______________________________________________

Digital Thread Enabled Manufacturing

Automation for Mass Personalization
______________________________________________

Engineering Management Project M Research Report

presented to The University of Auckland by
Student Name: Akhilnandh Ramesh

In partial fulfilment of the requirements for the award of the

Master of Engineering Management
2019-2020

Under the supervision of

Dr. Yuqian Lu
Faculty of Engineering
Abstract
In today’s era of globalization and consumerism, manufacturing industries are moving towards
adopting mass personalization, which refers to production of personalized products with high
efficiencies. This shift from push-type production to pull type mass personalization will result in
critical operational challenges for manufacturing organizations, ranging from eliciting customer
requirements, designing and manufacturing products to commissioning, installation and after sales
support. This research project addresses these operational challenges by identifying key
technological capabilities for manufacturing organizations in order to enable flexible yet resilient
engineering management systems for Mass Personalization. These capabilities identified have
been synthesized into a robust manufacturing automation framework which organizations can
adopt.

Central to this operational framework is the digital thread, which streamlines information flow
across different departments of the organization such as design, manufacturing, after sales and
field service. A digital replica of the product or digital twin is created at the beginning of the
product co-designing process and is updated with relevant product and process information as it
passes through different engineering stages such as manufacturing, installation, commissioning
and after sales support.

At each stage, efficient data exchange and communication across the digital thread can be enabled
through increased automation in business processes such as automatic integration of PDM and
ERP packages and improved real-time manufacturing operations management through superior
machine to machine communication.

The different digital twins are realized as physical products through self-organized shop-floor
consisting of self-aware and cognitive manufacturing entities. Manufacturing entities need to be
equipped with decision making capabilities and be capable of handling different scheduling
algorithms to facilitate dynamic decision making in shop-floor, aided by artificial intelligence and
extensible shop-floor control beyond traditional PLC/ SCADA based control paradigms. Unique
product installation procedures and consistency in product servicing can be ensured through a

ii |Digital Thread Enabled Manufacturing Automation for Mass Personalization

robust after sales service request management engine which capture human expertise in servicing
through artificial intelligence and natural language processing based methodologies.

Finally, a real world case study of an engineer to order industry in the construction sector has been
presented to validate the information flows between different digital threads. Finally the challenges
faced by the proposed automation framework and areas of future work are also discussed.

iii |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Acknowledgement

The author is thankful for the support and guidance of his project supervisor—Dr. Yuqian Lu, who
played a pivotal role with his feedback in shaping up this project. The author is thankful to
Professor Xun Xu and the Lab for Industry 4.0 Smart Manufacturing Systems at The University
of Auckland for the opportunity provided to work with the research team.

The author wishes to acknowledge his family member—his parents, Mr. N. Ramesh and
Mrs.Padmini Ramesh, his brother Mr. Abhinandh Ramesh and sister-in-law, Mrs. Girija
Abhinandh for their financial and moral support during author’s postgraduate study.

The author also wishes to thank his friends—Mr. Swapnil Chavan, Ms. Yasasvi Nepalli, Mr.
Joseph Xavier Baburaj, Mr. Zubair Ahmad Shalla , Mr. Kirthivasan Arul, Mr. Anandam Mallik,
Mr. Jaya Arun Prasanth, Mr. S. Ranganarasimman, Mr. Vinoth Siva, Mr. Srinath Manivannan, his
professional mentors—Mr. Muralikrishnan and Dr. Annamalai for their constant encouragement
and inputs during the course of the project.

iv |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 Research Outputs
The thesis is comprised of following research publication, which was an outcome of the research
carried out, alongside additional details and discussions.

Lu, Y., Ramesh, A. & Qin,Z. (2020). Digital Thread Enabled Manufacturing Automation
Towards Mass Personalization. In Manufacturing Science and Engineering Conference.
American Society of Mechanical Engineers (submitted).

v |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 Table of Contents
Declaration of Authorship i
Abstract ii
Acknowledgements iv
Research Outputs v
List of Figures vii
List of Tables viii
List of Abbreviations viii
1 Introduction 1
1.1 Need for Mass Personalization 1
1.2 Need for manufacturing automation framework 1
1.3 Research Objectives 3
1.4 Thesis Outline 4
2 Mass Personalization and its challenges 5
2.1 Evolution of mass personalization as a production paradigm 5
2.2 Mass Personalization Challenges 6
3 Digital thread enabled manufacturing automation framework 8
4 Technological Capabilities for information transfer in the digital thread 12
4.1 Technological capabilities for information transfer in Customer Co-Creation 12
4.2 Technological capabilities for information transfer in manufacturing execution 13
Technological capabilities for information transfer in commissioning and
4.3 after-sales service 15
5 Self-organized manufacturing for for Mass Personalization 18
5.1 Self-Organized Manufacturing for dynamic shop-floor planning 18
5.2 Scheduling Algorithm for a self-organized shop-floor 19
6 Case Study 26
7 Discussion 31
9 Conclusion and scope for future research 33
References 34

vi |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 List of Figures
Figure 1 The evolution of different production paradigms over the last century 5
Figure 2 Digital thread enabled manufacturing automation framework 8

Figure 3 ANSI/ISA-95 framework for manufacturing automation 9

Figure 4 Information flows in digital thread between different digital twins 11
Figure 5 Technological capabilities for effective customer co-creation 13
Figure 6 Technological capabilities for manufacturing execution 14

Figure 7 Shop-floor communication architecture for mass personalization 15

Information transfer architecture for installation, commissioning & after-sales

Figure 8 service 16
Self-organized shop-floor architecture in line with ANSI/ISA-95 framework
Figure 9 for manufacturing automation 20
Figure 10 Multi agent production scheduling algorithm 25
Figure 11 Interactive design platform for customer co-creation 27
Figure 12 Rule based classification algorithm for product features 28
Figure 13 “As-Designed Digital Twin” containing product data- BoM & CAD model 28
Figure 14 “As-Planned/ As-Built digital twin” containing process information 29
Figure 15 After sales support platform with data from “As-Maintained Digital Twin” 30

vii |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 List of Tables
Table 1 Comparison between different production paradigms 6
Table 2 Different shop-floor agents, their role description and attributes 22
Table 3 Proposed framework as a solution for challenges of mass personalization 32

List of Abbreviations
CAD Computer Aided Design
PDM Product Data Management
AI Artificial Intelligence
NLP Natural Language Processing
MQTT Message Queuing Telemetry Transport
IoT Internet of Things
M2M Machine to Machine
PLC Programmable Logic Controller
SCADA Supervisory Control And Data Acquisition
PAC Programmable Automation Controllers
ANSI American National Standards Institute
ISA International Society for Automation
PDM Product Data Management
OEE Overall Equipment Effectiveness
UI/UX User Interface/ User Experience
CAE Computer Aided Engineering
MES Manufacturing Execution System
XML eXtensible Markup Language
RDF Resource Description Framework
Ctot Total Cost of manufacturing
Ttot Total Time of manufacturing
SKU Stock Keeping Unit

viii |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 1
Introduction
In today’s globalized world, characterized by increased level of competition, technological
disruptions and increased focus on digital technologies, organizations often face the challenges of
having to stand out from the competition and adopt innovative practices such as launching products
with higher degree of personalization. As a result, many organizations are adopting production of
individualized products with a higher involvement of customer co-creation [1]. Furthermore, in
recent times, personalized products are gaining popularity among consumers, many of whom,
increasingly feel that mass produced products do not meet all their expectations and that
manufacturers need to personalize their product offerings to meet all their latent needs [2].

1.1 Need for Mass Personalization

According to market surveys carried out by different organizations, the demand for personalized
products ranges from 36 % among consumers in United Kingdom, ( from a survey by Deloitte) to
63 % of early technology adopters in the USA (from a study by the Forbes Magazine) [3][4].
Organizations increasingly feel that adopting product personalization will improve profit margins
( according to 83 % of retailers in apparel and footwear industry, as per survey by Lectra ), result
in increased sales, increased customer interaction ( as per a business study by Tata Consultancy
Services in 2019 ) and cultivate brand loyalty, with a net promoter score of more than 50 %
[5][6][7].

Personalization can result in product differentiation at a lower cost, reduce SKUs (stock keeping
options), reduce inventory and align production in-line with product pull [3] [8]. Efficient and
effective realization of personalized products can be achieved through Mass Personalization,
which is a strategy for manufacturing custom made products of higher degree of personalization
while at the same time achieving the operation efficiencies of mass production.

However, there have been only a few examples of successful deployments of Mass Personalization
in manufacturing organization and many are in the domain of apparel and fashion. A successful
example of an organization which has adopted Mass Personalization is Nike, whose product- Nike

1 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

ID facilitates customer co-creation of shoes through an interactive platform enabled through
internet and mobile applications and through specialized design studio [9][10].

Other examples of customer led Mass Personalization include Argos, British catalogue retailer
who has started producing personalized jewelry in small batches through 3D printing , with design
through a Web based UX and Kennedy City bicycles, who had used an interactive decision tree
for eliciting customer needs [11][12].Therefore, since Mass Personalization is a relatively new
production paradigm, an organization adopting the same effectively can benefit from the first
mover advantages in many of the global markets.

1.2 Need for a manufacturing automation framework

For organizations to effectively adopt Mass Personalization as a production paradigm, many
operational difficulties across different departments need to be overcome. Some of the operational
challenges that a manufacturing organization can face are lack of sufficient flow of product
information across various departments of the organization, difficulty in forecasting due to
constant changes in Bill of Materials ( BoM ) and production demands, need for efficient customer
relationship management and User interface to facilitate co-creation.

Information technology and increased manufacturing automation can help in can help in improving
interconnectivity between different manufacturing entities and serve as a solution for few of
these fundamental challenges. One such initiative which aims to improve efficiency in the
manufacturing organizations is Industry 4.0, mooted by the German government in 2011 [13].

Industry 4.0 aims to improve information flow in an organization by create synchronized digital
representations of physical systems , also known as digital twins , which can capture all critical
product and process information through data acquisition from physical resources [14]. The
created digital twin would be able to facilitate easier and automated exchange of data between
different entities, in a manufacturing organization, through an information network known as
digital thread [15]. The resulting digital entity, with interfaces with different physical systems
would constitute a cyber-physical model of a production system [16]. This cyber-physical
production system can help in building capabilities for Mass Personalization, when integrated with
technologies such as industrial automation, machine to machine communication protocols, greater
self-organization among machines and intelligent scheduling algorithms.

2 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

However, most of the current research in the domain of Mass Personalization is centered
predominantly around different methods of eliciting customer needs and product co-designing.
While digital twins and digital threads can enable Mass Personalization through improved
communication and manufacturing automation, there is a lack of research on developing a digital
thread enabled manufacturing automation and operations framework for implementing flexible,
robust and resilient engineering management systems for Mass Personalization. Therefore, in this
research, a digital thread based manufacturing automation framework for enabling engineering
management systems for mass personalization has been proposed. The different capabilities which
manufacturing organizations need to develop, for implementing the proposed digital thread based
manufacturing automation framework have also been identified and discussed.

1.3 Research objectives

This main motivation of this research is to develop flexible engineering management systems and
automated business processes for Mass Personalization through a digital thread based
manufacturing automation framework. This research also seeks to identify methods and
technological capabilities which can enable such an automation framework for Mass
Personalization.

In particular, there are four objectives for this research:

1. To carry out a qualitative analysis of available literature and case studies in the domain of
Mass Personalization and identify critical challenges faced by different departments of a
manufacturing organization while adopting mass personalization.
2. To develop a robust operational framework for manufacturing automation involving digital
twins and digital thread, so as to implement flexible engineering management systems
which can overcome the critical challenges identified.
3. To identify different capabilities which an organization has to develop, with respect to
information flows and in the shop-floor for effective implementation of the digital thread
based manufacturing automation framework
4. To explain the developed operational framework by means of a real world case study.

3 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

1.4 Thesis outline

This thesis describes in detail the research carried out in order to achieve the objectives. Both the
challenges faced in adopting Mass Personalization and the capabilities identified have been clearly
explained in the remainder of the thesis. The outline of the thesis are as follows:

The paradigm of Mass Personalization, and the resulting challenges arising from its
implementation are discussed in chapter 2.

Chapter 3 presents a composite framework for Mass Personalization consisting of digital twins
and digital threads. The flow of information across different departments of the organization has
been discussed in this section.

The different capabilities which an organization has to develop, with respect to improved
information flows and manufacturing have been discussed in chapters 4 and chapter 5.

In particular, Chapter 4 discusses in detail capabilities required for facilitating seamless transfer of
information to ensure efficient Mass Personalization by adopting the discussed framework. In this
section, different capabilities with respect to information technology such as software,
communication requirements have been discussed.

The concept of self-organized manufacturing and corresponding scheduling methodologies which

form a crucial part of the automation framework for Mass Personalization are explained in chapter
5. An example of a scheduling algorithm involving multi-agent mechanism is presented in this
chapter.

Chapter 6 discusses a case study which presents an application of the Mass Personalization
framework to a manufacturing enterprise engaged in made to order construction fittings.

Chapter 7 presents a discussion on how the capabilities identified and the digital thread enabled
Mass Personalization framework, further explained by means of a case study can overcome the
challenges of Mass Personalization.

Chapter 8 concludes the report and identifies scope for future research.

4 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 2
Mass Personalization and its challenges
In this chapter, the evolution of Mass Personalization as a production paradigm and the distinct
challenges caused by adopting this production paradigm are discussed.

2.1 Evolution of Mass Personalization as a production paradigm

Over the last 100 years, as shown by figure 1 [17], manufacturing has evolved from Mass
Production to Mass Customization to Mass Personalization.

Figure 1: The evolution of different production paradigms over the last century [17]

The main focus of Mass Production is to manufacture products, with little product variety through
dedicated production lines. Mass Customization focusses on realizing products with greater variety
through configurable and custom made modules. While the modules have been designed to cater
to the varied needs of the customers, the customer do not however, actively take part in product
co-creation.

The advent of flexible manufacturing systems and principles of lean manufacturing in 1980s has
enabled effective realization of Mass Customization, through batch production [18]. On the other
hand, Mass Personalization is a new paradigm of product realization in which the products are

5 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

tailor made to the needs and preferences of individual customers. [19]. In Mass Personalization,
customers are involved in product co-creation to produce a uniquely designed product which
incorporates latent and affective customer needs into the product structure, aesthetics and product
functions. [20][21]. Table 1 summarizes the key characteristics of different production paradigms.

