires
reaching a certain level of design maturity, additive manufacturing has the potential
to have a constructive impact on the industrial segment by minimizing production
costs, logistics costs, inventory costs, and costs involved in the advancement and
automation of new products [47].
262 Intelligent Analytics for Industry 4.0 Applications

18.3.9 Augmented Reality (AR)

In I4.0, as increasingly more data are generated and acquired from various sources,
it is no longer unimaginable to think of a factory where not only everything is con-
nected, but it is also viewable and interactive [48]. This is where AR comes in. AR
is a technology that superimposes the virtual world onto the real world to present an
enhanced customer experience [49]. AR helps users to visualize the real surroundings
with the virtual environment embedded in them [50]. AR displays the supplement
view of a real-world environment or objects with computer-generated inputs includ-
ing sound, graphics, and GPS location. AR is used to enhance the ­decision-making
capacity and efficiency of manufacturing stakeholders, typically with ­ handheld
­displays or wearables like glasses.
The applications of AR in assembly, maintenance, and repair are found in manu-
facturing. Instructions are better and more effectively understood if they are obtained
in the form of a superimposed physical-virtual experience rather than in the form
of texts and images. Through AR, a maintenance person might see precisely which
equipment needs servicing, including its potential issues, operation times, date of last
service, and probable sites of failure [51].

18.3.10 Digital Twin (DT)

During the product design and development process, the product definition of a 3D
model is created using a CAD software. Two types of data are mainly included in
a 3D model: the geometric and non-geometric information stored in the specifica-
tion tree. These data are then stored and managed in a product data management
(PDM) software [52]. Once the 3D model is realized, the tooling design, production
manufacturing processes, and even product function testing and verification process
simulations and optimization are performed based on the model. Design functional
proofing is done using computer-aided engineering (CAE), which simulates design
performance, that is, whether or not design meets the functional criteria [53].
Once the design is ready and released for manufacturing and later realized physi-
cally, it then goes under testing, showing the actual behavior of the product func-
tionality when subjected to real-world conditions. Results from the testing are then
verified with the pass and fail criteria. If passed, the product moves to the production
phase, and if fails, it undergoes the design modification until approved for passing
criteria. The above process of product development has physical and digital world
processes involved that are carried out separately. This gap of separation is bridged
by the digital twin technology. It is the mapping technology for the complete lifespan
process of physical systems in the virtual space. Thus, digital twin is a digital repre-
sentation of physical products and the technology that allows the mirroring of real-
time data between physical objects and digital models in the virtual world. Data from
all aspects of the product’s lifecycle are collated in the digital twin and analyzed for
getting insightful information at every step of its lifecycle, such as the product design
stage, production stage, and product service stage, which possibly could lead to per-
formance improvements, productivity improvements, and improvement in customer
satisfaction [53].
Technology in Industry 4.0 263

18.3.11 Blockchain
Blockchain in manufacturing is a distributed digital record of transactions of indi-
vidual records structured as “blocks”, which are linked into a single list, or “chain”,
hence the name blockchain. It is also known as distributed ledger technology [54].
Each transaction added to a blockchain is validated with consensus from the stake-
holders before being added to the blockchain [55].
Through a peer-to-peer network, blockchain enables manufacturers to streamline
operations, gain greater visibility into supply chains, and track assets with a higher
level of precision [56]. Blockchain serves the purpose of tackling the issue of possible
inaccurate and tampered data, which poses the risk of huge threats to the intercon-
nected networks of systems [57].

18.4  CONTEMPORARY SCENARIO

While it has been tough for several manufacturing enterprises to understand what
I4.0 really is and that they are still in the process of evaluating its merits and demer-
its, the adoption of I4.0-related technologies in industrial manufacturing is gaining
momentum with less than 30% adoption today and North American manufacturers
having the largest adoption rates, leading to a huge gap of regional difference across
nations [58]. At the same time, the adoption of I4.0 technologies by the automotive
manufacturing industry at an early stage highlights a vast gap across industry adop-
tions. Enterprises like Siemens, General Electric, and Boeing are considered to be
the leading manufacturers by adopting Industry 4.0.19
While countries all over the world such as the United States, Japan, China, the
United Kingdom, Ireland, Sweden, and Austria have started adopting Industry 4.0
technologies, as per estimation the global I4.0 market is expected to reach 13,90,647
crore rupees by 2023 [59]. Also, according to a report published by the World
Economic Forum, out of the 100 countries and economies included in the assess-
ment, only 25 countries were found well positioned to benefit from I4.0.20 India,
despite being the sixth-largest global manufacturing hub, is behind its global peers
in I4.0 adoption. A substantial fraction of the Indian industrial segment is still in the
postelectrification phase, with the technology usage restricted to the siloed systems
that operate independently of each other. CPS, which is the basis of I4.0, is still at
its budding stage [60]. Also, the micro, small and medium enterprises (MSMEs)
segment is yet to adopt the automation technology, owing to its high cost of imple-
mentation [59].
However, the Government of India’s vision and key focus on achieving its share
of manufacturing target GDP of 25% by 2022 have offered an immense potential to
the Indian Automobile Industry for possible expansion and embracement of the I4.0
technologies to create state-of-the-art manufacturing facilities in line with its global
peers. The Automotive Mission Plan 2016–26 is one such initiative that depicts the
government’s combined vision to grow the automotive sector with technological
maturity, global competitiveness, and a well-established institutional structure. The
goal is to brand India among the top three benchmark locomotive industries across
the globe and boost exports exponentially. With the combined effort of the Indian
264 Intelligent Analytics for Industry 4.0 Applications

Government and the Indian Automotive Industry, India is poised to create enabling
ecosystem which will help India become a world-class manufacturing hub in the
years to come [61]. With an investment from the Boeing Company and with a part-
nership with the Indian Institute of Science’s (IISc) Centre for Product Design and
Manufacturing, India’s first smart factory in Bengaluru, equipped with data exchange
in manufacturing and the Internet of Things (IoT) is being developed in line with the
I4.0 standard framework [62].
However, the overall current scenario of I4.0 adoption readiness challenges in
India can be due to (i) lack of understanding of the technologies, (ii) lack of an
organized approach toward digital transformation, (iii) lack of interest to adopt I4.0
technologies, (iv) consideration of cheap labor in lieu of adopting automation, (v)
a small volume of products making it not lucrative for investment, and (vi) skillset
unavailability to adopt automation [59].

18.5  NEED AND SIGNIFICANCE

As customers’ demands are increasing day by day, it becomes imperative for the
manufacturers to be more agile and responsive to the dynamics of customer pref-
erences by being mass-customization centric rather than being mass-production
­centric. To address this, I4.0 technology solutions provide manufacturers with inno-
vative forms of personalization. Direct customer inputs can be considered during the
product design stage, which will help establishments to progressively manufacture
customized products with faster delivery times and lower costs [63–65].
The COVID-19 pandemic led to a sudden increase in the gap to connect to the
manufacturing world, affecting operational visibility. Despite still trying to adjust to
the new normal, the necessity of digital transformation in manufacturing organiza-
tions has become the key priority for business leaders. The value of I4.0 is suddenly
recognized with the accelerated adoption of I4.0 technology solutions in advanced
and developing economies as it is based on the vision of a fundamental process
of innovation and digital transformation in industrial production.20 Impacted by
COVID-19, significant investments in digital transformation with an enhanced focus
on I4.0 to improve productivity can be witnessed by the automotive industry future-
proofing its technology estate.21
As I4.0 describes a fundamental innovation and transformation process for indus-
trial value creation [19], it is poised to bring about manufacturing revolution all over
the world, as the first three IRs did. This IR4.0 in the 21st century will be differ-
ent as compared to its predecessors as the ecosystem will be more interconnected,
much faster, and more responsive to the changing requirements. Information will
flow throughout the value chain by connecting the producers and the consumers in
this ecosystem. I4.0 will provide tremendous opportunities while creating the need
for a highly trained, skilled, and flexible workforce and a production capacity that
will answer the needs of tomorrow [61,66,67,68].

18.6 CONCLUSION
I4.0 is the convergence of the physical and digital worlds. Intelligent factories of
I4.0 will have machines and products communicating with each other cooperatively,
Technology in Industry 4.0 265

driving production in an efficient way to boost up productivity. Factories of I4.0 will

have raw materials and machines interconnected with each other within an ecosys-
tem of IoT. The vision of the I4.0 is to have smart factories with extremely flexible
individualized and resource-friendly mass production resulting in tailor-made prod-
ucts at relatively reasonable prices. It also means highly flexible mass production that
can be rapidly adapted to market variations because, in the future product lifecycles
will be even shorter. Emerging technologies such as IIoT, cloud, big data analytics,
and cybersecurity are pushing I4.0 forward, which promises to change the way busi-
ness is done.

REFERENCES
1. D. Rodrik, “Premature deindustrialization,” Journal of Economic Growth, vol. 21, no.
1, pp. 1–33, 2015.
2. J. Qin, Y. Liu, and R. Grosvenor, “A categorical framework of manufacturing for
Industry 4.0 and beyond: changeable, agile, reconfigurable & virtual production,” in
Procedia CIRP, 2016.
3. N. Crafts, “The first industrial revolution: A Guided Tour for Growth Economists, The
American Economic Review,” in Papers and Proceedings of the Hundredth and Eighth
Annual Meeting of the American Economic Association, San Francisco, CA, 1996.
4. N. F. R. Crafts and C. K. Harley, “Output growth and the British industrial revolu-
tion:  a  restatement of the Crafts-Harley,” Economic History Review, vol. 45, no. 4,
pp. 703–730, 1992.
5. N. F. R. Crafts, “Exogenous or endogenous growth? The industrial revolution reconsid-
ered,” Journal of Economic History, vol. 54, no. 4, pp. 745–772, 1995.
6. G. Nelson, “Rise and fall of American technological leadership: The postwar era in
historical perspective,” Journal of Economic Literature, vol. 34, no. 4, pp. 1931–1964,
1992.
7. M. Jacob, “Mechanical science of the factory floor,” History of Science, vol. 45, no. 148,
pp. 197–221, 2007.
8. H. K. Mohajan, “The second industrial revolution has brought modern social and eco-
nomic developments,” American Institute of Science, vol. 6, no. 1, pp. 1–14, 2020.
9. S. Kim, “Immigration, industrial revolution and urban growth in the United States,”
1820 1920: Factor Endowments, Technology and Geography, Working Paper 12900,
2007.
10. J. Rifkin, “Ecological, Ushering in a Smart Green Digital Global Economy to Address
Climate Change and Create a More,” WordPress, 2015.
11. T. Waghorn, “http://www.siaf.ch/files/130410-rifkin-2.pdf,” 12 December 2011.
[Online]. Available: http://www.siaf.ch/files/130410-rifkin-2.pdf
12. B. Roberts, “The third industrial revolution: Implications for planning cities and
regions,” 2015.
13. J. Rifkin, “The Third Industrial Revolution,” Spring, pp. 32–34, 15 November 2012.
14. IMF, “https:// www.imf.org/B/media/Files/Publications/WEO/2019/October/English/
text.ashx,” 2019. [Online].
15. J. Taalbi, “Origins and pathways of innovation in the third industrial revolution,”
Industrial and Corporate Change, vol. 28, no. 5, pp. 1125–1148, 2019.
16. Schwab, “https://www.weforum.org/agenda/2016/01/the-fourth-industrial-revolution-
what-it-means- and-how-to-respond/#,” 14 January 2016. [Online]. Available: https://
www.weforum.org.
17. J. M. Xu, “The fourth industrial revolution: opportunities and challenges,” International
Journal of Financial Research, vol. 9, no. 2, pp. 90–95, 2018.
266 Intelligent Analytics for Industry 4.0 Applications

18. H. Drath, “Industrie 4.0: hit or hype?,” IEEE Industrial Electronics Magazine, vol. 8,
no. 2, pp. 56–58, 2014.
19. W. Kagermann, “Securing the future of German manufacturing industry: recommenda-
tions for implementing the strategic initiative,” Final report of the Industry 4.0, Working
Group, European Commission, DG Employment, Social Affairs and Inclusion, 2013.
20. Y. M. Omar, M. Minoufekr, and P. Plapper, “Business analytics in manufacturing:
Current trends, challenges and pathway to market leadership,” Operations Research
Perspectives, vol. 6, no. 1, pp. 100127, 2019.
21. L. Pawar, “https://www.ukras.org/wp-content/uploads/2018/10/UK_RAS_wp_­
manufacturing_web.pdf” 2016. [Online]. Available: https://www.ukras.org.
22. H. P. Karabegovic, “The role of service robots in Industry 4.0—smart automation of
transport,” International Scientific Journal “Industry 4.0”, vol. 4, no. 6, pp. 290–292,
2019.
23. S. Lampropoulos, “Internet of Things in the context of Industry 4.0: An overview,”
International Journal of Entrepreneurial Knowledge, vol. 7, no. 1, pp. 4–19, 2019.
24. D. B. Sen, “Smart factories: a review of situation, and recommendations to accelerate
the evolution process,” in Proceedings of the International Symposium for Production
Research, 2018.
25. A. Gilchrist, “Designing Industrial Internet Systems,” in Industry 4.0: The Industrial
Internet of Things, Apress, pp. 87–118, 2016.
26. Z. Mosterman, “Industry 4.0 as a cyber-physical system study,” Software and Systems
Modeling, vol. 15, no. 1, pp. 17–29, 2015.
27. D. Lee, “Industrial artificial intelligence for industry 4.0-based manufacturing sys-
tems,” Manufacturing Letters, vol. 18, pp. 20–23, 2018.
28. H. D. Wehle, “Machine learning, deep learning, and AI: What’s the difference?,” in
International Conference on Data Scientist Innovation Day, Bruxelles, Belgium, 2017.
29. K. Atkeson, “The Transition to a New Economy after the Second Industrial Revolution,”
Working Paper 606, 2001.
30. W. Yin, K. Kann, M. Yu, and H. Schütze, “Comparative study of CNN and RNN for
natural language processing,” arXiv preprint arXiv:1702.01923, pp. 1–7, 2017.
31. Y. Wang, “Deep learning for smart manufacturing: methods and applications,” Journal
of Manufacturing Systems, vol. 48, pp. 144–156, 2018.
32. D. M. T. N. O’Connell, “Challenges associated with implementing 5G in manufactur-
ing,” Telecom, vol. 1, no. 1, pp. 48–67, 2020.
33. A. Fernández, S. del Río, V. López, A. Bawakid, M. J. del Jesus, J. M. Benítez, and F.
Herrera, “Big Data with Cloud Computing: An insight on the computing environment,
MapReduce, and programming frameworks,” WIRES Data Mining and Knowledge
Discovery, vol. 4, no. 5, pp. 380–409, August 2014.
34. M. Rüßmann, M. Lorenz, P. Gerbert, M. Waldner, P. Engel, M. Harnisch, and J. Justus,
“https://www.bcg.com/publications/2015/ engineered_products_project_business_
industry_4_future_productivity_growth_manufacturing_industries,” 9 April 2015.
[Online]. Available: https://www.bcg.com/.
35. D. Thames, “Software-defined cloud manufacturing for Industry 4.0,” in Procedia
CIRP, 2016.
36. B. Ervural, “Overview of cyber security in the Industry 4.0 Era,” in Managing the
Digital Transformation, 2018, pp. 267–284.
37. Hongmei He et al., “The security challenges in the IoT enabled cyber-physical systems
and opportunities for evolutionary computing & other computational intelligence,” in
IEEE Congress on Evolutionary Computation (CEC), 2016.
38. J. Lane Thames, “Distributed, collaborative and automated cybersecurity infrastruc-
tures for cloud-based design and manufacturing systems,” Cloud-Based Design and
Manufacturing (CBDM), pp. 207–229, 2014.
Technology in Industry 4.0 267

39. B. Purcell, “Big data using cloud computing,” Journal of Technology Research, vol. 5,
no. 8, pp. 1–8, 2014.
40. F. Li, “Big Data in product lifecycle management,” The International Journal of
Advanced Manufacturing Technology, vol. 81, no. 14, p. 667 684, 2015.
41. K. N. Palma-Mendoza, “A business process re-design methodology to support supply
chain integration: application in an airline MRO supply chain,” International Journal
of Information Management, vol. 35, pp. 620–631, 2015.
42. H. T. W. K. Ismail, “Manufacturing process data analysis pipelines: a requirements
analysis and survey,” Journal of Big Data, vol. 6, no. 1, pp. 1–26, 2015.
43. H.-N. Dai, H. Wang, G. Xu, J. Wan, and M. Imran, “Big Data analytics for manufactur-
ing internet of things: opportunities, challenges and enabling technologies,” Enterprise
Information Systems, vol. 14, pp. 1–25, 2015.
44. U. M. Dilberoglu, B. Gharehpapagha, U. Yamana, and M. Dolen, “The role of additive
manufacturing in the era of Industry 4.0,” in 27th International Conference on Flexible
Automation and Intelligent Manufacturing, FAIM2017, Modena, Italy, 2015.
45. J. Butt, “Exploring the Interrelationship between Additive Manufacturing and Industry
4.0,” Designs, vol. 4, no. 13, pp. 1–33, 2020.
46. S. A. M. Tofail, E. P. Koumoulos, A. Bandyopadhyay, S. Bose, L. O’Donoghue, and
C. Charitidis, “Additive manufacturing: scientific and technological challenges, market
uptake and opportunities.,” Materials Today, vol. 21, no. 1, pp. 22–37., 2018.
47. R. Godina, I. Ribeiro, F. Matos, Bruna T. Ferreira, H. Carvalho, and P. Peças, “Impact
assessment of additive manufacturing on sustainable business models in industry 4.0
context,”. Sustainability, vol. 12, no. 17, p. 7066, 2020.
48. F. DePace, “Augmented reality in industry 4.0,” American Journal of Computer Science
and Information Technology, vol. 6, no. 01, pp. 1–7, 2018.
49. P. Fraga-Lamas, T. M. FernáNdez-CaraméS, Ó. Blanco-Novoa, and M. A. Vilar-
Montesinos, “A review on industrial augmented reality systems for the industry 4.0
shipyard,” 2018. doi: 10.1109/ACCESS.2018.2808326
50. V. Paelke, “Augmented reality in the smart factory: Supporting workers in an industry
4.0. environment.,” in Proceedings of the 2014 IEEE emerging technology and factory
automation (ETFA), 2014.
51. C. Narcisa Deac, G. Calin Deac, C. Laurentiu Popa, M. Ghinea and C. Emil Cotet,
“Using  augmented reality in smart manufacturing,” in Annals of DAAAM &
Proceedings, 2017.
52. T. H.-J. Uhlemann, C. Lehmann, and R. Steinhilper, “The digital twin: Realizing the
cyber-physical production system for industry 4.0,” Procedia Cirp, vol. 61, pp. 335–340,
2017.
53. Z. Wang, “Digital twin technology, in Industry 4.0—Impact on Intelligent Logistics
and Manufacturing,” pp. 7–21, 2020.
54. T. Ko, J. Lee, and D. Ryu, “Blockchain technology and manufacturing industry: Real-
time transparency and cost savings,” Sustainability 1, vol. 10, no. 11, p. 4274, 2018.
55. S. Zhang and J.-H. Lee, “Analysis of the main consensus protocols of blockchain,” ICT
Express, vol. 6, pp. 93–97, 2019.
56. A. Vatankhah Barenji, Z. Li, W. M. Wang, G. Q. Huang, and D. A. Guerra-Zubiaga,
“Blockchain-based ubiquitous manufacturing: A secure and reliable cyber-physical
system,” International Journal of Production Research, vol. 58, no. 7, pp. 2200–2221,
2020.
57. T. Kobzan, A. Biendarra, S. Schriegel, T. Herbst, T. Müller, and J. Jasperneite, “Utilizing
blockchain technology in industrial manufacturing with the help of network simula-
tion,” in 2018 IEEE 16th International Conference on Industrial Inf, 2018.
58. A. Schumacher, S. Erol, and W. Sihn, “A maturity model for assessing Industry 4.0
readiness and maturity of manufacturing enterprises,” in Procedia Cirp, 2016.
268 Intelligent Analytics for Industry 4.0 Applications