Characteristic Mass Production Mass Customization Mass Personalization

Batch size Large Small to Medium Small, individualized
batches
Product Common product Basic Product Basic Product
architecture platform and parts architecture with architecture with custom
between different custom made and made/configurable
configurable modules modules and
[22] personalized modules
Production Dedicated production Flexible manufacturing Re-configurable cyber
method line systems, and lean physical production
manufacturing systems [24]
Customer’s role Buying Configuring and Customer Co-creation
selection of product
[23]
Efficiency Through realizing Through flexible batch Through effective
economies of scale production utilization of digital
technologies [25]
Table 1: Comparison between different production paradigms
2.2 Mass Personalization challenges

In Mass Personalization, considering that each product will be custom made, a unique pull type
workflow would accompany each product, unlike the push type work flow generally associated
with other production paradigms. Effective realization of Mass Personalization would result in
several distinct challenges across different departments in an organization:

 Engineering Design: Each personalized product has a unique product design and a bill of
materials. Considering that each product is co-created by the customer, a designer faces the
challenges of having to understand the manufacturing capabilities of the shop-floor while
designing a new product. Furthermore, to facilitate customer co-creation, a platform which
can capture customer requirements effectively in engineering design file needs to be
created.
 Manufacturing Planning: Production scheduling and manufacturing planning will face
the challenges of lack of timely knowledge of the process requirements for an incoming

6 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 product, constant changes to BoM between different orders, difficulties in adapting
common production control strategies, such as Kanban and CONWIP and limitations on
machine capacities and capabilities of existing MES and ERP Systems.
 Procurement and Purchase: Since each product is personalized with a unique BOM and
varying production runs, purchase team would face challenges of not having economies of
scale, difficulty in purchase scheduling, challenges in supplier evaluation and product
change management.
 Manufacturing: On the manufacturing front, a shop floor would face challenges of having
to schedule distinct process routes for each product, re-scheduling of manufacturing
operations based on dynamic changes in production and incoming orders.
 Commissioning and Field Service: Since each product would be unique with unique
service procedures, the commissioning and field service teams would need to have updated
service procedures and service history for each personalized product. This would need
robust information transfer and retrieval mechanisms, for ensuring consistent quality of
service during field service.
 After-Sales support: After-Sales teams will face challenges in tracing individual product
data throughout the lifecycle, obtaining information and procedures for product servicing
and training service personnel.

To summarize, as each product would be unique with unique BoM and process routings, all
departments and team need to adopt dynamic decision-making based on up-to-date information
about customers, product, process and resources. The following chapter present a holistic digital
thread based manufacturing automation framework, which can address these challenges.

7 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 3

Digital thread enabled manufacturing automation framework

In this section a holistic automation framework for Mass Personalization, which takes into account
the key technological capabilities across different departments of a manufacturing enterprise has
been presented. The aim of the framework is to streamline information flows between product
design, planning and manufacturing, installation, commissioning and servicing of product, and
hence improve operational efficiencies.

Figure 2 provides a comprehensive automation framework for realizing an individualized product.

The core of this framework is the digital thread, which facilitates seamless transfer of information
and data exchange across the organization. The digital thread connects different digital twins,
which are digital replicas of individualized products at different stages.

Figure 2: Digital Thread enabled manufacturing automation framework for Mass Personalization

8 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

When a product is co-designed by the customer, an “As-Designed Digital Twin” is established in
the computer aided design and engineering software environment. This “As-Designed Digital
Twin’ contains different product requirements such as product geometry, BoM and Product quality
requirements. In the shop-floor, the manufacturing resources coordinate with a Central Agent
Control System, which is responsible for initial task allocation and communication with ERP
systems, responsible for driving business processes and with digital twins. The product
information contained in the “As-Designed Digital Twin” is automatically transferred to the
Central Agent Control System, which manages all manufacturing tasks and supervises all
manufacturing systems, through the ERP system. An automated ERP-CAD bridge ensures process
consistencies and improves efficiencies by eliminating manual operations. The BoM structure
contained in “As-Designed Digital Twin” is parsed by examining each component against the parts
library stored in the ERP system so that off the shelf parts and in-house manufactured parts can be
separated. This will enable creation of automatic purchase orders and work flows in the ERP
system, so as to drive other business processes. A manufacturing BoM will also be created in
addition to design BoM with additional manufacturing planning information.

Figure 3: ANSI/ ISA-95 Framework for manufacturing automation [32]

The Central Agent Control System decides on an optimum production schedule for manufacturing
control for producing a production plan for each individualized product, each of which have their
own process requirements based on their product geometry manufacturing data. A shop-floor

9 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

producing individualized products may face uncertainties which can lead to production re-
scheduling such as delays in supply of various parts, changing priorities with order expedition,
machine downtime resulting in loss of manufacturing capabilities. Hence a shop-floor needs to be
reconfigurable, flexible and responsive to changes, with decentralized decision making for better
control [26][27][28]. A traditional top-down hierarchical communication model cannot be
employed for this purpose and instead shop-floor resources should be self-aware and communicate
with each other [29][30]. Hence the manufacturing system should exhibit a great degree of self-
organization with cognitive decision making abilities. While there are many approaches for self-
organization, the principle of multi-agent manufacturing systembased on ANSI/ISA-95 shop-floor
automation framework, presented in figure 3, has been employed in this framework [31] [32]
Shop-floor resources such as machines, conveying systems and an inspection systems are
classified as agents with pre-defined roles and responsibilities, resulting in realization of a smart
and connected shop-floor with greater cognitive abilities for dynamic decision making. The
product and its digital twins represent the physical production process which is the lowermost level
of ANSI/ISA-95 shop-floor automation framework, while different shop-floor resources represent
the levels of manufacturing control. The product information contained in the “As-Designed
Digital Twin” is transferred through the digital thread to the “As-Planned Digital Twin” which is
realized after coupling the product data with unique process routings and production schedules
pertaining to every individualized product.

At the same time, the “As-Designed digital twin” receives feedback through the digital thread
from the “As-Planned Digital Twin”, regarding manufacturing capabilities of the self-aware and
self-organized shop-floor. This feedback can be incorporated in the design process for better
manufacturability and personalization options. An “As-Built digital twin” is created after
production, and consists information regarding manufacturing BoMs and process information such
as process parameters and product traceability data. Simultaneously, information regarding
machine downtime and efficiency are sent as a feedback through the digital thread to the “As-
Planned digital twin” for more efficient scheduling. The “As-Designed Digital Twin”, the “As-
Planned Digital Twin” and the “As-Built Digital Twin” would constitute the “As-Maintained
Digital Twin” after creation of tailor-made service and installation procedures. The final product
BoM, process characteristics and production log, crucial for production traceability are transferred
via the digital thread to “As-Maintained Digital Twin”. At the same time Information regarding

10 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Product field performance and product service history are transferred through the digital thread as
a feedback to “As-Designed Digital Twin”.

Thus, the digital thread forms a seamless communication medium for information transfer between
different digital replicas of the personalized product. The forward and feedback information
transfer across the digital thread is described in figure 4.

The proposed operational framework can be used to realize an efficient engineering management
systems which can overcome many of the operational challenges discussed in chapter 2. For
effective implementation of this operational framework, technological capabilities which facilitate

Figure 4: Information flows in digital thread between different digital twins

efficient information transfer and self-organization in manufacturing have to be developed.

Chapter 4 discusses the critical technological capabilities which facilitate efficient information
transfer and data exchange between the digital twins across the digital thread. Chapter 5 discusses
in detail the concept of self-organized manufacturing along with a scheduling algorithm for
personalized products A real world case study has been explained in chapter 6 to explain the
application of the framework and the capabilities to a manufacturing organization engaged in
making individualized components for the construction industry.

11 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 4

Technological capabilities for information transfer in the digital

thread
In this chapter, different technological capabilities required for realizing consistent and efficient
sharing of information across the mass personalization digital thread have been discussed For
facilitating greater understanding, the discussion has been structured around different engineering
stages involved in realizing a personalized products, starting from co-designing and manufacturing
a personalized product to field service after installation and commissioning of a product.

4.1 Technological capabilities for information transfer in customer co-creation

Figure 5 highlights the key technological capabilities in order to facilitate gathering and
transferring of crucial product information over the digital thread, for establishing of an “As-
Designed Digital Twin”. Customer co-designing of a product can be effected through an UI/UX
software which facilitates participation through an interactive co-designing process. However,
there is a need for defining a set of parametric rules which facilitate this interaction. Firstly, product
features which can be personalized need to be identified and product architecture needs to be
classified into basic and personalized modules. Customer’s affective and cognitive needs can be
elicited through a data mining methodology employing recommendation engines, aided by
machine learning algorithms and techniques such as Citarasa descriptors [33]. A hierarchical
decision making process which classifies product features on the basis of a parametric composite
index, can be used to arrive at the set of personalized, configurable and basic modules, on the basis
of the identified customer needs, product utility and manufacturing complexity [34][35][36].

To ensure interactive co-designing, the user interface must have superior visualization and graphic
capabilities which facilitate free sketch CAD Modelling bound by set of design constraints such
as part interface rules and manufacturing constraints such as surface finish and tolerances [22][24]
UI/UX interface, simultaneously, also needs to interact in real-time design libraries stored in
existing PDM package to allow usage of existing product designs, store new designs and create
design BOM pertaining to every component involved in the product structure. This design BoM

12 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

along with other important product information would be stored in the “As-Designed Digital
Twin”.

Figure 5: Technological capabilities for effective customer co-creation

4.2 Technological capabilities for information transfer in manufacturing

execution

The technological capabilities which facilitate information transfer over the digital thread for
manufacturing execution in personalized production is explained in the figure 6. The product
information contained in the “As-Designed Digital Twin” and represented virtually in PDM
system would need to be parsed by the ERP system for identifying in-house manufactured and
off-the shelf parts. Automated integration of ERP Systems with PDM for improved efficiency in
generating workflows is required for mitigating inefficiencies associated with manual data
transfer [37][38]. This can be carried out through service mapping framework involving
semantic-web based ontologies delivered through XML and OWL [39]. Considering that each
product is uniquely manufactured, the ERP system must be able to custom-make work flows and
purchasing orders depending on analysis of lead time, quality assurance and reliability of various
off the shelf suppliers.

13 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

An information transfer and a communication architecture for manufacturing execution in the
multi-agent self-organized and self-aware shop-floor has been derived and the same is
represented by figure 7.This information transfer architecture is based upon the ANSI/ISA-95
manufacturing automation framework used for deriving the basic architecture for the multi-agent
self-organized shop-floor [32].

Figure 6: Technological capabilities for effective manufacturing execution

For establishing self-organization and dynamic responses in the shop-floor, manufacturing

resources should be equipped with superior machine to machine communication capabilities and
decision cognitive decision making abilities.

At the machine level, a service-oriented operational framework with a multi-agent methodology

identifying the function of each manufacturing entity can be adopted for production scheduling
and monitoring. Cognitive decision making can be enabled through machine intelligence aided by
AI and ML algorithms. Real-time sensing of the shop-floor environment and machine condition
monitoring can be enabled through data loggers and sensors. Machine to Machine communication
protocols such as MQTT tokens communicated over IoT can be used to enable a faster and more

14 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

responsive shop-floor which can react to dynamic changes better. Open-ended and extensible
manufacturing programming -beyond traditional PLC/SCADA such as through PAC is a method
by which smart shop-floor entities can be constructed.

For effective visualization of the “As-Planned/As- Built” digital twin across the digital thread, the
ERP system proposed would need to be built on a service-oriented architecture which can integrate
data from different shop-floor sources, provide business intelligence functions and data
visualization and be hosted on the cloud for rapid deployment and accessibility across different
enterprise functions [40][41][42].

Figure 7: Shop-floor communication architecture for mass personalization

4.3 Technological capabilities for information transfer in commissioning and
after-sales service

Once a personalized product is manufactured and shipped, it is vital that, product and process
information be retained for product traceability and warranty concerns management and also in
order to curate tailor made installation, maintenance and service procedures.

15 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

In many Engineer to Order industries which are engaged in personalized products, installation and
servicing often involve manual procedures involving human knowledge and expertise. Most of the
human knowledge is often captured in text-based documents involving technical terminologies,
resulting in inconsistencies [43][44]. Furthermore, the uniqueness of each product with respect to
product design and bill of materials can result in the need for tailor making unique servicing and
installation procedures to render consistent quality of service.

Figure 8: Information transfer architecture for installation, commissioning and after-sales service
management

Hence, it is required to capture human expertise of installation and servicing personalized products
and data available in text-documents to create a knowledge management tool- kit which interfaces
with product and process data available in PDM and ERP packages to create an automated and
robust after-sales service retrieval and request management engine, which when coupled with
AI/ML methodologies such as Natural language processing can be used to parse unstructured data
contained in text-documents onto different ontological data models represented by XML and RDF
topic maps [45][46]. This would also enable consistent information transfer across the digital

16 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

thread for the establishment of an “As-Maintained Digital Twin” which would contains unique
and tailor made installation and service procedures. The “As-Maintained Digital Twin” results in
a singular source of information regarding a personalized product, thereby ensuring consistent
quality of service while training field technicians and handling custom concerns

Figure 8 represents a summaries an architecture for information transfer for installation,

commissioning and after sales-service with respect to Mass Personalization.

17 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 5

Self-organized manufacturing for Mass Personalization

In this chapter, the key technological capabilities that are to be developed for a implementing a
self-organized shop-floor for mass personalization have been discussed. This chapter is organized
into two sections—the first section discusses the basic concept of self-organization for
manufacturing and identifies the key shop-floor requirements, while the second section presents a
production scheduling algorithm.

5.1 Self-organized manufacturing for dynamic shop-floor planning

Self-Organized manufacturing is defined as a collection of manufacturing resources (such as

machines, material handling systems such as conveyors, forklifts, human operators, supervisory
control centers and ERP systems) which can re-configure their organizational structure and
operational functions with minimal external interaction, so as to adapt to rapidly changing
production conditions, thereby achieving optimal performance. Self-Organization, on the whole,
aims at increasing autonomy among manufacturing resources through decentralized decision
making methodologies [47].

A Self-Organized shop-floor, coupled with robust scheduling algorithms can help in minimizing
downtimes and realizing efficiencies in a scenario of manufacturing components in low quantities
with fluctuating demand. A self-organized shop-floor can be realized by various principles such
as holonic manufacturing systems [48][49], bionic manufacturing systems [50][51] and multi-
agent manufacturing systems [52] [53][54].

In each of the three governing principles, a manufacturing resources treated as an autonomous

entity with a set of predefined roles and can communicate with each other through established
communication practices and plan for dynamic changes in a concurrent fashion, with different
levels of established hierarchy. Shop-floor control can be established by means of different
algorithms ranging from genetic and heuristic based algorithms to machine learning algorithm
such as reinforcement learning.