59. K. Pundir, “The role of Industry 4.0 of small and medium enterprises in Uttar Pradesh,”
ABS International Journal of Management, vol. 6, no. 2, pp. 68–74, 2018.
60. V. V. Jadhav, R. Mahadeokar, and S. D. Bhoite, “The Fourth Industrial Revolution (I4.
0) in India: Challenges & Opportunities,” International Journal of Trend in Scientific
Research and Development, pp. 105–109, 2019. https://doi.org/10.31142/ijtsrd23076
61. S. Chouhan, P. Mehra, and A. Dasot, “India’s readiness for Industry 4.0–A focus on
automotive sector,” Confederation of Indian Industry (CII), Grant Thornton-An
Instinct for Growth, 2017.
62. A. Gautam, “Industry 4.0: The Industrial Revolution and New Concepts for the Factory
of Future,” International Journal for Research in Applied Science and Engineering
Technology (IJRASET), vol. 8, no. 6, pp. 1718–1722, 2020.
63. M. Wang, “Industry 4.0: a way from mass customization to mass personalization pro-
duction,” Advances in Manufacturing, vol. 5, no. 4, pp. 311–320, 2017.
64. J. Greenwood, “The IT revolution and the stock market,” Economic Review, vol. 35, no.
2, pp. 2–13, 1999.
65. J. Greenwood, “The third industrial revolution: technology, productivity, and income
equality,” Economic Review, vol. 35, pp. 2–12, 2001.
66. G. Sarma, “Identities in the future Internet of Things,” Wireless Personal
Communications, vol. 49, no. 3, pp. 353–363, 2009.
67. L. R. D. D. Schaefer, “Distributed collaborative design and manufacture in the cloud—
motivation, infrastructure, and education,” in American Society for Engineering
Education Annual Conference Paper# AC2012–3017, San Antonio, 2012.
68. H. Tuptuk, “Security of smart manufacturing systems,” Journal of Manufacturing
Systems, vol. 47, pp. 93–106, 2018.

NOTES
1 https://history.state.gov/milestones/17761783
2 https://en.wikipedia.org/wiki/American_Revolutionary_War
3 https://en.wikipedia.org/wiki/Second_Industrial_Revolution
4 https://www.pbs.org/wgbh/theymadeamerica/whomade/slater_hi.html8
5 https://www.i-scoop.eu/industry-4-0/
6 h t t p s: // w w w. m a n u f a c t u r i n g t o m o r r ow. c o m /a r t i c l e / 2 016 / 0 7/ r o b o t s - i n -
manufacturing-applications/8333
7 https://www.statista.com/statistics/802690/worldwide-connected-devices-by-access-
technology/
8 https://rapidminer.com/blog/6-ways-machine-learning-revolutionizing-manufacturing/
9 https://nexusintegra.io/artificial-intelligence-the-driving-force-behind-industry-4-0/
10 ht t ps://www.gsma.com /iot /wp - content /uploads/2020/0 4/2020 0 4_GSM A _
SmartManufacturing_Insights_On_How_5G_IoT_Can_Transform_Industry.pdf
11 https://www.information-age.com/5g-is-the-heart-of-industry-4-0-123483152/
12 https://azure.microsoft.com/en-in/overview/what-is-cloud-computing/
13 https://www.loadspring.com/post/is-your-industry-fast-tracking-cloud-adoption/
14 https://www.euroscientist.com/modern-manufacturing-needs-cybersecurity
15 https://asia.nikkei.com/Business/Technology/Honda-and-other-smart-factories-fall-
prey-to-hackers
16 https://www.sas.com/en_in/insights/big-data/what-is-big-data.html
17 https://www.manufacturing.net/operations/article/13228439/using-big-data-
analytics-to-improve-production
18 https://am fg.ai /2019/03/28/industr y- 4 - 0 -7-real-world-examples-of-digital-­
manufacturing-in-action/
Technology in Industry 4.0 269

19 https://iot-analytics.com/industry-4-0-adoption-2020-who-is-ahead/
20 ht t p s://w w w.m ck i n s ey.c om / bu si ne ss -f u nct ion s /op e r a t ion s /ou r-i n sig ht s /
industry-40-reimagining-manufacturing-operations-after-covid-19
21 https://zinnov.com/covid19-and-the-automotive-industry-shifting-gears-on-the-other-
side-of-coronavirus/
 19 Intelligent Analytics in
Cyber-Physical Systems
Manjushree Nayak, Priyanka P. Pratihari,
Sanjana Mahapatra, and Shyam Sundar Pradhan
NIST Institute of Science and Technology (Autonomous),

CONTENTS
19.1 Introduction................................................................................................... 271
19.2 Cyber-Physical System.................................................................................. 272
19.2.1 Smart Connection.............................................................................. 272
19.2.2 Data-To-Information Conversion....................................................... 272
19.2.3 Cyber................................................................................................. 272
19.2.4 Cognition........................................................................................... 273
19.2.5 Configuration..................................................................................... 273
19.3 Internet of Things with the Advancement of Industries................................ 273
19.4 Methodology for Designing CPS-Based Industry 4.0 Systems..................... 274
19.5 Case Study..................................................................................................... 275
19.5.1 Industrial Robot................................................................................. 275
19.5.2 Virtual Battery................................................................................... 275
19.6 Conclusion..................................................................................................... 278
References............................................................................................................... 278

19.1 INTRODUCTION
In past decades, companies focus on the quality products and improvement of their
products. But now in today’s world, enterprises are enforced to maintain competency
and their position in the marketplace [1]. Recent technological advancements in com-
puter science and technology have provided solutions to obtain and transfer colossal
amounts of data from their fast-moving environment. Correspondingly, to handle
such a vast amount of data “big data” came into account. Methods and terms such as
the Internet of Things and interconnected systems are introduced to apply a solution
to the “big data” environment [2].
The concept of total productive maintenance [3] evolved from an organization-
centric focus on quality to a customer-centric focus on value creation and smart
­services. The evolution led to the development of prognostics and health management
(PHM). PHM policies are designed to predict component failures, thus ­minimizing
unexpected system downtimes. The underlying patterns are helping to avoid the
costly failures and downtime of machinery. Near-zero breakdown can be obtained

DOI: 10.1201/9781003321149-19 271

272 Intelligent Analytics for Industry 4.0 Applications

by such a maintenance scheme. To make the process more consistent by adjusting

and tuning processes, these invisibles visible can help.
Research in the field of intelligent maintenance has grown significantly and
­facilitated the development of intelligent maintenance systems. As computer and
information systems progress, sensors, data collection equipment, wireless commu-
nication devices, and remote computing solutions become available. Using predic-
tive analytics, such technologies are becoming the face of modern industries. An
advanced predictive analytics system integrated with communication technology and
machines is known as a cyber-physical system (CPS).
Since the concept came into account CPS has been the ever-expanding t­ erminology
in today’s developing world. It integrates the physical systems with computational
models. This includes process control, energy, transportation, medical devices, mili-
tary, automation, smart structures, etc. An intelligent CPS is capable of providing
self-awareness and self-maintenance. Implementing this predictive analysis with a
decision-making system can give proper service with productivity. The CPS is the
central hub of data management. A critical role is played at the fleet level.