There have been a few examples of self-organized manufacturing being applied to Engineer to
Order industries. A Production as a Service (PaaS) framework involving a service oriented

18 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

architecture coupled with multi-agent methodology based on a genetic algorithm was employed
by Balta et al for manufacturing customized plastic parts [55] A real-time workshop scheduling
system for a factory producing jet engine parts for an expected increase in operational efficiency
such as productivity, percentage of orders completed without delay by 15- 30 % [56]. Both these
examples employed a constraint based algorithm which dynamically compared each and every
manufacturing resource available on the basis a series of priority rules. A multi-agent
manufacturing framework involving different smart shop-floor agents communicating with each
other was developed by Luo et al for developing a multi-period hierarchical scheduling mechanism
for custom made dies and moulds [57]

In this research, a similar multi-agent cl[5assification of manufacturing entities along with a

scheduling algorithm has been proposed for optimum production scheduling for mass
personalization.

5.2 Scheduling Algorithm for a self-organized shop-floor

Figure 9 presents the proposed shop-floor framework with different agents classified in accordance
with the ISA-95 shop-floor architecture. The physical production process is represented by a
product agent which is created each time a new order is received and for every individual
components which constitute a personalized batch. The product agent is connected through the
digital thread to the various digital twins of the product and receives real time information update
regarding the production status, as it passes through the shop-floor. This information which is
acquired by also serves as an input for different digital twins, which are updated dynamically
through the digital thread.

The manufacturing processes, at level 1 and level 2, are represented by a network of smart
manufacturing agents such as machining agents, conveying agents and inspection agents. These
agents are equipped with cognitive decision making abilities and can communicate with each other
through machine to machine communication protocols. At level 3, which represents manufacturing
operations management, is a Centralized Agent Control system. The Centralized Agent control
system forms an automated bridge between various business processes and the shop-floor. The
centralized agent control system is a computerized shop-floor agent which liaises with the ERP
system for daily sales plans uploaded by the sales team and also to acknowledge production
completion and material receipt, in order to drive other business processes such as procurement

19 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

functions. The Centralized Agent Control System also receives information from different
manufacturing entities such as their downtime, the machine parameters, the fixture setup in the
machine etc. Based on this information, the centralized agent control system creates an initial
process route, which is sent to the manufacturing agent.

At the business level is the ERP system, which communicates with the CAD/PDM package to
create automated work flows for each product, parse the product structure into in-house and bought
out parts, receives and updates product information and drives other business processes. The
different agents and their unique roles are mentioned in the table 2.

Figure 9: Self-organized Shop-floor framework in line with ANSI/ISA-95 framework for

manufacturing automation

Efficiency of a shop-floor, is measured by overall equipment effectiveness, which is a composite

index which takes into account machine availability, performance and quality requirements [58]

OEE is calculated as a composite index which is as follows: [59]

𝑂𝐸𝐸 = 𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑋 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑋 𝑄𝑢𝑎𝑙𝑖𝑡𝑦

𝐴𝑐𝑡𝑢𝑎𝑙 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 𝑋 100
𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑇𝑖𝑚𝑒
𝑁𝑒𝑡 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑇𝑖𝑚𝑒
𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 = 𝑋100
𝐴𝑐𝑡𝑢𝑎𝑙 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑇𝑖𝑚𝑒

20 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 𝑇𝑜𝑡𝑎𝑙 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑜𝑓 𝑑𝑒𝑓𝑒𝑐𝑡 𝑓𝑟𝑒𝑒 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝑄𝑢𝑎𝑙𝑖𝑡𝑦 = 𝑋 100
𝑇𝑜𝑡𝑎𝑙 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑜𝑓 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
Machine downtime is quantified as a sum of planned and unplanned losses, which are as follows:
𝐷𝑜𝑤𝑛𝑡𝑖𝑚𝑒 = 𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 + 𝑢𝑛𝑝𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒
𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 = 𝐶ℎ𝑎𝑛𝑔𝑒𝑜𝑣𝑒𝑟 𝑡𝑖𝑚𝑒 + 𝑞𝑢𝑎𝑙𝑖𝑡𝑦 𝑖𝑛𝑠𝑝𝑒𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 +
𝑝𝑙𝑎𝑛𝑛𝑒𝑑 𝑚𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝑡𝑖𝑚𝑒
𝑈𝑛𝑝𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 = 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝑏𝑟𝑒𝑎𝑘𝑑𝑜𝑤𝑛 𝑡𝑖𝑚𝑒 + 𝑡𝑜𝑜𝑙 𝑓𝑎𝑖𝑙𝑢𝑟𝑒 𝑡𝑖𝑚𝑒
Figure 10 presents an algorithm for enabling production planning and process routing for
manufacturing a personalized product. In order to enable manufacturing, product agents are
grouped into product categories based on the product features. Machining agents are also
categorized depending on their attributes such as machine capabilities, set-up capabilities and
process nature. For any product, there are “n” machines which can perform a required operation.
An iterative bidding mechanism is used by the central agent control system and each self-aware
smart machining agent to determine the optimum process routing. This mechanism is explained as
follows:

Any new product instituted by the central agent control system requires n operations, represented
by 𝑂𝑃 = {𝑂𝑃1, 𝑂𝑃2, 𝑂𝑃3, … … , 𝑂𝑃𝑛}.

The shop-floor has a number of resources, denoted by set A, comprising of manufacturing agents
, conveying agents and inspection agents which can be used to manufacture a personalized product
The sets of agents denoted by as follows:

Set of Inspection agents: 𝐼 = {𝐼1, 𝐼2, 𝐼3, … … , 𝐼𝑛}

Set of conveying agents: 𝐶 = {𝐶1, 𝐶2, 𝐶3, … … , 𝐶𝑛}

Set of machining agents: 𝑀 = {𝑀1, 𝑀2, 𝑀3, … … , 𝑀𝑛}

For any product, the actual combination set of agents required to perform the operations set OP is
denoted by the set A1, such that

𝐴1 ⊂ 𝐴

In order to ensure that each of the operations can be completed in case of any unplanned
downtimes, machines are selected in such a way that a particular operation can be performed in
more than one machine. This is to ensure that the product can still be manufactured in case the

21 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

destined machine breaks down. The central agent control system creates an initial process routing
for each product by matching the operations set OP with the set of shop-floor resources A, so as
to arrive at the actual combination set A1, which can perform a particular manufacturing operation.

Agents Role Description Attributes

Product Agents Represents a product, contains Product geometry,
product information and is dimensions and tolerances,
generated for each product quality requirements,
being manufactured manufacturing
requirements, BoM,
Machining Agents Machines are smart entities Machine capability, set up
which are responsible for requirements , machine
realization of smart products parameters, energy
from WIP. consumption, cycle time ,
Machining agents can take communication
orders from other machining methodology
system and also from
centralized agent control system
Conveying Agents Conveying agents perform in Load capacity, speed,
transportation and material throughput, application,
handling functions. power consumption
These include conveyors,
AGVs, smart forklifts and smart
workers
Inspection Agents Inspection agents perform Device Resolution, type,
inspection operations- these data acquisition method,
include vision system, tool feedback systems,
setting devices, CMMs Repeatability
Central Agent Control Automated bridge between Scheduling algorithms,
System shop-floor and ERP System Communication and data
Responsible for creation of acquisition method
initial process routings and
process schedules, monitors
machine status
ERP System Drives business processes Modularity & Service
Enables information sharing Composition, scalability
across all verticals and extensibility with
Performs control at business business intelligence
level
Table 2: Different shop-floor agents, their role description and attributes
The product agents are assigned to particular machines in the combination set on the basis of
series of bids and auctions and also depending on the real time information status of each machine,

22 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

such as downtime, the fixture running in the machine, previous orders which have been assigned
earlier to the machine and the machine parameters.

The central agent control system sends out a call for each of the machines in the combination set
which can perform the first operation on the basis of algorithms focused on minimizing total cost
of production or the total time of manufacturing.

The total cost of manufacturing a particular product would involve following elements:

Machining cost: This is proportional to the time taken by the machine and the unit power
consumption - depends on machine parameters and also machine specifications (such as a
hydraulic drive or an electrical drive)

Tooling cost: Tooling cost is the cost incurred in the purchase of cutting tools and fixtures—for
example, one cutting tool or an insert maybe be utilized to produce few similar components, before
being scrapped or sent for a re-grind. In such a case, the contribution of the tooling cost to the
product cost is fractionally proportion to the total number of components manufactured with the
tool. Similarly, the cost contribution of fixtures is calculated as the cost of fixtures divided by the
estimated useful life

Conveying cost: Cost of transportation of product agent from one machining agent to another
across the shop-floor. This is proportional to the power consumption expounded in the process of
transportation.

Cost of downtime: Cost of downtime is the cumulative cost involved in restoring a manufacturing
entity to its original state and the additional cost, if any, which would be incurred due to time delay
involved ( such as additional machining cost due to cost of expedition). This is generally calculated
as an operational cost proportional to the time loss due to unplanned and planned downtimes [60]

Each machine in the combination set sends out a series of bids with information regarding cycle
time and total cost. The central agent control system determines the process routes and assigns the
product agent to machining agents in the combination set which can manufacture the product in
such a way that the total cost is minimal or the total manufacturing time is minimal ( in case
products need to be expedited ).

23 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

This is given by:

𝑀𝑖𝑛 ( 𝐶𝑡𝑜𝑡 = ∑ 𝐶𝑖)

𝑀𝑖𝑛 ( 𝑇𝑡𝑜𝑡 = ∑ 𝐶𝑖)

Within a particular time frame, the central agent control system determines the combination set of
all the machines to determine a rough production plan. An iterative approach which analyzes all
the possible machining combinations to determine the lowest cost or lowest time is employed this
purpose.

After completion of the first operation, the product agent is updated with process information such
as machine parameters and product traceability details and these are transferred over the digital
thread to the “As-Planned/As-Built Digital Twin”. The machining agent which has carried out the
first operation calls for bid again, using a similar methodology and assigns component to the next
machining agent in the combination set which can perform the subsequent operation in the process
route after checking the individual status of each machine.

The exact combination set may vary from the resource combination originally decided, on the
basis of real-time machine status, such as to account for machine breakdown, quality defects,
urgent orders which may have to be expedited.

24 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 Figure 10: Multi-Agent production scheduling algorithm

25 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 6
Case Study
In this chapter, the proposed framework is validated by means of a real life case study which
involves manufacturing personalized aluminium fins, external cladding, balustrades and louvres
used for building facades. Since each construction project is unique and custom made, each order
is personalized and is manufactured in varying batch sizes. At the same time, the products would
be having a basic open product architecture based on calculations such as wind load and other
design constraints. The manufacturing enterprise would also have a few constraints such as
organization’s manufacturing and sourcing capability, choice of materials, sectional profile, size
and colour .

In the operational framework proposed for this industry, the unique customer requirements such
as choice of basic design, colour and size requirements are captured through an interactive product
design interface, which allows product co-creation. The product co-creation interface would
facilitate co-creation by presenting choices of a common platform along with configurable and
personalized modules, which can be visualized through a 3D rendering package. The Visualization
platform will be interfaced with CAD/PDM packages to create design BOM on the basis few
design rules. This is used to create the “As-Designed Digital Twin” of the product. The “As-
Designed Digital Twin” would contain the preliminary BOM of the product along with product
geometry. Figure 11 represents a visual of the interactive design platform in which customers can
co-create custom-made aluminium fins and balustrades. A rule based classification algorithm used
for categorizing product features into personalized, configurable and common product
characteristics is presented in the figure 12. The “As-Designed digital twin”, represented in figure
13 forms the preliminary information for sourcing the components and is used by centralized
planning agent in the self-organized shop-floor.

The centralized planning agent parses the BoM into bought out and in-house parts and uploads
onto the ERP system. The centralized agent control system also liaises with the ERP system to
gather production demand and creates process plan and production schedule for facilitating
distributed decision making among various manufacturing agents in shop-floor. The different
manufacturing entities such as assembly agents, inspection agents, conveying agents can interact
with each other through a mechanism of bids and auctions, represented to determine a dynamic

26 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

production schedule, considering the production demand, shipping date and machine specification.
There would also be changes in part numbers, to take into account work in progress and this is
represented by dummy part numbers. The process plan, process parameters and part traceability
data along with Manufacturing BoM would constitute the “As-Planned Digital twin”, which along
with production information such as operator name, time of processing and forms “As-Built
Digital Twin”. This is represented by figure 14.

Figure 11: Interactive design platform for customer co-creation

27 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Figure 12: Rule based classification algorithm for product features

Figure 13: As-Designed Digital Twin containing product data- BoM and CAD model

The commissioning and after-sales service team can access the “As-Planned/As-Built Digital
Twin” in the ERP System for production data such as process parameters and product traceability
details and also the “As-Designed Digital Twin” from the PDM system, visualized through a 3D
rendering interface to view the installation condition. Each of the digital twins, enabled thru a

28 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

digital thread would provide information for creation of tailor made product installation procedures
and service instructions. These serve as the “As-Maintained Digital Twin” for the product,
represented by figure-15. Furthermore, with the incorporation of AI/ML and NLP procedures,
information contained in manual data and documents can be captured effectively to create a
Service Request Management engine. Hence, this would serve as a robust, up-to date, singular and
consistent source for handling service requests for a personalized products.

Since at each of the engineering stages, from the point of co-designing to manufacturing and
servicing of a tailor made product, the product and process information is stored as a digital twin,
a closed loop feedback system for better design of personalized product is established. The service
history and product performance data can be transferred to the “As-Designed Digital Twin” and
stored in PDM repositories for building upon product knowledge base, for better and unique design
of personalized products.

Figure 14: “As-Planned/ As-Built Digital Twin” containing process information

29 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Figure 15: After sales support platform with data from “As-Maintained Digital Twin”

For example, data regarding the failure at mounting points and bolt-failure at critical points is sent
back to the “As-Designed Digital twin” for arriving at possible combination sets for basic, custom-
made and personalized modules.

30 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 7

Discussion
Mass Personalization is a production strategy which aims to bring efficiency and effectiveness
similar to mass production in manufacturing personalized products. Personalized products are
typically manufactured in Engineer to Order Industries. Engineer-to-Order industries
manufacturing personalized products have historically suffered from low levels of productivity
arising uncertainties in orders, difference in BoM between every order and scheduling
complexities [60][61] .

These fundamental challenges faced by such organizations can be overcome by adopting industry
the digital thread based manufacturing automation framework presented, encompassing the key
technological capabilities for information transfer, which have been discussed. The framework can
lead to the establishment of a flexible engineering management systems for mass personalization.
Table 3 presents how the challenges in adopting mass personalization can be overcome by the
proposed framework.

Different product digital twins will enable an adaptive, responsive and decentralized-distributed
manufacturing automation. The case study presented has demonstrated the application of proposed
framework by presenting a cyber-physical representation of the manufacturing system that stores
product information and process information virtually and can communicate with each other
through information technology protocols. The cyber physical production process is actualized in
a self-organized and reconfigurable shop-floor.

However, for establishment of a reconfigurable muti-agent shop-floor , there is need for robust
production scheduling algorithm, which integrates regulatory methodologies for control,
coordination and production scheduling [59], and a knowledge base, containing product history
and technical information regarding other shop floor entities such as machines [60][61]. The digital
thread which facilitates seamless information transfer between the product and process digital
twins enables in the creation of this knowledge base through the computerized Central Agent
Control system. Furthermore, there is also a need for a production scheduling strategy which with
provides a clear process flow for job allocation. One such possible algorithm, based on production
costs and production time which can be adopted is presented in chapter-5.

31 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 Organizational Challenge Solution from the framework
Departments
Engineering Suitable interface for customer- Customer Co-designing through a
Design co-creation CAD interface.
Data-mining for customer’s affective
through recommendation engine
Classification algorithm for product
features
Need for Designer to Feedback systems in the digital
understand the feedback thread through automatic integration
regarding manufacturing between PDM and ERP Packages
capabilities and shop-floor
Manufacturing Uncertainties in forecasting Multi-agent production scheduling
Planning algorithm
Inefficiency in understanding Automatic integration of ERP
product and process Packages with PDM packages.
requirement for incoming
product
Unable to apply traditional Dynamic manufacturing planning
production control through “Smart” manufacturing
methodologies such as Kanban, entities, equipped with AI/ML and
CONWIP real time production monitoring
Procurement and Challenges in Change A singular source of information in
Purchase management “As-Designed digital twin”, accesses
through the digital thread
Difficulties in purchase Dynamic demand planning by
scheduling Central Agent Control System, and
automatic linkage with ERP system.
Manufacturing Distinct process routings Process routings generated by central
and Shop-floor agent in self-organized shop-floor
Need for adapting to dynamic Dynamic production scheduling
changes algorithms
Limitations and Challenges of Integration of CAD and ERP
existing ERP and MES packages through service mapping
and MES programming beyond
PLC/SCADA.
Commissioning Need to generate unique “As-maintained Digital Twin”
and Field Service installation and commissioning containing unique service procedures.
procedures.
Consistency in quality of Capturing human knowledge in As-
service Maintained digital twin through NLP
and AI.
After Sales Product traceability Product traceability provided through
Support requirements digital twins communicating through
digital thread
Table 3: Proposed framework as a solution for challenges of mass personalization

32 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

Chapter 8

Conclusion and scope for future research

In this project thesis, different operational challenges faced by manufacturing organizations while
adopting mass personalization, such as uncertainties in forecasting, difficulties in information
transfer across the organization, distinct process plans for each orders have been identified and
analyzed. These fundamental challenges can be overcome by adopting the presented digital thread
enabled manufacturing automation framework. The different crucial technological capabilities
which can enable implementation of such an automation framework have been identified and
discussed. A case study of an Engineer to Order enterprise in the construction sector has been
discussed to explain the real-world application of the proposed framework. This research report
would benefit manufacturing organizations and research in developing an understanding of the
technological changes which have to be undertaken to enable flexible engineering management
systems for mass personalization.

Future scope for research exists with respect to real world deployment of the self-organized
production scheduling algorithm presented. The feedback gained from such a practical
implementation will be useful in contributing to the knowledge bank existing in the domain of
mass personalization, which, currently, is predominantly, focused on product design.

33 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

References
[1] Lu, Y. and Xu, X., (2018). “Resource virtualization: a core technology for developing cyber-
physical production systems”.Journal of manufacturing systems, 47, pp.128-140.

[2] Capegemini.(2018). The shift from mass production to mass personalization. Retrieved from
https://www.capgemini.com/gb-en/2018/05/the-shift-from-mass-production-to-mass-
personalisation/

[3] The Deloitte Consumer Review. (2015). Made-to-order: The rise of mass personalization .
Retrieved from https://www2.deloitte.com/ch/en/pages/consumer-business/articles/made-to-
order-the-rise-of-mass-personalisation.html.

[4]. Forbes (2020). New Research shows consumers already expected mass personalization.
Time to get ready. Retrieved from https://www.forbes.com/sites/jimvinoski/2020/01/20/new-
research-shows-consumers-already-expect-mass-personalization-time-to-get-
ready/#7d3df8d1223e/

[5] Sourcing Journal (2019). Personalization is Everywhere Because Consumers Want it.
Retrieved from https://sourcingjournal.com/topics/consumer-insights/personalization-report-
demand-margins-151680/

[6] Forbes (2019). Is Mass Personalization The Next Big Hurdle for Manufacturers?. Retrieved
from https://www.forbes.com/sites/jimvinoski/2019/10/17/is-mass-personalization-the-next-big-
hurdle-for-manufacturers/#46f770d97203

[7] Bain & Company (2013). Making it personal: Rules for success in product customization.
Retrieved from https://www.bain.com/insights/making-it-personal-rules-for-success-in-product-
customization/

[8] HP Megatrends (2017). From mass production to mass personalization. Retrieved from
https://hpmegatrends.com/from-mass-production-to-mass-personalization-1ad6746dc24d

[9] Sourcing Journal (2019). Apparel Brands Are Testing the Waters of Mass Customization.
Retrieved from https://sourcingjournal.com/topics/technology/personalization-report-
investments-adaptations-152409/

34 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

[10] Protolabs. (2019). Smart Manufacturing Enables Mass Customization Trend. Retrieved
from https://www.protolabs.com/resources/blog/smart-manufacturing-enables-mass-
customization-trend/

[11] Wired. (2015). Argos sells customisable 3D printed jewellery. Retrieved from
https://www.wired.co.uk/article/argos-3d-printing-jewellery

[12] Kennedy City Bicycles (2020). The KCB Story. Retrieved from
https://www.kennedycitybicycles.cc/

[13] Kagermann, H., Helbig, J., Hellinger, A. and Wahlster, W., (2013). Recommendations for
implementing the strategic initiative INDUSTRIE 4.0: Securing the future of German
manufacturing industry; final report of the Industrie 4.0 Working Group. Forschungsunion.

[14] Lu, Y., Liu, C., Kevin, I., Wang, K., Huang, H. and Xu, X., 2020. Digital Twin-driven smart
manufacturing: Connotation, reference model, applications and research issues. Robotics and
Computer-Integrated Manufacturing, 61, p.101837.

[15] Singh, V., & Willcox, K. E. (2018). Engineering design with digital thread. AIAA
Journal, 56(11), 4515-4528.

[16] Zhuang, C., Liu, J., & Xiong, H. (2018). Digital twin-based smart production management
and control framework for the complex product assembly shop-floor. The International Journal
of Advanced Manufacturing Technology, 96(1-4), 1149-1163.

[17] Hu, S.J., (2013). Evolving paradigms of manufacturing: from mass production to mass
customization and personalization. Procedia CIRP, 7, pp.3-8.

[18] Koren, Y. (2010). The global manufacturing revolution: product-process-business

integration and reconfigurable systems (Vol. 80). John Wiley & Sons.

[19] Wang, Y., Ma, H. S., Yang, J. H., & Wang, K. S. (2017). Industry 4.0: a way from mass
customization to mass personalization production. Advances in Manufacturing, 5(4), 311-320

[20] Zhou, F., Ji, Y., & Jiao, R. J. (2013). Affective and cognitive design for mass
personalization: status and prospect. Journal of Intelligent Manufacturing, 24(5), 1047-1069.

[21] Zheng, P., Yu, S., Wang, Y., Zhong, R.Y. and Xu, X., (2017). User-experience based
product development for mass personalization: a case study. Procedia CIRP, 63, pp.2-7

35 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

[22] Koren, Y., Shpitalni, M., Gu, P., & Hu, S. J. (2015). Product design for mass-
individualization. Procedia Cirp, 36, 64-71.

[23] Tseng MM, Jiao J, Merchant ME, (1996), Design for Mass Customization. CIRP Annals-
Manufacturing Technology 45(1): 153-156.

[24] Tan, C., Hu, S. J., Chung, H., Barton, K., Piya, C., Ramani, K., & Banu, M. (2017). Product
personalization enabled by assembly architecture and cyber physical systems. CIRP
Annals, 66(1), 33-36.

[25] Wang, Y., Ma, H. S., Yang, J. H., & Wang, K. S. (2017). Industry 4.0: a way from mass
customization to mass personalization production. Advances in Manufacturing, 5(4), 311-320.

[26] Pourabdollahian, G., Taisch, M. and Tepe, G., (2013), September. Exploring different faces
of mass customization in manufacturing. In IFIP International Conference on Advances in
Production Management Systems (pp. 13-20). Springer, Berlin, Heidelberg.

[27] Davrajh, S. and Bright, G., (2013). Advanced quality management system for product
families in mass customization and reconfigurable manufacturing. Assembly Automation, 33(2),
pp.127-138.

[28] Gilke, N., Mantha, S. and Thampi, G., (2013). Infrastructural Back Bone Of Enabling And
Converging Technologies For Mass Customization Manufacturing System In Automotive
Industries. International Journal for Quality Research, 7(1), pp.153-162.

[29] Essers, M.S. and Vaneker, T.H., (2014). Evaluating a data distribution service system for
dynamic manufacturing environments: a case study. Procedia technology, 15, pp.621-630.

[30] Schild, K. and Bussmann, S., (2007). Self-organization in manufacturing

operations. Communications of the ACM, 50(12), pp.74-79.

[31] Tharumarajah, A., (2003). A self-organising view of manufacturing enterprises. Computers

in Industry, 51(2), pp.185-196.

[32] Brandl, D. (2008). What is ISA-95? Industrial best practicesof manufacturing information
technologies with ISA-95 models. Retrieved from http://apsom.org/docs/T061_isa95-04.pdf

36 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

[33] Mavridou, E., Kehagias, D. D., Tzovaras, D., & Hassapis, G. (2013). Mining affective needs
of automotive industry customers for building a mass-customization recommender
system. Journal of intelligent manufacturing, 24(2), 251-265.

[34] Berry C, Wang H, Hu SJ (2013) Product Architecting for Personalization. Journal of

Manufacturing Systems 32(3):404–411.

[35] Kaul A, Rao VR (1995) Research for Product Positioning and Design Decisions: An
Integrative Review. International Journal of Research in Marketing 12 (4):293–320.

[36] Koren Y, Hu SJ, Gu P, Shpitalni M (2013) Open-Architecture Products. CIRP Annals—

Manufacturing Technology 62(2):719–729.

[37] Stojkić, Ž., Veža, I., & Bošnjak, I. (2016, January). Concept of information system
implementation (CRM and ERP) within industry 4.0. In 26th DAAAM International Symposium
on Intelligent Manufacturing and Automation.

[38] Basl, J. (2018, September). Analysis of industry 4.0 readiness indexes and maturity models
and proposal of the dimension for enterprise information systems. In International Conference
on Research and Practical Issues of Enterprise Information Systems (pp. 57-68). Springer,
Cham.

[39] Lu, Y., Wang, H., & Xu, X. (2019). ManuService ontology: A product data model for
service-oriented business interactions in a cloud manufacturing environment. Journal of
Intelligent Manufacturing, 30(1), 317-334.

[40] Andročec, D., Picek, R., & Mijač, M. (2018, January). The ontologically based model for
the integration of the IoT and Cloud ERP service. In Proceedings of the 8th International
Conference on Cloud Computing and Services Science, Funchal, Madeira, Portugal (pp. 481-488

[41] Koupaei, M. N., Mohammadi, M., & Naderi, B. (2016). An Integrated Enterprise Resources
Planning (ERP) Framework for Flexible Manufacturing Systems Using Business Intelligence
(BI) Tools. International Journal of Supply and Operations Management, 3(1), 1112.

[42] Ghazanfari, M., Rouhani, S., Jafari, M., & Taghavifard, M. T. (2009). ERP requirements for
supporting management decisions and business intelligence. IUP Journal of Information
Technology, 5(3).

37 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

[43] R. Owen, I. Horváth. Towards product-related knowledge asset warehousing in

enterprises, Proceedings of the 4th International Symposium on Tools and Methods

of Competitive Engineering, TMCE 2002 (2002) 155–170.

[44] Z.Y. Wu, X.G. Ming, Y.L. Wang, L. Wang. Technology solutions for product lifecycle

knowledge management: framework and a case study, International Journal of Production

Research. 52 (21) (2014) 6314–6334.

[45] Klein, Dan; Manning, Christopher D. (2002). Natural language grammar induction using a
constituent-context model. Advances in Neural Information Processing Systems.

[46] Chungoora, N., Cutting-Decelle, A. F., Young, R. I. M., Gunendran, G., Usman, Z.,
Harding, J. A., & Case, K. (2013). Towards the ontology-based consolidation of production-
centric standards. International Journal of Production Research, 51(2), 327-345.

[47] Schild, K. and Bussmann, S., (2007). Self-organization in manufacturing

operations. Communications of the ACM, 50(12), pp.74-79.

[48] Botti, V., & Giret, A. (2008). Holonic manufacturing systems (pp. 7-20). Springer London.

[49] Valckenaers, P., & Van Brussel, H. (2005). Holonic manufacturing execution
systems. CIRP annals, 54(1), 427-432.

[50] Tharumarajah, A. (1996). Comparison of the bionic, fractal and holonic manufacturing
system concepts. International Journal of Computer Integrated Manufacturing, 9(3), 217-226.

[51] Gu, W., Tang, D., Zheng, K., & Wang, L. (2011). A neuroendocrine-inspired bionic
manufacturing system. Journal of Systems Science and Systems Engineering, 20(3), 275.

[52] Ouelhadj, D., Hanach, C., & Bouzouia, B. (1998, May). Multi-agent system for dynamic
scheduling and control in manufacturing cells. In Proceedings. 1998 IEEE International
Conference on Robotics and Automation (Cat. No. 98CH36146) (Vol. 3, pp. 2128-2133). IEEE.

[53] Kouiss, K., Pierreval, H., & Mebarki, N. (1997). Using multi-agent architecture in FMS for
dynamic scheduling. Journal of intelligent Manufacturing, 8(1), 41-47.

[54] Guo, Q., & Zhang, M. (2009). A novel approach for multi-agent-based intelligent
manufacturing system. Information Sciences, 179(18), 3079-3090.