19.2  CYBER-PHYSICAL SYSTEM

The CPS structure consists of five levels which are called 5c architecture [4]. This
provides a guideline for CPS for various applications. This CPS structure consists
of two main components, i.e., (i) streamlining data from physical space into cyber-
space and receiving feedback from cyberspace through advanced connectivity and
(ii) intelligent data analytics that helps to construct cyberspace. The 5c structure
shows the workflow of the CPS system. The 5c structure of CPS consists of smart
connections, data-to-info conversion, cyber, cognition, and configuration.

19.2.1 Smart Connection
Connecting machines and their components to obtain accurate data is the first step to
develop CPS for modern industries. Various devices or sensors are used to acquire a
variety of data and store and transfer it to a central server. At this level, the workflow
has a significant impact on the performance of CPS in the next levels, and it is pos-
sible to discover quality and accuracy through the system.

19.2.2 Data-To-Information Conversion
Analyzing and converting data into information is the core of the architecture. In
recent years, algorithms and techniques for data mining have been developed. A
variety of data sources, from machinery to business management data, will be used
to implement such algorithms.

19.2.3  Cyber
The cyber level plays the role of a central information hub, which collects massive
information from machines in machinery work. To extract additional data, specific
Intelligent Analytics in Cyber-Physical Systems 273

analytics are used that provide a better understanding of individual machines within
a fleet.

19.2.4  Cognition
At this level, users decide by information provided to them through the presentation.
Comparative information and individual machine status can make it easier to deter-
mine the priority of maintenance tasks.

19.2.5  Configuration
This level gives feedback from cyberspace to physical space. A supervisory level
­performs the self-configuration and self-adaptation of machines. By applying correc-
tive and preventive decisions, it serves as a resilience control system.

19.3 INTERNET OF THINGS WITH THE

ADVANCEMENT OF INDUSTRIES
IoT holds a very crucial role in building the foundation for the transformation of
manufacturing using the latest technologies. It is also a communication platform for
the CPS [5]. IoT consists of different interacting smart devices with IP addresses,
which enables each device with a unique IP address.
IoT is very much capable of collecting, sorting, and merging different kind of data
that comes from the different sources within the network. It provides an indepen-
dent way of data management with the capability to get real-time status and access.
IoT helps in the implementation of big data for the conversion of data to a piece of
­valuable information and then that to a knowledge which is then converted to action
via a CPS structure.
The IoT platform can be effectively used for many applications in the CPS and
some of the applications are customer-service management, resource planning,
­managing the total supply chain, etc. These all applications are only made possible
due to the collection and categorization of those massive data that is being c­ ollected
from different sources. IoT collects those data from the machines through the
­controller s­ ignals and from the sensors which are being installed within the ­network
[6].
IoT helps in industrial informatics which includes the location and sensor and
for data mining, and it tells us about the unlabeled data. The data that are collected
instantly are sent to the platform and then it is analyzed to bring out the results in the
form of tables, figures, and projected graphs. The enormous amount of data which
are collected by IoT devices leads to the need for analysis which generally includes
data mining, visualization, normalization, etc. The whole data abstraction process is
made easier due to the help of the new embedded systems. Big data analytics has a
five-layered architecture of a context intelligence platform out of which IoT is a layer
that captures data from the IoT-enabled devices like sensor signals, ­geographical
coordinates, etc. Out of the sensors, a very useful one is that which detects the
temperature and intensity of light in that area. This sensor is important because a
274 Intelligent Analytics for Industry 4.0 Applications

luminous environment affects the productivity of an industry. IoT provides many

such innovative ideas to improve the productivity of the industries. IoT-based CPS
also helps the service providers in many different ways. It is a kind of multitier cloud-
based system with a hierarchical structure including the global and local devices [7].
IoT also plays an important role in industry 4.0 which takes industrialization
to a different level. Industry 4.0 is capable of making production more efficient,
­sustainable, and flexible with the help of developed technologies. Often called the
fourth industrial revolution, Industry 4.0 encompasses a wide range of technologies
and devices and on this basis, Germany is leading others. Industry 4.0 also refers to
the integration of industrial technologies with the internet to make productivity up
to the need with better efficiency and adaptability [8]. Industry 4.0 with IoT is the
future of industrial production [9]. Nine technologies help in the transformation of
industry 4.0 and out of those nine, the industrial Internet of Things (IIoT) has an
important role. The IIoT links all devices and components together with the sensing
sensors to get data from them [10]. The technology of the internet provides us with a
very better approach to the performance of the CPS [11]. Industry 4.0 is an integra-
tion of many technologies like data analysis, data mining, security of the network,
cloud computing, and intelligence. These technologies help to increase the efficiency
of production [12].
Industry 4.0 is mainly implemented in the industries to make the m ­ anufacturing
sector digital. Due to these functionalities, the mobile data network or internet
has become a compulsory element in the production sector. Machine-to-machine
­communication (M2M) is an integral part of IoT technology. It allows the exchange
of info between CPSs [13]. Visual computing is a kind of technology that helps
Industry 4.0 and IIoT systems to make their impact on the manufacturing process.
Using Industry 4.0, it is possible to plan, configure, order, design, manufacture, and
operate according to custom customer criteria. A vital component of the fourth-gen
industrial revolution is the capability to offer good performance-based services for
working on the data in these coming days [14].

19.4 METHODOLOGY FOR DESIGNING CPS-

BASED INDUSTRY 4.0 SYSTEMS
Based on the capacities of digital actual frameworks, a better system can be ­developed
for the applications of fourth generation industrial era in the process of CPS. In a pre-
vious segment, we discussed how inter-linkage provides access to information in a
massive amount. In any case, sole accessibility of information does not make a criti-
cal benefit. Along these lines, a versatile yet strong approach is expected to make due,
arrange and process information for additional investigation by PHM calculations.
This technique must be adequately expansive to genuinely use all the benefits of
digital actual frameworks. In this particular part, we bring forward “Time Machine
Methodology for Digital Physical Systems”. The purpose of this is to integrate acces-
sible information in the ecosystem of the big data, to enable the calculations in PHM
to take place efficiently.
Each and every part of the armada will make some agent Memories Machine
record in the internet. This digital representative concentrates advantageous data from
Intelligent Analytics in Cyber-Physical Systems 275

the pool of accessible information and standardizes it for additional investigation.

Separated data incorporate yet not restricted to execution history, stress and burden,
activity boundaries, framework designs, and support records. A fizzled genuine part
should be eliminated from the parent device and it should not participate in testing
again. The virtual representation of an entity which is also termed as digital-twin,
will remain unaffected by time. These digital-twins bring about consistent collec-
tion of time machine records and subsequently assemble different activity boundaries
from wide scope of indistinguishable parts. Standardization of boundaries, in which
further examination endeavors must be led, guarantees the likeness of time machine
records with one another for indistinguishable parts. Time machine records obey the
various leveled connection of genuine parts also and each digital-twin has access to
the records or data of its ancestor parts with progenitor parts. An environment so rich
in data enriches PHM calculations and enables continuous anticipation with accurate
observing the lines of production. At last, this philosophy brings extreme execution of
digital physical framework right into it for planning an Industry 4.0 production line.

19.5  CASE STUDY

19.5.1 Industrial Robot
This particular section provides a better way for the industrial robot health moni-
toring system. The main goal of this study was to design a method or approach for
monitoring the health and condition of 30 industrial robots that are being used in the
production process. By using torque and speed data to construct a robust prognos-
tic and health management algorithm, we had to use a more complex multiregime
approach due to the different speed ranges within a CPS. Monitoring torque has
become an increasingly popular technique for detecting faults in industrial robots
because of its nature, and most research efforts in this area are focused on deter-
mining torque. Furthermore, the nonlinear relationship between operating speed and
torque makes determining the health of the robot a challenge for PHM algorithms.
The model is used to determine the speed and torque of all robots as well as vari-
ous settings like the ratio of the gears and loads, and also for the measure of pres-
sure gauge, and the products or devices that are being assigned from the production
part’s allotted robots. These configuration parameters help the model standardize
and adjust how data clusters provide a more accurate analysis of operations data.
Consequently, every aspect of the study was performed on the cloud, where cloud
platform was used to keep ta dataset securely, and the algorithms for the monitor-
ing of health, were based on the details and those data which we already kept in our
cloud system. Each instance’s health was calculated using PHM algorithms. The user
was presented with highly structured designs (graphs, charts, and diagrams) via a
web interface (Figure 19.1).

19.5.2 Virtual Battery
An electric or hybrid vehicle is not complete without a battery pack. In addition to
concerns about battery safety, uncertainty about driving range, and reliability and
276 Intelligent Analytics for Industry 4.0 Applications

FIGURE 19.1  Industrial robot monitoring.

life of batteries, it is hard to achieve more widespread adoption of electric vehicles.

Battery technology is becoming increasingly important to meet the demands of
hybrid and electric vehicles have become increasingly popular in recent years due
to challenges and rising demands [15]. Understanding battery performance requires
a comprehensive understanding of its dynamics it is vital to know the conditions
under several factors such as the environment temperature, humidity, driving style,
charging level, and discharge level in rates and roads. A battery model, as a means
for evaluating the health of a battery and predicting its failure in different conditions,
will be necessary in order to achieve this understanding. With the help of such a
battery model, suppliers can understand how changes in design and manufacturing
processes impact battery performance and feed that information back to design and
­manufacturing partners [16]. An overview of an electric vehicle and battery health
monitoring and prediction system is shown in Figure 19.2, including algorithms
for calculating charge state and health, and classifying driving behavior. There is
­currently no battery model that considers the dynamics within a battery, but ­virtually
all focus on individual cell models or battery packages. A battery’s components,
however, are interconnected. It is likely that interactions among all cells, c­ onductors,
Intelligent Analytics in Cyber-Physical Systems 277

FIGURE 19.2  Virtual battery.

BMSs, and environment temperature affect battery performance significantly.

Furthermore, manufacturing methodologies have a strong impact on the application
conditions and behaviors of the battery.
In order to evaluate the impacts of multiregime changes in battery parameters and
inputs on battery functionality, a simulated framework is required that is capable of
executing the functions discussed above.
By combining new battery technology with intelligent and enabling tools, a
­“virtual battery” provides the following solutions:

1. By leveraging prognostic tools, health information can be transformed into

useful information about battery performance, reliability, and readiness.
2. Delivering the right information at the right time through an adequate visu-
alization system. The dashboard may display some information the driver
needs to take immediate action, whereas other forms of information (such
as more detailed diagnostic information) may be of more significance to
maintenance and logistics personnel. Schedules maintenance or replace-
ments at the center.
278 Intelligent Analytics for Industry 4.0 Applications

3. The tether-free communication enables real-time battery information to be

gathered from vehicles and sent to a central logistics center, which can use
GPS data to instruct the driver where to take the vehicle for service.
4. The ability to predict battery behavior based on external conditions (e.g.,
driving style, environment) and internal conditions (e.g., battery type and
age) within new operations conditions, without having to make complex
calculations.
5. Provide the designers with better estimates of the pack’s lifespan expectation
in the specific application as a result of a better hybrid cycle management.
6. Identify faulty packs and assist the battery manufacturers in refining their
designs, manufacturing procedures, or raw material selections.
7. In addition to providing better vehicle monitoring during standard or
extended warranty periods, wireless solutions could be integrated into prog-
nostic tolls to detect faulty systems long before they need repair. The faulty
system could be repaired prior to the usual inspection period. As a result,
such instances may result in lower warranty costs. it may also be possible to
integrate this system into engine management sensors.

19.6 CONCLUSION
This paper examined current trends regarding the implementation of CPSs
in ­ manufacturing industry. A generic architecture for implementing CPS in
­manufacturing is necessary since industrial big data has become a challenge for
­factories. In this article, we discuss the 5C architecture for automating and ­centralizing
data processing, health assessment, and prognostics. This architecture covers all the
steps involved in generating and analyzing data, presenting it to users, and ­supporting
decisions. It is possible to use the health information generating by a system to
­perform higher-level functions such as maintenance scheduling and improve control
to enhance overall system productivity. Using a case study of CNS saw machines, a
brief look at the capabilities of the 5C architecture is demonstrated. The case study
shows the application of the 5C architecture in a manufacturing ­environment for the
processing and management of a fleet of CNC saw machines. In the current state of
integration, the 5C CPS architecture is in its infancy. Advancements in all five levels
of the architecture are therefore realistic. A distributed data management system
with new algorithms for fleet-level analysis of machine performance has significant
potential for advancement at the cyber level.

REFERENCES
1. L. Da Xu (2011) “Enterprise Systems: State-of-the-Art and Future Trends.” Industrial
Informatics.
2. E. Dumbill (2013) “Big Data Is Rocket Fuel.” Big Data.
3. Big Data in Cyber-Physical Systems, Digital Manufacturing and Industry 4.0 Lidong
Wang, Guanghui Wang.
4. Industrial Big Data Analytics and Cyber-Physical Systems for Future Maintenance &
Service Innovation Jay Lee, Hossein Davari Ardakani, Shanhu Yang, Behrad Bagheri.
Intelligent Analytics in Cyber-Physical Systems 279

5. Research on IoT Based Cyber Physical System for Industrial Big Data Analytics C.K.M.
Lee, C.L. Yeung, M.N. Cheng.
6. A. Stork (2015) Visual Computing Challenges of Advanced Manufacturing and
Industrie 4.0. IEEE Computer Graphics and Applications, March/April 21–25.
7. M. Anne and M. Gobble (2014) News and Analysis of the Global Innovation Scene.
Research Technology Management, November/December, 2–3.
8. Audio-Tech Business Book Summaries, Inc. (2015) Industry 4.0 and the U.S.
Manufacturing Renaissance. Trends E-Magazine, June, 4–10.
9. I. R. Anderl (2014) Industrie 4.0- Advanced Engineering of Smart Products and Smart
Production, Technological Innovations in the Product Development. 19th International
Seminar on High Technology, Piracicaba, Brasil, October 9th, 2014, 1–14.
10. I. Singh, N. Al-Mutawaly, and T. Wanyama (2015) Teaching Network Technologies
That Support Industry 4.0. Proc. 2015 Canadian Engineering Education Association
(CEEA15) Conf., CEEA15; Paper 119, McMaster University; May 31–June 3, 1–5.
11. J. Bechtold, C. Lauenstein, A. Kern, and L. Bernhofer (2014) Industry 4.0-The
Capgemini Consulting View. Technical Report, Capgemini Consulting.
12. R. Davies (2015). Industry 4.0 Digitalization for Productivity and Growth, European
Parliamentary Research Service (EPRS), Briefing, 1–10.
13. J. Lee, B. Bagheri, and H. A. Kao. (2015) “A Cyber-Physical Systems Architecture for
Industry 4.0-Based Manufacturing Systems.” Manufacturing Letters, 3: 18–23.
14. F. T. S. Chan, H. C. W. Lau, R. W. L. Ip, H. K. Chan, and S. Kong. (2005) “Implementation
of Total Productive Maintenance: A Case Study.” International Journal of Production
Economics, 95(1): 71–94.
15. F. Ju, J. Wang, J. Li, and G. Xiao. (15 January 2013) “Virtual Battery: A Battery
Simulation Framework for Electric Vehicles.” IEEE Transactions on Automation
Science and Engineering, 10(1): 5.
16. Y. He, W. Liu, and B. J. Koch. (2010) “Battery Algorithm Verification and Development
Using Hardware-in-the-Loop Testing.” Journal of Power Sources, 195: 2969–2974.
20 An Overlook on
Security Challenges
in Industry 4.0
R. Ramya, S. Sharmila Devi, and Y. Adline Jancy
Sri Ramakrishna Engineering College

CONTENTS
20.1  Introduction to Industrial Internet of Things................................................. 281
20.1.1  Security Troubles and Issues in Enforcing the IIoT.......................... 282
20.2  Security Attacks in Field Devices..................................................................284
20.2.1  Securing an IIoT Device.................................................................... 285
20.2.2  A Step-by-Step Security Strategy for IIoT Devices........................... 286
20.3  Security Breaches in Industrial Control Systems........................................... 286
20.3.1  Attacks in ICS Systems...................................................................... 287
20.4  Security Challenges in Communication Networks........................................ 287
20.5  Necessity of Data Protection.......................................................................... 288
20.5.1  Industry 4.0 Cybersecurity................................................................. 288
20.5.2  IIoT and Big Data Analytics (BDA).................................................. 288
20.5.3  IoT Data Privacy Leaking.................................................................. 289
20.5.4  Standard Security Solutions............................................................... 289
20.5.5  Opportunities to Overcome BDA Challenges....................................290
20.6 Conclusion.....................................................................................................290
References...............................................................................................................290

20.1  INTRODUCTION TO INDUSTRIAL INTERNET OF THINGS

The Internet has modified the manner humans communicate, do commercial
­enterprise, and collaborate. Now, the point of interest has switched to attaining the
identical outcomes for machines. Systems builders have spent the previous couple
of years that specialize in integrating sensors, facet nodes, and analytics to create
clever structures that flip operations into high-productiveness settings. The industrial
Internet of things (IIoT) is made up of those interconnected structures [1]. This is the
maximum disruptive business revolution in history, touching regions as numerous
as healthcare, strength, transportation, and production. Engineers in each enterprise
will discover a manner to make use of the brand new abilities created via way of
means of connecting system and methods to more effective compute and ­analytics
abilities within side the coming years. The IIoT is the extension and use of the IoT