38 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

[55] Balta, Efe C., Lin,Y., Barton,K., Tilbury, D.M., Mao, Z.M., (2018). Production as a Service:
A Digital Manufacturing Framework for Optimizing Utilization. IEEE Transactions on
Automation Science and Engineering, 15(4), 1483-1493

[56] Shpilevoy, V., Shishov, A., Skobelev, P., Kolbova, E., Kazanskaia, D., Shepilov, Y. and
Tsarev, A., (2013). Multi-agent system “Smart Factory” for real-time workshop management in
aircraft jet engines production. IFAC Proceedings Volumes, 46(7), pp.204-209.

[57] Luo, H., Fang, J. and Huang, G.Q., (2015). Real-time scheduling for hybrid flowshop in
ubiquitous manufacturing environment. Computers & Industrial Engineering, 84, pp.12-23.

[58] Dal, B., Tugwell, P., & Greatbanks, R. (2000). Overall equipment effectiveness as a
measure of operational improvement–a practical analysis. International Journal of Operations &
Production Management.

[59] Nakajima, S. (1988). Introduction to Total Productive Maintenance (TPM), Productivity

Press, Cambridge. MA.

[60] Bertrand, J.W.M and Muntslag D.R., (1993), “Production Control in engineer-to-order
firms”, International Journal of Production Economics,30-31, 3-22

[61] Hicks, C. and Braiden, P. M., (2000), “Computer-aided production management issues in
the Engineer-to-order Production of Complex Capital goods explored using a simulation
approach.”, International Journal of Production Research, 38, 4783–4810

39 |Digital Thread Enabled Manufacturing Automation for Mass Personalization

 ______________________________________________

Digital Thread Enabled Manufacturing

Automation for Mass Personalization
_____________________________________________

Engineering Management Project M Company Report

presented to The University of Auckland by
Student Name: Akhilnandh Ramesh
Student ID: 189204075

In partial fulfilment of the requirements for the award of the

Master of Engineering Management
2019-2020

Under the supervision of

Dr. Yuqian Lu
Faculty of Engineering
DIGITAL THREAD AKHILNANDH
ENABLED RAMESH

MANUFACTURING
AUTOMATION FOR
Digital Thread Enabled MASS
Manufacturing Automation for Mass Personalization i

PERSONALIZATION
 Declaration of authorship

I, Akhilnandh Ramesh, declare that this thesis titled, “Enabling Mass Personalization in
Engineering Industries” and the work presented in it are my own. I confirm that:

 This work was done wholly or mainly while in candidature for the degree—Master of
Engineering Management at The University of Auckland
 Where any part of this thesis has previously been submitted for a degree or any other
qualification at this University or any other institution, this has been clearly stated.
 Where I have consulted the published work of others, this is always clearly attributed
 Where I have quoted from the work of others, the source is always given. With the
exception of such quotations, if any, this thesis is entirely my own work.
 I have acknowledged all main sources of help.
 Where the thesis is based on work done by myself jointly with others, I have made clear
exactly what was done by others and what I have contributed myself

Signed: __________________
Date: 27/06/2020

Digital Thread Enabled Manufacturing Automation for Mass Personalization i

 Executive summary

In today’s era of globalization, consumerism, technological disruptions and increased

competitions, organizations are increasingly launching tailor made individualized products with
bespoke customization options and customer co-creation. Furthermore, the popularity of
personalized products and services has been the rise over the last decades, due to attitudinal
shifts among consumers, who increasingly feel that mass produced products do not meet all their
expectations and needs.

For organizations, personalization is a lower cost option for product differentiation and can help
in aligning customer pull with production, thereby reducing inventory. Effective realization of
product personalization can be achieved through Mass Personalization, which is a strategy for
manufacturing custom-made individualized products in varying batch sizes, while at the same
time, achieving the operational efficiencies of mass production.

However, organizations need to overcome many critical operational challenges such as ensuring
efficient information flows across all the departments of an organization, increased manual
workflows due to uniqueness of each order, facilitating effective customer co-creation,
uncertainties in production planning, need for efficient production execution systems and
individualized servicing procedures.

This project addresses these operational challenges by identifying key technological capabilities
for manufacturing organization in order to enable flexible yet resilient engineering management
systems for Mass Personalization. The identified capabilities have been synthesized into a robust
Manufacturing automation framework for adopting mass personalization.

Central to this operational framework is the digital thread, which enables efficient information
transfer of information across different departments of the organization such as design,
manufacturing, after-sales and field service. A digital representation of the product called digital
twin is created at the beginning of the co-design process and is updated with relevant information
at each of the engineering stages such as manufacturing, installation, commissioning and after
sales support.

At each stage, efficient data exchange and communication across the digital thread can be
enabled through increased automation in business processes such as automated bridging between
CAD and ERP systems and real-time manufacturing operations management through machine to
machine communication. Unique product servicing procedures can be enabled through
automated after sales service request management aided by artificial and intelligence and natural
language processing for capturing human knowledge and eliminate manual workflows, thereby
ensuring service consistency.

The proposed digital thread based manufacturing automation framework is explained by means
of a real-world case study of an engineer to order industry in the construction sector. Finally,
scope for future advancements in this domain has been discussed.

Digital Thread Enabled Manufacturing Automation for Mass Personalization ii

 Acknowledgement
The author is thankful for the support and guidance of his project supervisor—Dr. Yuqian Lu,
who played a pivotal role with his feedback in shaping up this project. The author is thankful to
Professor Xun Xu and the Lab for Industry 4.0 Smart Manufacturing Systems at The University
of Auckland for the opportunity provided to work with the research team.

The author wishes to acknowledge his family member—his parents, Mr. N. Ramesh and
Mrs.Padmini Ramesh, his brother Mr. Abhinandh Ramesh and sister-in-law, Mrs. Girija
Abhinandh for their financial and moral support during author’s postgraduate study.

The author also wishes to thank his friends—Mr. Swapnil Chavan, Ms. Yasasvi Nepalli, Mr.
Joseph Xavier Baburaj, Mr. Zubair Ahmad Shalla , Mr. Kirthivasan Arul, Mr. Anandam Mallik,
Mr. Jaya Arun Prasanth, Mr. S. Ranganarasimman, Mr. Vinoth Siva, Mr. Srinath Manivannan,
his professional mentors—Mr. Muralikrishnan and Dr. Annamalai for their constant
encouragement and inputs during the course of the project.

Digital Thread Enabled Manufacturing Automation for Mass Personalization iii

 Table of contents
Declaration of authorship i
Executive summary ii
Acknowledgements iii
List of Figures v
List of Tables v
List of Abbreviations vi
1 Introduction 1
1.1 Need for Mass Personalization 1
1.2 Need for a manufacturing automation framework for Mass Personalization 3
1.3 Project objectives 3
1.4 Project thesis methodology 4
1.5 Project report outline 4
2 Mass personalization and its challenges 5
3 Digital thread enabled manufacturing automation framework 7
4 Technological capabilities for digital thread information transfer 10
4.1 Technological capabilities for information transfer in customer co-creation 10
4.2 Technological capabilities for information transfer in manufacturing execution 11
4.3 Technological for information transfer in commissioning and after-sales service 12
5 Self-organized manufacturing for Mass Personalization 15
6 Case study 22
7 Proposed framework as a solution for challenges 26
8 Conclusion and scope for future advancements 28
References 29

Digital Thread Enabled Manufacturing Automation for Mass Personalization iv

 List of Figures
Mass Personalization application example: Custom made turbo charger housings for
Figure 1 older vehicles, 1
Figure 2 Nike ID Co-creation platform 2

Figure 3 Bike from Kennedy City bicycles 2

Figure 4 Evolution of different production paradigms over last century. 5
Figure 5 Challenges posed by Mass Personalization in different engineering stages of a product 6
Figure 6 Digital thread enabled manufacturing automation framework for Mass Personalization 7

Figure 7 ANSI/ISA-95 framework for manufacturing automation 8

Figure 8 Information flows in the digital thread between different digital twins 9
Figure 9 Technological capabilities for effective customer co-creation 10
Figure 10 Technological capabilities for manufacturing execution 11
Figure 11 Shop-floor communication architecture for Mass Personalization 14
Information transfer architecture for installation, commissioning and
Figure 12 after-sales service 14
Figure 13 Multi agent self-organized shop-floor framework 17
Figure
14i &14ii Multi-agent production scheduling algorithm 21
Figure 15 Interactive design platform for co-creation 23
Figure 16 Rule based classification algorithm for product features 23
Figure 17 “As-Designed Digital Twin” containing product data- BoM and CAD model 24
Figure 18 “As-planned/As-built Digital Twin containing process information 24
Figure 19 After-sales support platform with data from “As-Maintained Digital Twin” 25

List of Tables
Table 1 Comparison between different production paradigms 6
Table 2 Different shop-floor agents, their role description and attributes 20
Table 3 Proposed framework as a solution for challenges of mass personalization 27

Digital Thread Enabled Manufacturing Automation for Mass Personalization v

 List of Abbreviations
CAD Computer Aided Design
PDM Product Data Management
AI Artificial Intelligence
NLP Natural Language Processing
MQTT Message Queuing Telemetry Transport
IoT Internet of Things
M2M Machine to Machine
PLC Programmable Logic Controller
SCADA Supervisory Control And Data Acquisition
PAC Programmable Automation Controllers
ANSI American National Standards Institute
ISA International Society for Automation
PDM Product Data Management
OEE Overall Equipment Effectiveness
UI/UX User Interface/ User eXperience
CAE Computer Aided Engineering
MES Manufacturing Execution System
XML eXtensible Markup Language
RDF Resource Description Framework
Ctot Total Cost of manufacturing
Ttot Total Time of manufacturing
SKU Stock Keeping Unit

Digital Thread Enabled Manufacturing Automation for Mass Personalization vi

 1. Introduction
On the operations front, personalization can
With increased competition, higher levels of result in product differentiation at lower costs,
consumerism and technological disruptions, help in reduction of inventory through aligning
Organizations face a organizational by reducing the need for many
lot of challenges of stock keeping units (SKUs) and can align
having to stand out production more realistically with customer pull
from competition [3][8].
through innovation and Mass personalization is a production strategy
adding more value to which can help in effective and efficient
their product offerings. realization of personalized products. Mass
Many organizations in personalization can facilitate production of
recent times are adopting to producing individualized products, while at the same time,
individualized products in order to stand apart achieving operational efficiencies of mass
from competition [1]. production.
Furthermore, with changing consumer attitude An application of mass personalization is given
over the last decades, personalized products and in figure 1.
services are fast gaining traction, with consumers
across all age groups wanting personalized
options and a significant proportion willing to
pay extra and wait for longer for receiving their
personalized products and services [2].

1.1. Need for Mass

Personalization
Personalization is a product differentiation
strategy which organizations can adopt to stand
out from competition and retain their competitive
advantages. Personalization can help in
improving brand loyalty, increase customer
satisfaction and increase manufacturing volumes
[3][4][5]. Across the world, as per different
market research surveys carried out among Figure 1: Mass Personalization application
different category of consumers, the demand for example: Custom made turbo charger housings
personalized product ranges from 36 % in select for older vehicles,
international markets like the United Kingdom to
63 % among, early technology adopters [6][7].

Digital Thread Enabled Manufacturing Automation for Mass Personalization 1

However, there have been only a few successful
deployments of mass personalization in
manufacturing and many examples are from the Nike
domain of apparel and fashion. A successful  Design your own pair of shoes
 Early pioneer of personalization since 1999
example of an organization which has adopted
 Co-creation through website and mobile and
mass d personalization is Nike (figure 2), whose
Nike-ID studios.
product- Nike ID facilitates customer co-creation
of shoes through an interactive platform enabled  Benefits for the brand:
through internet and mobile applications and
through specialized design studio [9][10].  Huge options without increase in
inventory.
Other examples of customer led mass
personalization include Argos, a British  Premium pricing and higher profits
catalogue retailer, who has started to
manufacture personalized jewelry in small
batches through 3D printing , with customer co-
creation through a web based UX and Kennedy
City bicycles (figure 3) , who had used an
interactive decision tree for eliciting customer
needs [11][12].

Kennedy City Bicycles

 Active in UK -2012- 2016

 Completely seamless online

experience- no physical stores

 Employed decision trees to help

gather customer requirements.

 Was selling 30 unique and

personalized bicycles per month.

Figure 2 and 3 (top to bottom): Nike ID co-creation platform and a bike from Kennedy City bicycles

Digital Thread Enabled Manufacturing Automation for Mass Personalization 2

1.2. Need for a manufacturing Cyber physical organizational representation,
coupled with technologies such industrial
automation framework for automation, machine to machine communication,
Mass Personalization increased cognition and predictive demand
modelling can serve as an enabler for Mass
For organizations to successfully adopt Mass Personalization [16].
Personalization, many operational challenges
need to be overcome. These include lack of However, currently, a flexible, yet robust and
sufficient flow of product information across the resilient engineering management systems for
organization, forecasting difficulties due to Mass Personalization has not been developed.
constant changes in bill of materials, need for Current developments in the domain of Mass
effective interface to facilitate customer co- Personalization is centered around different
creation and effective manufacturing operations methods which can elicit customer needs and
management, actualize individualized product enable product co-designing
with unique process flow changes easily.
Therefore, in this project report, a digital thread
Information technology and manufacturing based manufacturing automation and operations
automation can improve interconnectivity framework for establishing an engineering
between different manufacturing entities and management systems for mass personalization
departments in an organization, thus solving has been proposed. The different capabilities
some of the fundamental challenges posed by which manufacturing organizations have to
Mass Personalization. An initiative which aims develop, for implementation of this proposed
to improve efficiency in manufacturing framework have also been identified and
organizations is Industry 4.0, mooted by German discussed.
government in 2011[13]. 1.3. Project objectives
Industry 4..0 aims to facilitate improved The main motivation of this project is to develop
information transfer by creating digital a digital thread based manufacturing automation
representations of physical systems , also known framework which can help in realizing flexible
as digital twins, which can capture all critical engineering management systems and automated
product and process information through data business processes for mass personalization. This
acquisition from physical resources [14]. project also seeks to identify different
technological capabilities for adoption of such an
These digital twins, would be able to automation framework for Mass Personalization.
automatically exchange data between different
manufacturing entities, through an The objectives of this project are:
interconnected information network known as
 To carry out a qualitative analysis of
“digital thread”, resulting in creation of a available literature and case studies in the
physical production system with virtual domain of mass personalization and
interfaces, also known as “Cyber-Physical identify crucial challenges faced by
production system” [15]. different departments of a manufacturing

Digital Thread Enabled Manufacturing Automation for Mass Personalization 3

 organization while adopting mass created after identifying different technological
personalization. capabilities required for effective information
 To develop a robust operational flow and shop-floor responsiveness.
framework for manufacturing automation,
incorporating on digital twins and digital The operational framework created were
thread, so as to enable flexible validated by means of a real world case study, to
engineering management systems which explain the suitability of the framework to an
can overcome critical challenges posed Engineer to Order enterprise manufacturing
by Mass Personalization.
personalized construction fittings for the
 To identify critical capabilities with
construction industry.
respect to information transfer, data
exchange and in the shop-floor for robust 1.5 Project report outline
implementation of the digital thread This report describes in detail the research
based manufacturing automation carried out in order to achieve the stated
framework. objectives. The remainder of the report is
 To explain a practical application of the organized into chapters as follows:
operational framework by means of a Chapter 2 discusses the paradigm of Mass
real world case study. Personalization and the challenges involved in
adopting it
1.4. Project methodology Chapter 3 presents a composite framework for
Firstly, the project objectives were finalized, Mass Personalization based on digital thread and
after identifying the key gaps, in the domain of manufacturing automation
Mass Personalization, from an anaysis of The different technological capabilities required
available literature. for seamless information through the digital
thread have been discussed in chapter 4.
Existing research predominantly focusses only Chapter 5 discusses the concept of self-organized
on different methods of eliciting customer’s manufacturing and corresponding scheduling
affective and cognitive needs and not on a methodologies which are a crucial part of the
complete engineering management systems for proposed digital thread based automation
mass personalization. framework.
A practical application of the proposed
Once the gap was identified, an exhaustive framework–through a real world case study of a
literature survey was carried out to analyze manufacturing enterprise engaged in Made to
relevant practices with respect to digital threads Order construction fittings has been explained in
and digital twins, manufacturing execution chapter 6.
systems, ERP, information flows in Chapter 7 discusses how the digital thread based
manufacturing entity, reconfigurable manufacturing automation framework and the
manufacturing systems. different capabilities identified can overcome the
challenges posed by Mass Personalization.
From the literature survey carried out, a robust Chapter 8 concludes the report and discusses the
operational framework for shop-floor automation scope for future advancements in the domain.
involving digital twins and digital threads was

Digital Thread Enabled Manufacturing Automation for Mass Personalization 4

2. Mass personalization and its challenges
3. challenges
Over the last 100 years, manufacturing has customers do not actively take part in product co-
transformed from push type Mass Production to designing but participate only in selection of
pull type Mass Personalization, as explained in different products.
figure-4 [17].
The advent of flexible manufacturing systems,
lean manufacturing methodologies such as
Kanban systems, rapid and quick die change
methodologies for faster set up and changeover
allowed for effective mass customization [18].

In mass personalization, customers are actively

involved in product co-creation to design
products which incorporate their latent, affective
and cognitive needs [19]. Each product is custom
made, with different bill of materials and
different process routings, along with custom
made work orders [20] [21].
___________________________________
“Effective realization of mass
Figure 4: Evolution of different production
personalization would pose several
paradigms over last century [17]. operational challenges for different
departments of the organization.”
Mass Production, which originated in the early
20th century, focused on building economies of
Table-1 summarizes the key differences between
scale, at low cost, with less product variety and
the different production paradigms.
with dedicated production lines. The main focus
Effective realization of mass personalization
of Mass Production was to manufacture products
would result in several operational challenges
in high quantities, with little product variety
across different departments of the organization
through dedicated production lines. . Mass pro
and there is a need for effective information
transfer and engineering management systems
“Over the last 100 years, across the organization.
manufacturing has evolved from The same has been described in figure-5.
mass production to mass
customization to mass
personalization”
In order to meet the needs of the customers,
manufacturers started to adopt Mass
Customization and offer increased product
options through configurable and custom made
product feature modules. In Mass Customization,

Digital Thread Enabled Manufacturing Automation for Mass Personalization 5

Characteristic Mass Production Mass Customization Mass Personalization
Batch size Large Small to Medium Small and individualized
batches
Product architecture Common product platform A common Product Open product architecture with
and interchangeable parts architecture with custom custom made/configurable
made and configurable modules and personalized
modules [22] modules
Production method Dedicated production line Flexible manufacturing Re-configurable cyber physical
systems, and lean production systems [24]
manufacturing.
Degree of customer Buying Configuring and selection Customer Co-creation
involvement of product [23].
Efficiency Through realizing Through flexible batch Through effective utilization of
economies of scale production digital technologies to facilitate
information transfer and self-
organization. [25]

Table 1: Comparison between different production paradigms

Figure 5: Challenges posed by Mass Personalization in different engineering stages of a product

Production
Planning •Difficulties with purchase
•Evolving product structure scheduling due to demand
and ground rules •Timely knowledge of fluctuations
•Developing user interface process specific •Challenges with dealing with
platform for capturing requirements part change management
customer requirements •Need for production and supplier mangement
control strategies like
Engineering Kanban, batch sizing and Procurement
Design CONWIP

Commissioning
& installation •Retrieving product and
•Distinct process routings for process traceability
each product •Updated procedures and data
•Responsiveness for service history for each •Obtaining information
dynamic changes in product and procedures for
production & orders •Robust information product servicing and
Manufacturing & transfer and retrieval training personnel.
shop-floor mechanism. After sales
support

Digital Thread Enabled Manufacturing Automation for Mass Personalization 6

3. Digital thread enabled manufacturing
automation framework
In this section, a holistic automation framework .
for Mass Personalization which takes into The main feature of this operation framework is
account key technological capabilities across the digital thread, which enables seamless
different departments of a manufacturing information transfer and data exchange across the
organization has been presented. The aim of this different departments of an organization. The
framework is to improve information flows different digital replicas of the individualized
between design, planning, manufacturing, product, known as digital twins are connected by
installation and commissioning a product and the digital thread.
hence realize superior operational efficiencies.
When a product is co-designed by the customer,
The digital thread based automation framework an “As-Designed Digital Twin” is established in
has been presented in figure 6. the environment of a CAE software. The “As-

Figure 6: Digital thread enabled manufacturing automation framework for Mass Personalization

Digital Thread Enabled Manufacturing Automation for Mass Personalization 7

Designed Digital Twin” contains product data As a result, the shop-floor resources should be
such as product geometry, BoM and product self-aware and employ decentralized
quality requirements. communication methodologies [29] [30].
In the shop-floor, automation framework is Self-organized manufacturing systems, which
established through a Central Agent Control exhibit a great degree of self-awareness with
System, which coordinates with all resources. cognitive decision making abilities can be
The Central Agent Control system employed to enable decentralized shop-floor
communicates with the ERP system for decision making. In this project, such a
performing initial task allocation for different manufacturing system enabled through the
shop-floor agents and receives product principles of multi-agent manufacturing, based
information contained in “As-Designed Digital on ANSI/ISA-95 shop floor automation
Twin, through an automated ERP-CAD bridge”. framework has been employed. Figure 7 presents
The BoM structure stored in the “As-Designed the ANSI/ISA-95 shop-floor automation
Digital Twin” is separated into off the shelf and framework [31] [32].
in-house manufactured parts by the ERP system, Each shop-floor resource is classified into an
which creates automated workflows such as agent with pre-defined roles and responsibilities,
purchase orders. A manufacturing BoM is also such as carrying out machining operations,
created from the design BoM, so as to account product transportation and inspection.
for shop-floor level product movements such as At the lowermost level of the shop-floor
WIP. automation framework, the product is
The Central Agent Control System creates an represented as a product agent which is
optimum production schedule and a process plan connected to the different digital twins. An “As-
is produced for each individualized products Planned Digital Twin” is realized after
A shop-floor involved in manufacturing transferring through the digital thread, the
individualized products faces challenges such as product data stored in “As-Designed’ Digital
production re-scheduling due to delays in Twin” with unique process routings and
procurement of individualized components, production schedules for every individualized
changing order priorities etc. In such a scenario, product.
a shop-floor should be reconfigurable and
responsive to changes [26] [27] [28].

Figure 7: ANSI/ISA-95 Framework for manufacturing automation [32]

Digital Thread Enabled Manufacturing Automation for Mass Personalization 8

The different digital twins are realized system is established in the digital thread as
progressively, as the product passes through the follows:
self-organized shop-floor and is shipped after it  The “As-Designed Digital Twin” receives
has been sold. The creation of different digital information regarding manufacturing
twins is described as follows: capabilities, from the “As-Planned Digital
 “As-Planned Digital Twin” evolves into Twin”. This feedback can be incorporated
the “As-Built Digital Twin” after the into the co-designing process for
manufacturing of product in the shop- producing products with better
floor. This digital twin would contain manufacturability.
crucial process information such as  The “As-Planned Digital Twin” receives
Manufacturing BoM and process information regarding machine downtime
parameter data. from the “As-Built Digital Twin”. This
 An “As-Maintained Digital Twin” is can be utilized for creating updated
created after coupling the product production schedules..
information stored in the “As-Designed  The information regarding product field
Digital Twin” and the process performance and service history are
information contained in “As-Planned transferred over the digital thread from
Digital Twin” and “As-Built Digital twin” the “As-Maintained Digital Twin” to the
with tailor made service and installation “As-Designed Digital Twin” for realizing
procedures. better product designs.
The feedback system through the digital thread is
presented in figure 8.
At the same time, a simultaneous feedback

Figure 8: Information flows in the digital thread between different digital twins

Digital Thread Enabled Manufacturing Automation for Mass Personalization 9

4. Technological capabilities for digital thread
information transfer
This chapter presents different technological However, to facilitate co-designing, the User
capabilities which facilitate consistent and interface for co-creation must incorporate the
efficient data exchange across the mass following (represented by figure 9):
personalization digital thread. This chapter is  Classification of Product features into
structured around different engineering stages basic and personalized modules on the
involved in realizing a personalized product, basis of a part function, product hierarchy
from co-designing to manufacturing to after sales and manufacturing complexity. A weight
support. based classification index can be used for
this purpose [34][35][36].
4.1. Technological capabilities for  Determination of customer’s affective
information transfer in customer and cognitive needs through citarasa
descriptors and through recommendation
co-creation engine aided by machine learning
Customer co-designing of a product can be algorithms [33].
enabled through an interactive UI/UX software,
 Superior visualization and graphics
which incorporates a co-designing process.
capabilities, in order to facilitate co-

Figure 9: Technological capabilities for effective customer co-creation

Digital Thread Enabled Manufacturing Automation for Mass Personalization 10

 creation through free sketch CAD 4.2. Technological capabilities for
modelling.
 Rule based design constraints for creating
information transfer in
correct part interfaces based on manufacturing execution
manufacturing constraints [22] [24].
The Information technology capabilities for
 A module which interfaces in real time
manufacturing execution in Mass personalization
existing PDM package to facilitate access
are explained in the figure 10.
to product design libraries and create
design BoMs The product information contained in the PDM
system would need to be parsed by the ERP
The co-created product along with relevant system for identifying in-house manufactured
product information forms the “As-Designed and off the shelf parts.
Digital Twin” of the product”. Most often, ERP systems and CAD Packages are
two separate stand -alone entities, resulting in a
increased manual work-flows and hence
inefficiencies, in a mass personalized
environment, where each order is unique and
custom made [37][38] .

Figure 10: Technological capabilities for manufacturing execution

Digital Thread Enabled Manufacturing Automation for Mass Personalization 11

Automatic integration of ERP systems with CAD  Equipping manufacturing agents with
Packages through a service mapping framework cognitive decision making abilities
involving semantic web ontologies realized delivered through AI and ML algorithms
through XML and OWL is capability which  Adopting intelligent scheduling
organizations can adopt to improve efficiencies algorithms which make use of predictive
[39]. data modelling and real time data
“ERP Systems also need to acquisition from machines through data
loggers and sensors.
incorporate a service oriented
 Open ended and extensible machine
architecture, which can integrate programming beyond traditional
data from different shop-floor PLC/SCADA—enabled by technologies
resources, provide business such as Programmable Automation
intelligence and data visualization Controller for easier integration of
functions and be equipped with decision making and information transfer
capabilities.
ability to be hosted on cloud for
rapid deployment and scalability.” 4.3. Facilitators for
_________________________________ information transfer in
Such integration would be useful in driving
installation, commissioning
processes such as custom made workflows and
purchasing orders based on lead times, quality and after-sales service
assurance and reliability [40][41][42]. Retaining product and process
Manufacturing agents should be self-aware, information is vital for product
equipped with superior machine to machine traceability and warranty concerns
communication capabilities and increases self- management and for curating tailor-made
awareness An information transfer architecture installation, maintenance and service
in line with ANSI/ISA-95 framework is procedures, once a product has been
presented in figure 11 [32]. manufactured and shipped.
Information transfer in the shop-floor for _____________________________
effective self-awareness and self-organization “Installation and servicing
can be enabled through: often involves manual
 Real-time shop-floor condition procedures involving human
monitoring and machine condition knowledge and expertise,
monitoring through data loggers and
which is often captured in
sensors.
text-based documents, with
 Superior machine to machine
communication, enabled through technical terminologies,
technologies such as IOT and MQTT leading to service
protocols. inconsistencies.”

Digital Thread Enabled Manufacturing Automation for Mass Personalization 12

There is also a need for tailor making unique
servicing and installation procedures to render
consistent quality of service [43][44].

Hence, the installation, commissioning and

service management team of a manufacturing
enterprise must develop capabilities of:
• Natural Language processing tools, along with
Artificial intelligence and Machine learning
methodologies which can capture unstructured
data in text documents [45][46]
• Automated Service Request Management
Engine which can interface with PDM and ERP
packages to retrieve latest product and process
information.