DOI: 10.1201/9781003321149-20 281

282 Intelligent Analytics for Industry 4.0 Applications

in business sectors and packages [2]. By that specializing in system-to-system (S2S)

connectivity, huge statistics, and system learning, the IIoT allows industries and
groups to grow their performance and dependability of their operations. The IIoT
consists of business packages along with robotics, clinical gadgets, and software
program-described production methods.
The IIoT is going past the conventional customer electronics and bodily tool
inter-networking that the IoT is understood for. It is prominent via way of means of
the mixing of statistics generation and operational generation. Operational meth-
ods and business management structures are networked with human–machine
interfaces (HMIs), supervisory control and data acquisition (SCADA) structures,
distributed control systems (DCSs), and programmable logic controllers (PLCs).
Industry can take advantage of more machine integration in phrases of automation
and o­ ptimization, in addition to better visibility of the delivery chain and logistics,
way to the convergence of information technology (IT) and operational technology
(OT). Smart sensors and actuators make it less complicated in capturing data and
make robust infrastructure in industries, extending the same to other domains like
agriculture, healthcare, production, and transportation.
As a part of the fourth business revolution, referred to as Industry 4.0 [3], the IIoT
is essential to how cyber body structures and manufacturing methods will adapt
with using huge statistics and analytics. Industrial system and infrastructures use
actual-time statistics from sensors and different reassets to resource in “decision-
making,” permitting them to produce insights and carry out unique actions. The
IIoT is vital in a broader feel to be used instances regarding networked ecosystems
or surrounds, along with how towns and factories grow to be clever towns and clever
factories. The normal series and transmission of statistics through clever gadgets and
systems present many possibilities for sectors and groups to grow. The IIoT also can
be used to optimize asset use, expect failure locations, or even set off upkeep sports
on their own. Businesses can also additionally acquire and examine large quantities
of ­statistics at quicker speeds via way of means of the usage of linked and clever gad-
gets. This will assist to bridge the space among manufacturing and standard offices,
in addition to growth scalability and performance.
IIoT is a subset of sensors, automation, machine-to-machine (M2M), e­ mbedded
systems, and industrial controls. They follow IT protocols rather than OT p­ rotocols
where personnel is trained in IT. Connected with trillions of devices, IIoT has
an umbrella of nontraditional systems with a combo of standardless protocols,
remote management, and authorized/nonauthorized devices. Evaluation of variable
­environments in IIoT leads to security breaches. The chapter focuses on various
attacks faced by target entities like devices, DCSs, and communication networks
(Figure 20.1).

20.1.1 Security Troubles and Issues in Enforcing the IIoT

The adoption of business IoT technology has significantly benefited companies
within side the production and transportation sectors (IIoT). Since IIoT structures
have grown to be extensively used, significantly in safety, formerly ­insurmountable
An Overlook on Security Challenges in Industry 4.0 283

FIGURE 20.1  Various security attacks in IIoT.

t­echnical problems were without problems overcome. Predictive protection, for

instance, can assist a water remedy commercial enterprise with the aid of using
­placing sensors on its machinery. The sensors acquire overall performance data,
examine it, and flag ability dangers, permitting the commercial enterprise to assume
protection desires and run extra effectively. Despite the plain tangible welfare of
extra virtual connectivity, IIoT deployment does offer a few IoT safety worries and
challenges.
To sell industry-huge standards, the World Economic Forum (WEF) has p­ repared
an insightful observation on IIoT protection and safety fine practices. Here are
four IoT safety-demanding situations noted within side the notice that producers
should grapple with. One of the maximum large issues [4] is while integrity ­legacy
­software program with IIoT systems is a loss of maintenance. While checking
out legacy ­software program is one challenge, it is far vital that IIoT gadgets (and
related ­software programs) are examined carefully in the course of their lifespan.
Methodologies need to include unit, system, acceptance, regression checking out,
and danger modeling, together with retaining a stock of the supply for any 1/3 party/
open supply code and additives utilized. As greater gadgets hook up with a shared
community, it will become tougher for IT groups to discover risks. As the WEF
284 Intelligent Analytics for Industry 4.0 Applications

suggests, IIoT deployment groups should construct modules that outline all of the
virtual and bodily belongings that want protection. They additionally want to higher
accumulate records on capacity threats.
The configuration of hardware additives is a hugely unregulated sphere that calls
for near interest while constructing an IIoT system. It all comes right all the way
down to endpoint safety, namely, stopping unauthorized get admission to customers
who make out-of-control modifications to the hardware. The major work for an IIoT
deployment group is assembly encryption standards. Every interplay with the system
(from any device) should undergo usually regularly occurring cryptography proto-
cols. The extent of private statistics being shared each day throughout the community
will most effective upward thrust withinside the years to come – making encryption
one of the maximum essential measures from a privacy perspective.

20.2  SECURITY ATTACKS IN FIELD DEVICES

The IIoT quarter is unexpectedly expanding. By the cease of 2015, the worldwide
marketplace have been saturated with an envisioned 134 billion IoT-linked devices.
By 2020, the variety of connected gadgets is anticipated to attain 38.5 billion. Because
maximum business methods are not completely automated, they may be capital and
labor demand, and they require a powerful method to find out community troubles
and beautify performance, many groups are trying to broaden IIoT gadgets. IIoT
gadgets are appealing due to the fact they could growth protection, dependability,
and strength performance. These gadgets encompass sensors that accumulate statis-
tics in actual time or over the years and connect with statistics analytics and manage
structures. In order for utilities to prioritize community upkeep, it is far crucial to
examine statistics traits on the way to supply actionable insights. Data styles should
be analyzed so as for utilities to offer significant insights and prioritize community
enhancements or screen delivery call for mismatches.
To lay out protection, it is far vital to apprehend from which the ­capability
threats arise. Attacks may be released over all community links: via employer
­connections, connections via different networks on the manage community layer,
and/or ­connections at the sector tool degree [5]. The Stuxnet assault may also had
been the ­riding aspect that made protection researchers aware of vulnerabilities in
area ­gadgets themselves. According to many sources, Stuxnet’s goal changed into
an Iranian ­uranium enrichment plant. Its purpose changed into to adjust this sys-
tem jogging on particular PLCs to alternate centrifuge rotor speeds, inflicting harm
to the rotors, at the same time as reporting fake values to cover the adjustments
from machine operators. Stuxnet used more than one exploit to unfold to computer
­systems over networks and via thumb drives. Once Stuxnet inflamed a pc jogging a
Siemens PLC configuration tool, it checked the configuration to peer if it had reached
its goal, then changed the good judgment programming on particular PLCs in that
machine. The complexity of the assault changed into wonderful as it required more
than one skillsets, took a huge quantity of resources, and changed into very centered
and self-limiting. Although the Stuxnet assault changed into extraordinarily com-
plex, the real subversion of the PLCs changed into made trustworthy with the aid of
using their design.
An Overlook on Security Challenges in Industry 4.0 285

Some of the usual place assaults are as follows:

• Eavesdropping assault
By tracking the community, an adversary can benefit from touchy
­statistics approximately the behavior of the community to perpetrate ­similar
assaults. Network site visitors’ analysis, even of encrypted packets, can
screen statistics and compromise privates.
• Man-in-the-center assault
In this assault, the adversary sits among speaking gadgets and relays
conversation among them.
• False statistics injection assault
A fake injection assault is a deception assault, wherein the adversary
injects fake statistics into the community; for example, via way of means of
sending malicious instructions on a subject bus.
• Time postpone assault
An attacker injects more time delays into measurements and manages
values of the structures, which could disturb balance of the machine and
reason system to crash.
• Replay assault
In a replay assault, valid packets may be retransmitted via way of means
of an adversary. This can appear in numerous ways: a true however compro-
mised node may want to ship the statistics.
• Spoofing assault
Spoofing assaults are in which an attacker’s node impersonates a
machine entity. A loss of good enough authentication management mecha-
nisms method that entities can masquerade as each other via way of means
of falsifying their identification to benefit illegitimate get entry to.
• Side channel assaults
Side channel assaults are done the usage of numerous strategies that
examine statistics leakage from hardware and software program along with
analyzing strength consumption, mild emissions, optical signal, site visitors
flow, timings, electromagnetic, acoustic, and thermal emission from hard-
ware additives and faults that arise with inside the machine [6].
• Covert-channel assaults
This is an assault that uses a compromised tool and valid conversation
channels to leak touchy statistics out of a stable environment, bypassing
security features.

20.2.1 Securing an IIoT Device

Ayyeka, as an IIoT commercial enterprise, created a device that consists of present-
day cybersecurity great practices. The essential ranges to shielding IIoT gadgets
and recognizing tool vulnerabilities may be protected on this article. Embedding
cybersecurity into IT layout, safeguarding conversation protocols, and integrating
extra security features for offering statistics to SCADA structures are only some
examples. Installing a firewall and an intrusion detection machine is the maximum
286 Intelligent Analytics for Industry 4.0 Applications

f­undamental degree of safety intrusion detection systems (IDS). Firewalls, on the

alternative hand, may be set to pick out while a suspicious man or woman is ­coming
near their property. As safety toward unauthorized intruders, a firewall is man-
aged via way of means of protection guidelines that adjust incoming and outgoing
­community site visitors.
Encrypting sensitive statistics is the subsequent step. Hackers are nearly positive
to try and intercept conversation from subject gadgets with inside the application
enterprise. As a result, it is crucial to assume like a hacker and find out the communi-
ty’s weakest hyperlink in order that extra security features can be installed place. For
example, all subject tool communications are encrypted with the SCADA machine.