Figure 12 has presents the Information transfer

architecture for installation, commissioning and
after-sales service management.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 13

Figure 11: Shop-floor communication architecture for mass personalization
Figure 12: Information transfer architecture for installation, commissioning and after-sales service

Digital Thread Enabled Manufacturing Automation for Mass Personalization 14

 5. Self-organized manufacturing
“conditions infor Mass
order to achieve
Personalization optimal performance.

This chapter discusses the key technological Self-organization as a whole, aims at increasing
capabilities that are to be developing for a self- autonomy in the shop-floor through decentralized
organized shop-floor for mass personalization. decision making methodologies [47].
_________________________________ A Self-organized shop-floor , along with robust
scheduling algorithms can help in minimizing
“For Effective manufacturing
downtime and improving efficiencies when
of personalized products, a manufacturing orders fluctuate in demand and
shop-floor needs to be re- are often in low quantities. A Self-organized
shop-floor can be realized through various
configurable, adaptable and
principles like holonic manufacturing systems,
situation dependent” bionic manufacturing systems, and multi-agent
_________________________________ manufacturing systems.
A shop-floor must cater to dynamic changes in The different principles are described below:
process schedules such as the dynamic batch
sizing, re-scheduling production orders to • In a Holonic manufacturing system, a
account for uncertainties in the shop-floor with manufacturing entity is treated as a holon, which
expediting urgent orders, delays in procurement is an autonomous, co-operative and intelligent
of material on time. entity taking part in autonomous and distributed
__________________________________ decision making [48][49].
“Self-organized manufacturing • In a Bionic manufacturing system, a
manufacturing entity is treated similar to a
is defined as a collection of
biological cells, which have their own individual
manufacturing entities ( such functions and yet can spontaneously respond to
as machines, material handling dynamic external events, by means of regulatory
mechanisms involving hormones and enzymes,
systems like forklifts,
which in a manufacturing system would be
operators , supervisory control information and materials flowing from one
centers and ERP systems) , manufacturing entity to another [50][51].
• In a Multi-agent manufacturing system, a
which can re-configure their
manufacturing system is treated as an
organizational structure and autonomous entity, each with defined set of
operational functions with activities and functions.[52][53][54]
minimal external interaction,
Thus in each of these governing principles, a
so as to adapt to rapidly manufacturing resource is treated as an
changing production” autonomous entity with a set of predefined roles
and can communicate with each other through
established communication practices and plan for

Digital Thread Enabled Manufacturing Automation for Mass Personalization 15

dynamic changes in a concurrent fashion with The different shop-floor resources have been
different levels of established hierarchy. classified into manufacturing agents, on the basis
Shop-floor control can be established by means of ANS/ISA-95 shop-floor architecture and
of regulatory mechanisms involving genetic and presented in figure-11. The physical production
heuristic based algorithms, machine learning and process is represented by the product agent
reinforcement learning. which is created every time a new order is
Few examples of self-organized manufacturing received and for every individual components
control have been established for Engineer to which constitute a personalized batch.
Order Industries, across both academia and in
industries and the same have been summarized in
the below.
A similar production scheduling algorithm has
been explained in this section, as a possible
method for production control for self-organized
shop-floor for mass personalization.

JSC Kuznetsov, University of

Russia Tool & die Michigan 3D
manufacturing- Printing
 Real time workshop- China laboratory
scheduling system  By researchers of  Established a Self-
for a factory Shenzen university and organized
manufacturing jet The University of manufacturing
engine parts. Hong Kong.
framework for
 Employed a multi-
 Expected machining, 3D
agent mechanism
improvement in involving smart printing and assembly
operational manufacturing agents customized plastic
efficiencies and connected by RFID parts [55].
percentage orders tags.  Adopted a Production
completed without  Used a multi-period as a Service
delay by 15-30% hierarchical scheduling framework involving
[56]. mechanism for a shop- service oriented
floor producing dies architecture delivered
 Employed a and molds. through multi-agent
constraint based  Significantly improved methodologies.
algorithm- efficiency &  Employed Genetic
dynamically compare coordination between
each and every algorithm based
shop-floor and line
manufacturing shop-floor scheduling
managers [57].
resource on basis of and control
priority rules. mechanism.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 16

 Figure 13: Multi-agent self-organized shop-floor framework
 Monitoring of other manufacturing
The product agents represent a digital replica of entities for downtime, machine
the product and contain product information parameters, fixture set up.
which serves as an input for different digital  Creation of initial process routes, which
twins. are transferred onto the manufacturing
The manufacturing processes, at level-1 and level agents.
2 of the ANSI/ISA-95 automation framework,
are represented by a network of smart At the business level is the ERP system, which
manufacturing agents such as machining agents, communicates with CAD/PDM system to create
conveying and inspection agents. Manufacturing automated work flows for each product, parses
agents are equipped with cognitive decision the product BoM into in-house and brought out
making abilities and can communicate with each parts and updates product information and drives
other through machine to machine other business processes. The efficiency of the
communication protocols. shop-floor is measured by overall equipment
At level 3, manufacturing operations effectiveness as given below[58]
management is ensured through a Centralized
Agent Control System, which forms a bridge 𝑂𝐸𝐸 =
between various business processes and shop- 𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑋 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑋 𝑄𝑢𝑎𝑙𝑖𝑡𝑦 [59]
floor such as: 𝐴𝑐𝑡𝑢𝑎𝑙 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒
𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑖𝑙𝑖𝑡𝑦 = 𝐴𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 𝑋 100
 Liaising with the ERP system for daily
(𝑁𝑒𝑡 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒)
sales plan uploaded by the sales and 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 = (𝐴𝑐𝑡𝑢𝑎𝑙 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒) 𝑋 100;
marketing team and also to acknowledge
production completion and material
𝑄𝑢𝑎𝑙𝑖𝑡𝑦
receipt in order to drive other business
𝑇𝑜𝑡𝑎𝑙 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑜𝑓 𝑑𝑒𝑓𝑒𝑐𝑡 𝑓𝑟𝑒𝑒 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠
processes such as procurement . = 𝑋 100
𝑇𝑜𝑡𝑎𝑙 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑜𝑓 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠

Digital Thread Enabled Manufacturing Automation for Mass Personalization 17

 Machine downtime is quantified as the sum of • The central agent control system
planned and unplanned losses, which are as follows : determines the initial process routing for each
𝐷𝑜𝑤𝑛𝑡𝑖𝑚𝑒 = 𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 product, which is to be manufactured, and
+ 𝑈𝑛𝑝𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 determines the agent combination set.
𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒
• The product agents are assigned to a
= 𝐶ℎ𝑎𝑛𝑔𝑒𝑜𝑣𝑒𝑟 𝑡𝑖𝑚𝑒
+ 𝑄𝑢𝑎𝑙𝑖𝑡𝑦 𝑖𝑛𝑠𝑝𝑒𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 particular machines in the combination set on the
+ 𝑃𝑙𝑎𝑛𝑛𝑒𝑑 𝑚𝑎𝑖𝑛𝑡𝑒𝑛𝑎𝑛𝑐𝑒 𝑡𝑖𝑚𝑒 basis of series of bids and auctions and also
𝑈𝑛𝑝𝑙𝑎𝑛𝑛𝑒𝑑 𝑑𝑜𝑤𝑛𝑡𝑖𝑚𝑒 depending on the real time information status of
= 𝐸𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 + 𝑏𝑟𝑒𝑎𝑘𝑑𝑜𝑤𝑛 each machine, such as downtime, the fixture
+ 𝑡𝑜𝑜𝑙 𝑓𝑎𝑖𝑙𝑢𝑟𝑒𝑠 running in the machine, previous orders which
have been assigned earlier to the machine and the
machine parameters.
The algorithm for enabling production planning
and process routings for mass personalization is • The central agent control system sends
mentioned in the figure 14. out a call for each of the machines in the
combination set for actualizing first operation.
For any product in part family, there are “n” • An algorithms focused on minimizing
machines which can perform the required total cost of production or the total time of
operation. An iterative bidding mechanism is manufacturing is employed to determine process
used by the central agent control system and each
routing and the same is mentioned in figure 14i
smart manufacturing entity, to determine process
routing. This mechanism is explained as follows: & 14ii.

A new product instituted by the central agent The total cost of manufacturing a particular
control system requires n operations, represented product would be comprised of:
by
𝑂𝑃 = {𝑂𝑃1, 𝑂𝑃2, 𝑂𝑃3, … . , 𝑂𝑃𝑛} Machine cost: This is proportional to the time
taken by the machine and the unit power
The shop-floor has a number of manufacturing
agents, conveying agents and inspection agents, consumption - depends on machine parameters
which are denoted by: and also machine specifications (such as a
Set of inspection agents: 𝐼 = {𝐼1, 𝐼2, … , 𝐼𝑛} hydraulic drive or an electrical drive).
Set of conveying agents: 𝐶 = {𝐶1, 𝐶2, … , 𝐶𝑛}
Set of machining agents: 𝑀 = {𝑀1, 𝑀2 … , 𝑀𝑛} Tooling cost: Cost of cutting tools and fixtures.
For example, one cutting tool or an insert can be
For any product, the actual combination set of
utilized to produce a few similar components,
agents required is denoted by:
A1⊂A before being scrapped or sent for a re-grind. In
Machines are selected in such a way that a such a case, the contribution of the tool to the
particular operation can be performed in more product cost tool. Similarly, the cost contribution
than one machine, so as to ensure that each of the of fixtures and tooling is fractionally proportional
operations can be completed, in case of any to the total number of components manufactured
unplanned downtimes. with the tool. Similarly, the cost contribution of
The manufacturing control is worked out as the fixtures divided by the estimated useful life.
follows:

Digital Thread Enabled Manufacturing Automation for Mass Personalization 18

Conveying cost: Cost of transportation from one machining agent, after monitoring individual
machining agent to another or the cost of machine status.
transportation. This is proportional to the power The exact combination set may vary from the
consumption expounded in the process of original resource combination, based on real-time
transportation. machine status.

Cost of downtime: Cost of downtime is the Thus, the given scheduling algorithm provides a
cumulative cost involved in restoring a methodology by which the need for accounting
manufacturing agent to its original state and for dynamic changes in process routings arising
additional cost, if any, for the time delay due to variations in product structure and also
involved. ( such as machining cost due to problems such as machine breakdown, quality
expedition). defects and expedition of urgent orders.
Each of the machine in the combination set sends
out bids consisting of data regarding cycle time
and total manufacturing cost.
The central agent control system selects a
machining agent for the first operation, from the
bids in such a way that the total cost is minimal
or the total manufacturing time is minimal ( in
case products need to be expedited ).

Min( Ctot=∑Ci)
Min( Ttot=∑Ci)

The central agent control system determines an

approximate machine combination set and
process routing for each of the products being
manufactured, within a particular time frame, by
employing an iterative approach which analyzes
all the possible machine combinations, to
determine the lowest cost or lowest time.

On completion of first operation, the product

agent is updated with process information such
as machine parameters and product traceability
details.

The machine agent which has carried out the first

operation calls for bids again, using a similar
methodology and assigns components to the next

Digital Thread Enabled Manufacturing Automation for Mass Personalization 19

Agent Role Description Attributes
Product Represents a product, contains Product geometry, dimensions
product information and is generated and tolerances, quality
for each product being manufactured requirements, manufacturing
requirements, BoM,

Machines Machines are smart entities which Machine capability, set up

are responsible for realization of requirements , machine
smart products from WIP. parameters, energy
Machining agents can take orders consumption, cycle time ,
from other machining system and communication methodology
also from centralized agent control
system

Conveying Systems Conveying agents perform in Load capacity, speed,

transportation and material handling throughput, application, power
functions. consumption
These include conveyors, AGVs,
smart forklifts and smart workers

Inspection Agents Inspection agents perform inspection Device Resolution, type, data
operations- these include vision acquisition method, feedback
system, tool setting devices, CMMs systems, Repeatability

Central Agent Control System Automated bridge between shop- Scheduling algorithms,
floor and ERP System Communication and data
Responsible for creation of initial acquisition method
process routings and process
schedules, monitors machine status

ERP System Drives business processes Modularity & Service

Enables information sharing across Composition, scalability and
all verticals extensibility with business
Performs control at business level intelligence

Table 2: Different shop-floor agents, their role description and attributes

Digital Thread Enabled Manufacturing Automation for Mass Personalization 20

Figure 14 i & 14 ii: Multi-agent production scheduling algorithm

Digital Thread Enabled Manufacturing Automation for Mass Personalization 21

6. Case study
In this section, the proposed framework is Interfacing with a CAD/PDM package to arrive
validated by means of a real life case study at design bill of materials and information about
which involves manufacturing personalized product geometry.
aluminium fins, external cladding, balustrades
and louvres used for building facades. Figure-15 represents an interactive design
___________________________________ platform which can facilitate customer co-
“Since each construction creation. The set of parametric rules which are
used for categorizing products into personalized,
project is unique and custom configurable and basic product architecture is
made, each order is presented in figure 16.
personalized and is
From the “As-Designed Digital Twin”
manufactured in varying batch (represented by figure-17), the centralized agent
sizes” control system parses the preliminary BoM
___________________________________ information into in-house and bought out parts
and loads the details onto the ERP system.
At the same time, the products would be having a Furthermore, the centralized agent control system
basic open product architecture based on also liaises with the ERP system to create process
calculations such as wind load and other design plans and production schedules for the decision
constraints. The manufacturing enterprise would making in self-organized shop-floor.
also have a few constraints such as
organization’s manufacturing and sourcing The different manufacturing entities can interact
capability, choice of materials, sectional profile, with each other through a series of bids and
size and colour. auctions to create a dynamic production schedule,
considering demand, shipping date and machine
In this case study, Customer co-creation is specification. A new manufacturing BoM can be
facilitated through an interactive product design created, considering parts which can be
interface which allows customers to make substituted and also in order to take into account
selection of basic design, colour and size of the work in progress, which can be denoted by
materials. Customers would be able to visualize dummy part numbers.
the product through a 3D rendering package
which would also allow customers to configure The individual process plan, process parameters
their products on the basis of a series of and part traceability data along with
personalized modules. manufacturing Bill of Materials

From the 3D rendering created, an “As-Designed

Digital Twin” can be created, after

Digital Thread Enabled Manufacturing Automation for Mass Personalization 22

Figure 15: Interactive design platform for customer co-creation

Figure 16: Rule based classification algorithm for product features

constitute the “As-Planned Digital Twin” of the through a 3D rendering interface can be
product. accessed by the commissioning and after sales
“As-Planned Twin”, along with real time service team through the digital thread to create
production information such as time of tailor made product installation procedures and
processing and operator name forms the “As- service instructions.
Built Digital Twin”.
The “As-Planned/AS-Built Digital Twin”
(represented by figure-18), stored in the ERP
system and the “As-Designed Digital Twin”
stored in the PDM software and visualized

Digital Thread Enabled Manufacturing Automation for Mass Personalization 23

Figure 17: “As-Designed Digital Twin” containing product data- BoM and CAD model

Figure 18: “As Planned/As-Built Digital twin” containing process information

Digital Thread Enabled Manufacturing Automation for Mass Personalization 24

These would serve as the “As-Maintained Digital repositories for building upon product knowledge
Twin” for the product, represented by figure 19. base, to actualize better product design.
A service request management engine, also
represented in the figure 16, has been created, For example, the data regarding failure at
after incorporation of AI/ML and NLP mounting points and bolt-failure at critical points
procedures. This would serve as a robust, up-to is sent back to the “As-Designed digital twin for
date, singular and consistent source of handling arriving at an updated personalization option set.
service requests for personalized products.
A closed loop feedback system is thus
established throughout the product lifecycle,
from the point of co-designing to manufacturing
and servicing of a tailor made product, as a result
of storing all product and process information as
a digital twin. The service history and product
performance data can be transferred to the “As-
Designed Digital Twin” and stored in PDM

Figure-19: After sales support platform with data from “As-Maintained Digital Twin

Figure 19: After-Sales support platform with data from “As-Maintained Digital Twin”

Digital Thread Enabled Manufacturing Automation for Mass Personalization 25

 The cyber-physical representations of a product
7. Proposed framework as aimplemented
can be solution in realityfor the
through a
reconfigurable shop-floor with self-aware and
challenges
Mass personalization is a production strategy self-organized manufacturing entities.
which aims to bring efficiency and effectiveness
similar to mass production, in manufacturing A shop-floor with a higher degree of self-
personalized products. Personalized products are organization, however, needs to incorporate a
typically manufactured in Engineer to order robust scheduling algorithm, which integrates:
industries. • Regulatory mechanisms for control,
coordination and production scheduling,
Engineer to order Industries manufacturing involving different operational entities.
personalized products suffer from historically • Knowledge base also known as belief
low levels of productivity, arising due to system, containing product history and technical
scheduling complexities, uncertainties and information regarding shop-floor entities such as
difference in BoM between every order [60][61]. machines and conveyors.
Hence, for organizations to adopt mass
personalization as a differentiation strategy, One such possible algorithm which can be
operational challenges across the domains of adopted is presented in chapter-4.
product design, manufacturing planning, demand
forecasting, manufacturing operations This algorithm takes into account the operating
management, field services and after sales cost for each production resource, the orders
support have to be overcome. available in hand and machine downtime,
The digital thread and digital twin based modelled using system reliability equations.
operational framework which has been explained
in this project, along with the identified Table 3 presents how challenges in mass
technological capabilities for efficient and personalization can be overcome using the
effective information transfer in the shop-floor proposed framework.
can be utilized to overcome these key operational
challenges.