20.2.2 A Step-by-Step Security Strategy for IIoT Devices

To assist corporations to plan their business virtual transformation correctly and
securely, AWS recommends a multilayered method to stable the industrial control
systems (ICS)/OT, IIoT, and cloud environments, which are captured in the follow-
ing steps [7]:

• Conduct a cyber-safety threat evaluation of a not unusual place framework

(which includes MITRE ATT&CK) and use it to tell gadget design.
• Maintain an asset stock of all related properties and updated community
architecture.
• Provision cutting-edge IIoT gadgets and structures with specific identities
and credentials, practice authentication and get admission to manipulate
mechanisms.
• Prioritize and put into effect OT and IIoT precise patch control and out-
line suitable replacement mechanisms for software program and firmware
updates.
• Secure production facts at the threshold and within the cloud through
encrypting facts at relaxation and create mechanisms for stable facts shar-
ing, governance, and sovereignty.

20.3  SECURITY BREACHES IN INDUSTRIAL CONTROL SYSTEMS

In today’s Internet-connected society, concerns about more prominent cyber attacks.
Multiple devices, computers, controllers, communications channels, and software are
integrated to communicate and operate industrial processes in the IoT. As a result,
both inside and outside the communication network, the IoT is vulnerable to physical
intrusion and viruses. A hacker with awareness of software, process equipment, and
networks can employ electronic means to really get access to IoT if it is not protected.
The main reasons most industry devices get hacked are as follows [8]:

a. Many plants’ devices go weeks or months without receiving any security

upgrades or antivirus software.
b. Because many of the controllers used in IoT networks were created in an era
when cybersecurity was not a priority, they can be disturbed by malformed
network traffic or even by large volumes of correctly formed traffic.
An Overlook on Security Challenges in Industry 4.0 287

c. Many IoT networks have several points of entry for cybersecurity threats,
allowing them to overcome existing security measures. Laptops brought
into and out of facilities, as well as USB sticks used by several PCs without
being properly scanned for malware, are common examples.
d. Many IoT networks are still built as a single, big, flat network with no physi-
cal or virtual protection between networks. This aids in the rapid spread of
malware, even to remote locations devices.

20.3.1 Attacks in ICS Systems

Zero-day attack happened in SCADA where the nature of attack is not predicted
[9]. False logic attack refers to giving false command to SCADA and wrong data
is obtained by ICS [10]. Deception is another attack on SCSDA and DCS where
wrong commands are issued. The operators are subjected to reconnected mean
while the hacker issues the incorrect command to the system. The Stuxnet attack
is a major attack on PLC leading to misbehavior of it and causing the controller to
reprogram [11].

20.4  SECURITY CHALLENGES IN COMMUNICATION NETWORKS

The IoT integrates everything on the Internet and in the telecommunications ­network.
But it also introduces a special set of challenges for the development of technology
due to the recent revolution in communications and computers. Signaling, security,
authorization, open ports, and encryption are some of the main security issues affect-
ing IoT connectivity, as follows [12]:

Signaling: bidirectional strong signal is necessary when IoT devices

a.
are connected in order to collect and route data between smart devices
effectively, as the acquired data must move fast from point A to point B.
Manufacturing message specification (MMS) packet modification attack is
done while communication between SCADA and control center [13].
Security: developers must come up with some innovative solutions because
b.
the majority of smart devices do not have the basic necessities of hacker
security. Security measures should offer enough privacy and security for
end users for a constant flow of data moves between devices, networks,
and gateways when millions of smart devices are interconnected. Collision
attack leads to rearrange the checksum in the message frame. Fake identi-
ties of nodes are generated in Sybil attack. Node replication happens by
cloning the sent nodes [14].
Authorization: as there are more connected devices, the IoT will produce
c.
and store an increasing amount of personal data, making it more attractive
to hackers and increasing the risk of security breaches. We must confirm
that the server or smart device being used has the appropriate authorization
to send and receive data before sending or receiving any data.
Open ports: the issue is that bidirectional connection is required in order
d.
to send and receive data, but open ports must not exist because an open port
exposes a system open to attack. Wormhole attack makes the node to be in
288 Intelligent Analytics for Industry 4.0 Applications

a perception of it is in the right routing direction. Sinkhole attack makes to

transmit fake data to neighbor node [14].
Encryption: The use of encrypted data transmission services by 70% of
e.
smart devices makes personal information accessible. It requires end-to-end
encryption between our smart devices and the servers being used in order to
send and receive any data secure manner.
Although there is no dispute that the IoT is transforming industries and
improving everyone’s consumer experience, security and data misuse are
perennial concerns. There will be circumstances at which data may be
misused because there is so much key information traveling around. The
industries are still implementing the IoT, and it will take some time before
users begin to notice and explore with the changes. The IoT concept relies
on device communication, and smart things can communicate with one
another using network technology.

20.5  NECESSITY OF DATA PROTECTION

The recent expansion of information and communication technology (ICT) has
offered significant economic and technological prospects. Meanwhile, this has
become the hornet’s nest of cybercrime, allowing for the theft of personal informa-
tion and identities, further more eavesdropping on people’s professional and private
life and other connected difficulties. Another significant challenge is that IoT, which
links a multitude of devices for a two-way flow of information, and captures and
archives zillions of personal data that can be exploited or misused by third parties.

20.5.1 Industry 4.0 Cybersecurity

The handling of personal data belonging to employees and consumers poses the
greatest danger in a smart factory. Personal data may be captured, analyzed, and in
certain cases transmitted combined with other data in any scenario of networking
between humans, manufacturing unit, and logistical systems. Pellucid for employees
and consumers about the technology being used, as well as appropriately qualified
data protection officers who monitor data acquisition and processing in the context
of Industry 4.0, are essential first steps toward a data-protection-compliant embrace
of Industry 4.0.

20.5.2  IIoT and Big Data Analytics (BDA)

Enormous volumes of data are generated by IIoT systems. In industrial systems, data
are generated through smart sensor networks, devices, log files, and other sources.
Inevitably, with so many distinct and varied data sources, it is likely that the data will
be highly diverse and structured, unstructured, or both. Volume, reliability, velocity,
diversity, validity, and volatility are all characteristics of big data. As a consequence
of the partnership with the IoT and cyber-physical systems (CPS), these produc-
tion systems generate a tremendous amount of data, a wide diversity of data, and a
high level of veracity. Metadata are collected from manufacturers, vendors, product
An Overlook on Security Challenges in Industry 4.0 289

designers, clientele, and consumers for BDA. Despite its unrivaled significance in
furthering Industry 4.0, big data’s distributed structure poses a number of issues.
Data security, privacy, and access control, as well as data storage, are among the most
important concerns.
Manufacturers can use BDA in manufacturing to discover the most up-to-date
information and recognize patterns, allowing them to improve processes, regulate
supply chain, and identify variables that influence production. According to reports,
BDA in the manufacturing industry would approach $4.55 billion by 2025 [15].
Intelligent manufacturing ship floors are constructed using BDA, which represents
tangible Internet-based logistics data. To collect data, they used radio-frequency
identification readers, tags, and wireless communication networks on the factory
floors. They used this information to create a visualization of the logistics chain and
to assess the efficiency of logistics operators and operations [16].

20.5.3 IoT Data Privacy Leaking

Most businesses are increasingly aware of the importance of external security threats.
As a result, we create a security system and a new security solution, as well as handle
the outside security. Internal security concerns have become increasingly relevant
as a result of recent insider data leaks and system sabotage. Data ingress by internal
personnel is 50%, and data leakage by executive officers is up to 20%, reported on a
study done by Websens Security Labs. This study demonstrates that only the outflow
is no longer significant. Finally, the significance of user behavior analysis (UBA)
technology is highlighted [17].
Many companies are investing on UBA and a few are narrated here. IBM uses
analytics-based intelligence, and Netcruz focused on a combined platform of
finding external infringement and internal information control. Through aber-
rant traffic categorization and traffic source inquiry, export management company
(EMC) developed network forensics that can detect and respond to sophisticated
assaults [18].
Typical raw data storage, energy limitations, encryption limitations, lack of stan-
dardized structure, and a few more factors all contribute to privacy leaks [19,20].
Data and intellectual property are disseminated across supply chains and stakehold-
ers with Industry 4.0. Consumers and suppliers are integrating their systems. Data
are dispersed among the systems, implying a broader security scope. As the number
of stakeholders grows, so does the number of entry points that are subject to assault.
Another assault vector is the convergence of information and operational technology.
This becomes harder to develop an end-to-end security system.

20.5.4 Standard Security Solutions

Simple, well-known solutions, such as standard IT systems, access authentication,
access control, and key management, may help to lessen the likelihood of a breach.
Application protection, antimalware, and antivirus are a few examples of standard
information security techniques. Network consumption, traffic, and logs may all be
tracked, as well as network segmentation. Personnel security training is also a viable
290 Intelligent Analytics for Industry 4.0 Applications

option. Patching methods could be useful for keeping industrial systems current and
secure. Regular system audits and penetration testing are also important safeguards.

20.5.5 Opportunities to Overcome BDA Challenges

An analytic team is to be recruited to frame strategies on data analysis in manufac-
turing companies. Standards need to be introduced in data collection, storage, and
transmission from the floor or site. As data obtained from field devices are unstruc-
tured, a synchronized and structured database needs to be established. Data storage
becomes costly as the data are expensive, and as the data processing algorithms works
on large data size, size reduction techniques are being encouraged. A backup system
for data is required when there is a possibility of negative BDA analysis. Finally,
standards need to be improved in legislation policies and information s­ ystems [21].

20.6 CONCLUSION
In this chapter, the primary aim was to explore the current state of the art of security
challenges in Industry 4.0 relating technologies in the manufacturing and automation
industry. In Industry 4.0, industrial revolution is happening and it is accomplished
with smart in production, logistics, storage, maintenance, and control system. On
the other hand, there these smarter things are data hungry and prone to security
attacks. As IIoT is a combination of field devices, control devices, and communica-
tion networks, attackers can target on any of the mentioned areas. Hackers targeting
the field devices like sensors or industrial machines leads to hindrance in receiving
data from the field. Different ICS system attacks are discussed in this chapter. Usage
of dynamic IP for SCADA gives a better solution for ICS attack. In networks attack,
wireless media are more prone compared to wired as the hacker can falsely route
or trap data packets. The final chapter concludes the necessity of data protection, as
IIoT are used in data-driven applications. Industry 4.0 extends its wings to BDA and
cloud storage, and this tells the security breaches in data storage and analysis. UBA
is implemented in leading industries as internal data leakage overcomes the external
data hack.

REFERENCES
1. Matthew N. O. Sadiku, Yonghui Wang, Suxia Cui, and Sarhan M. Musa, “The indus-
trial internet of things,” IJASRE, vol. 3, no. 11, pp. 1–5, Dec. 2017.
2. A. Gilchrist, Industry 4.0: The Industrial Internet of Things. Springer, 2016.
3. Adam Mentsiev, Elina Guzueva, and Tamirlan Magomaev, “Security challenges of the
Industry 4.0,” Journal of Physics: Conference Series, vol. 1515, 032074, 2020.
4. S. R. Chhetri, N. Rashid, S. Faezi, and M. A. Al Faruque, “Security trends and advances
in manufacturing systems in the era of industry 4.0,” 2017 IEEE/ACM International
Conference on Computer-Aided Design (ICCAD), 2017, pp. 1039–1046.
5. Nilufer Tuptuk, and Stephen Hailes, “Security of smart manufacturing systems,”
Journal of Manufacturing Systems, vol. 47, pp. 93–106, 2018.
6. M. A. Al Faruque, S. R. Chhetri, A. Canedo, and J. Wan, “Acoustic side-channel attacks
on additive manufacturing systems,” 2016 ACM/IEEE 7th International Conference on
Cyber-Physical Systems (ICCPS), 2016, pp. 1–10.
An Overlook on Security Challenges in Industry 4.0 291

7. S. Lin, B. Miller, J. Durand, G. Bleakley, A. Chigani, R. Martin, B. Murphy, and

M. Crawford, “The industrial internet of things volume G1: reference architecture”
Industrial Internet Consortium, pp. 10–46, 2017.
8. P. Rizwan, M. R. Babu, B. Balamurugan, and K. Suresh, “Real-time big data comput-
ing for Internet of Things and cyber physical system aided medical devices for better
healthcare,” 2018 Majan International Conference (MIC), pp. 1–8, 2018.
9. M. Lezzi, M. Lazoi, and A. Corallo, “Cybersecurity for industry 4.0 in the current
­literature: A reference framework,” Computers in Industry, vol. 103, pp. 97–110, 2018.
10. W. Li, L. Xie, Z. Deng, and Z. Wang, “False sequential logic attack on scada system and
its physical impact analysis,” Computers & Security, vol. 58, pp. 149–159, 2016.
11. S. Karnouskos, “Stuxnet worm impact on industrial cyber-physical system security,”
IECON 2011–37th Annual Conference of the IEEE Industrial Electronics Society, pp.
4490–4494, IEEE, 2011.
12. W. Yu, T. Dillon, F. Mostafa, W. Rahayu, and Y. Liu, “Implementation of industrial
cyber physical system: challenges and solutions,” 2019 IEEE International Conference
on Industrial Cyber Physical Systems (ICPS), pp. 173–178, 2019.
13. S. Lee, S. Lee, H. Yoo, S. Kwon, and T. Shon, “Design and implementation of cyber-
security testbed for industrial IoT systems,” The Journal of Supercomputing, vol. 74,
no. 9, pp. 4506–4520, 2018.
14. F. Januario, C. Carvalho, A. Cardoso, and P. Gil, “Security challenges in scada sys-
tems over wireless sensor and actuator networks,” 2016 8th International Congress
on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT),
pp. 363–368, IEEE, 2016.
15. Ray Y. Zhong, Chen Xu, Chao Chen, George Q. Huang, “Big data analytics for physi-
cal internet-based intelligent manufacturing shop floors,” International Journal of
Production Research, vol. 55, pp. 2610–2621, 2017.
16. Available online: https://www.anblicks.com/blog/how-big-data-analytics-in-manufac-
turing-strengthens-the-industry (accessed on 10 June 2022).
17. S. Ryu, Y.-J. Kang, and H. Lee, “A study on detection of anomaly behavior in auto-
mation industry,” 2018 20th International Conference on Advanced Communication
Technology (ICACT), pp. 377–380, 2018.
18. M. M. H. Onik, C. -S. Kim, and J. Yang, “Personal data privacy challenges of the
fourth industrial revolution,” 2019 21st International Conference on Advanced
Communication Technology (ICACT), pp. 635–638, 2019.
19. Thuy Duong Oesterreich, Frank Teuteberg, “Understanding the implications of digiti-
sation and automation in the context of Industry 4.0: A triangulation approach and ele-
ments of a research agenda for the construction industry,” Computers in Industry, vol.
83, pp. 121–139, 2016.
20. J. Butt, “A strategic roadmap for the manufacturing industry to implement industry
4.0,” Designs, vol. 4, no. 2, p. 11, May 2020.
21. B. Bajic, A. Rikalovic, N. Suzic, and V. Piuri, “Industry 4.0 implementation challenges
and opportunities: A managerial perspective,” IEEE Systems Journal, vol. 15, no. 1,
pp. 546–559, March 2021.
Index
additive manufacturing 261 gated recurrent unit 121
agile IoT platform 132 Gaussian Naive Bayes 67
Amazon Web Services 91 Google Cloud Platform 92
anti-collision system 167
Apache Cassandra 107 Hadoop File System 90
Apache Kafka 101 hash function 218
Application Programming Interface 2 hierarchical programming 108
area under curve 124 hold out validation 122
artificial intelligence 28 human computer interaction 38
augmented reality 29 hybrid ontology 80
automatic speaker verification 113
automation systems 59 IBM Cloud 92
immutable 234
bagging 67 industrial Internet of Things (IoT) 3
big data analytics 31 industrial manufacturing 60
big data 86 industrial robot 275
Bitcoin protocol 245 infrastructure as a service 88
blockchain 25 insecure APIs 203
intelligent cyber-physical system 4
classification models 115 intelligent risk management 30
cloud computing 28 Internet of Things (IoT) 2
CNN model 66
cold-start problem 75 k-fold cross validation 122
confusion matrix 190 k-modes clustering 69
consensus algorithms 246 KNN 80
content-based methods 77
COVID-19 66 localization 61
customized healthcare services 150 logistic regression 67
cyber threats 197 long short-term memory 115
cyber-physical system 3 LoRaWAN Architecture 176

data analytics 32 MAC 175

data integrity 135 machine learning 25
data security and privacy 158 MapReduce 88
decision tree 67 Mean Average Precision 125
density evaluation model 169 message encryption 216
deployment of big data 99 Microsoft Azure 93
digital twins 23 Monte Carlo cross validation 123
direct acrylic graph 90 MQ telemetry transport 133

EEG signals 66 NMAC 213

ensemble learning 67 nonce 243
enterprise resource planning 103
ethereum 237 Open Connectivity Foundation 134
ETL 100 Oracle IoT 141
explainable artificial intelligence 47
physical threats 200
F1 Score 125 platform as a service 88
fault tolerance 4 powerful analytics 137

293
294 Index

precision 124 SQL ETL 100

PTC Thing work 132 stratified k-fold cross validation 123
supervised learning 187
quality healthcare 65 SVM 187
quarantine 221
textual documents clustering 219
random forest 67 ThingWorx 141
recall 124 trust-dependent CF algorithm 80
recommender system 74
recurrent neural network 115 usability 138
robot control systems 61
robots 60 virtual battery 275
ROLAP 103 virtual IoT slice services functions 133
routing protocol 11 virtual reality 29
voice conversion 114
Samsung Artik Cloud 141 vulnerabilities 200
Scala 100
ScienceSoft 104 Weka 109
singular value decomposition 220 wireless transmission 62
smart agriculture and forming 151
smart grids 151 x-rays 150
software as a service 88
speech synthesis 114 Zigbee 176
spoofing assault 285