The case study presented has demonstrated the

application of proposed framework by presenting
a cyber-physical representation which can store
product information and process information
virtually and can communicate with each other
through information technology protocols over
the digital thread.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 26

Organizational Vertical Challenge Solution
Engineering Design Evolving product structure and Hierarchical product feature and
ground rules. function based classification of parts
into basic, configurable and
personalized modules
Developing user interface platform Determination of customer needs
for capturing customer requirements through citarasa descriptors and AI
based recommendation engines.
CAD based software for co-creation.

Manufacturing Planning Timely knowledge of process Automatic Integration of ERP

specific requirements Packages with PDM packages
through service mapping by
semantic web ontological models
Need for production control Dynamic production scheduling
strategies like Kanban, batch sizing through self-organized self-aware
and CONWIP production entities.

Procurement Difficulties with purchase Dynamic demand planning by

scheduling due to demand central agent control system, and
fluctuations automatic linkage with ERP system.

Challenges with part change A single point of information in “As-

management and supplier Designed digital twin”, for
management singularity of information ,
communication between different
Digital Twins, through industry 4.0
framework
Manufacturing and Shop-floor Distinct process routings for each Process routings generated by self-
product aware manufacturing entities in the
shop-floor
Responsiveness for dynamic Dynamic scheduling algorithms
changes in production & orders aided by “Smart” Machines

Commissioning and Field Updated procedures and service Personalized installation and service
Service history for each product procedures through capturing human
. knowledge and through integration
with real time product and process
data.
Robust information transfer and Product and process information is
retrieval mechanism. stored in different digital twins and
procedures through semantic web
ontologies.
After Sales Support Retrieving product and process Product traceability provided
traceability data through “As-Planned/As-Built
Digital Twin” coupled with “As-
Maintained “ Digital twin.
Obtaining information and Product and process data stored in
procedures for product servicing and different digital twins.
training personnel.

Table 3: Proposed framework as a solution for challenges

Digital Thread Enabled Manufacturing Automation for Mass Personalization 27

8. Conclusion and scope for future advancements
In this project report, a robust digital thread based manufacturing automation framework which would
overcome critical challenges of Mass Personalization has been proposed. The different technological
capabilities with respect to information transfer and in the shop-floor, which can facilitate the
implementation of such a digital thread based manufacturing automation framework has been identified
and discussed. A practical application of this framework has been discussed through a case study of an
Engineer to order enterprise engaged in manufacturing individualized products for the construction sector.

The author believes that this work will act as a guide for organizations looking to adopt mass
personalization as a business model. The author also believes that this project would bridge the gap
existing with respect to research in mass personalization, which is predominantly focused on product
design, with lesser focus on the operational aspects of mass personalization. Scope for future
advancements exist with respect to development of different robust production scheduling algorithms for
the self-organized shop-floor and deploying them in real world. The feedback gained from a practical
implementation will be useful in contributing to further advancements in the domain of dynamic shop-
floor control.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 28

References
[1] Lu, Y. and Xu, X., (2018). “Resource virtualization: a core technology for developing cyber-physical
production systems”.Journal of manufacturing systems, 47, pp.128-140.
[2] Capegemini.(2018). The shift from mass production to mass personalization. Retrieved from
https://www.capgemini.com/gb-en/2018/05/the-shift-from-mass-production-to-mass-personalisation/
[3] Sourcing Journal (2019). Personalization is Everywhere Because Consumers Want it. Retrieved from
https://sourcingjournal.com/topics/consumer-insights/personalization-report-demand-margins-151680/
[4] Forbes (2019). Is Mass Personalization The Next Big Hurdle for Manufacturers?. Retrieved from
https://www.forbes.com/sites/jimvinoski/2019/10/17/is-mass-personalization-the-next-big-hurdle-for-
manufacturers/#46f770d97203
[5] Bain & Company (2013). Making it personal: Rules for success in product customization. Retrieved
from https://www.bain.com/insights/making-it-personal-rules-for-success-in-product-customization/
[6] The Deloitte Consumer Review. (2015). Made-to-order: The rise of mass personalization . Retrieved
from https://www2.deloitte.com/ch/en/pages/consumer-business/articles/made-to-order-the-rise-of-mass-
personalisation.html.
[7]. Forbes (2020). New Research shows consumers already expected mass personalization. Time to get
ready. Retrieved from https://www.forbes.com/sites/jimvinoski/2020/01/20/new-research-shows-
consumers-already-expect-mass-personalization-time-to-get-ready/#7d3df8d1223e/
[8] HP Megatrends (2017). From mass production to mass personalization. Retrieved from
https://hpmegatrends.com/from-mass-production-to-mass-personalization-1ad6746dc24d
[9] Sourcing Journal (2019). Apparel Brands Are Testing the Waters of Mass Customization. Retrieved
from https://sourcingjournal.com/topics/technology/personalization-report-investments-adaptations-
152409/
[10] Protolabs. (2019). Smart Manufacturing Enables Mass Customization Trend. Retrieved from
https://www.protolabs.com/resources/blog/smart-manufacturing-enables-mass-customization-trend/
[11] Wired. (2015). Argos sells customisable 3D printed jewellery. Retrieved from
https://www.wired.co.uk/article/argos-3d-printing-jewellery
[12] Kennedy City Bicycles (2020). The KCB Story. Retrieved from https://www.kennedycitybicycles.cc/
[13] Kagermann, H., Helbig, J., Hellinger, A. and Wahlster, W., (2013). Recommendations for
implementing the strategic initiative INDUSTRIE 4.0: Securing the future of German manufacturing
industry; final report of the Industrie 4.0 Working Group. Forschungsunion.
[14] Lu, Y., Liu, C., Kevin, I., Wang, K., Huang, H. and Xu, X., 2020. Digital Twin-driven smart
manufacturing: Connotation, reference model, applications and research issues. Robotics and Computer-
Integrated Manufacturing, 61, p.101837.
[15] Singh, V., & Willcox, K. E. (2018). Engineering design with digital thread. AIAA Journal, 56(11),
4515-4528.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 29

[16] Zhuang, C., Liu, J., & Xiong, H. (2018). Digital twin-based smart production management and
control framework for the complex product assembly shop-floor. The International Journal of Advanced
Manufacturing Technology, 96(1-4), 1149-1163.
[17] Hu, S.J., (2013). Evolving paradigms of manufacturing: from mass production to mass customization
and personalization. Procedia CIRP, 7, pp.3-8.
[18] Koren, Y. (2010). The global manufacturing revolution: product-process-business integration and
reconfigurable systems (Vol. 80). John Wiley & Sons.
[19] Wang, Y., Ma, H. S., Yang, J. H., & Wang, K. S. (2017). Industry 4.0: a way from mass
customization to mass personalization production. Advances in Manufacturing, 5(4), 311-320
[20] Zhou, F., Ji, Y., & Jiao, R. J. (2013). Affective and cognitive design for mass personalization: status
and prospect. Journal of Intelligent Manufacturing, 24(5), 1047-1069.
[21] Zheng, P., Yu, S., Wang, Y., Zhong, R.Y. and Xu, X., (2017). User-experience based product
development for mass personalization: a case study. Procedia CIRP, 63, pp.2-7
[22] Koren, Y., Shpitalni, M., Gu, P., & Hu, S. J. (2015). Product design for mass-individualization.
Procedia Cirp, 36, 64-71.
[23] Tseng MM, Jiao J, Merchant ME, (1996), Design for Mass Customization. CIRP Annals-
Manufacturing Technology 45(1): 153-156.
[24] Tan, C., Hu, S. J., Chung, H., Barton, K., Piya, C., Ramani, K., & Banu, M. (2017). Product
personalization enabled by assembly architecture and cyber physical systems. CIRP Annals, 66(1), 33-36.
[25] Wang, Y., Ma, H. S., Yang, J. H., & Wang, K. S. (2017). Industry 4.0: a way from mass
customization to mass personalization production. Advances in Manufacturing, 5(4), 311-320.
[26] Pourabdollahian, G., Taisch, M. and Tepe, G., (2013), September. Exploring different faces of mass
customization in manufacturing. In IFIP International Conference on Advances in Production
Management Systems (pp. 13-20). Springer, Berlin, Heidelberg.
[27] Davrajh, S. and Bright, G., (2013). Advanced quality management system for product families in
mass customization and reconfigurable manufacturing. Assembly Automation, 33(2), pp.127-138.
[28] Gilke, N., Mantha, S. and Thampi, G., (2013). Infrastructural Back Bone Of Enabling And
Converging Technologies For Mass Customization Manufacturing System In Automotive Industries.
International Journal for Quality Research, 7(1), pp.153-162.
[29] Essers, M.S. and Vaneker, T.H., (2014). Evaluating a data distribution service system for dynamic
manufacturing environments: a case study. Procedia technology, 15, pp.621-630.
[30] Schild, K. and Bussmann, S., (2007). Self-organization in manufacturing operations.
Communications of the ACM, 50(12), pp.74-79.
[31] Tharumarajah, A., (2003). A self-organising view of manufacturing enterprises. Computers in
Industry, 51(2), pp.185-196.
[32] Brandl, D. (2008). What is ISA-95? Industrial best practicesof manufacturing information
technologies with ISA-95 models. Retrieved from http://apsom.org/docs/T061_isa95-04.pdf

Digital Thread Enabled Manufacturing Automation for Mass Personalization 30

[33] Mavridou, E., Kehagias, D. D., Tzovaras, D., & Hassapis, G. (2013). Mining affective needs of
automotive industry customers for building a mass-customization recommender system. Journal of
intelligent manufacturing, 24(2), 251-265.
[34] Berry C, Wang H, Hu SJ (2013) Product Architecting for Personalization. Journal of Manufacturing
Systems 32(3):404–411.
[35] Kaul A, Rao VR (1995) Research for Product Positioning and Design Decisions: An Integrative
Review. International Journal of Research in Marketing 12 (4):293–320.
[36] Koren Y, Hu SJ, Gu P, Shpitalni M (2013) Open-Architecture Products. CIRP Annals—
Manufacturing Technology 62(2):719–729.
[37] Stojkić, Ž., Veža, I., & Bošnjak, I. (2016, January). Concept of information system implementation
(CRM and ERP) within industry 4.0. In 26th DAAAM International Symposium on Intelligent
Manufacturing and Automation.
[38] Basl, J. (2018, September). Analysis of industry 4.0 readiness indexes and maturity models and
proposal of the dimension for enterprise information systems. In International Conference on Research
and Practical Issues of Enterprise Information Systems (pp. 57-68). Springer, Cham.
[39] Lu, Y., Wang, H., & Xu, X. (2019). ManuService ontology: A product data model for service-
oriented business interactions in a cloud manufacturing environment. Journal of Intelligent Manufacturing,
30(1), 317-334.
[40] Andročec, D., Picek, R., & Mijač, M. (2018, January). The ontologically based model for the
integration of the IoT and Cloud ERP service. In Proceedings of the 8th International Conference on
Cloud Computing and Services Science, Funchal, Madeira, Portugal (pp. 481-488
[41] Koupaei, M. N., Mohammadi, M., & Naderi, B. (2016). An Integrated Enterprise Resources Planning
(ERP) Framework for Flexible Manufacturing Systems Using Business Intelligence (BI) Tools.
International Journal of Supply and Operations Management, 3(1), 1112.
[42] Ghazanfari, M., Rouhani, S., Jafari, M., & Taghavifard, M. T. (2009). ERP requirements for
supporting management decisions and business intelligence. IUP Journal of Information Technology, 5(3).
[43] R. Owen, I. Horváth. Towards product-related knowledge asset warehousing in
enterprises, Proceedings of the 4th International Symposium on Tools and Methods
of Competitive Engineering, TMCE 2002 (2002) 155–170.
[44] Z.Y. Wu, X.G. Ming, Y.L. Wang, L. Wang. Technology solutions for product lifecycle
knowledge management: framework and a case study, International Journal of Production Research. 52
(21) (2014) 6314–6334.
[45] Klein, Dan; Manning, Christopher D. (2002). Natural language grammar induction using a
constituent-context model. Advances in Neural Information Processing Systems.
[46] Chungoora, N., Cutting-Decelle, A. F., Young, R. I. M., Gunendran, G., Usman, Z., Harding, J. A., &
Case, K. (2013). Towards the ontology-based consolidation of production-centric standards. International
Journal of Production Research, 51(2), 327-345.
[47] Schild, K. and Bussmann, S., (2007). Self-organization in manufacturing operations.
Communications of the ACM, 50(12), pp.74-79.
[48] Botti, V., & Giret, A. (2008). Holonic manufacturing systems (pp. 7-20). Springer London.

Digital Thread Enabled Manufacturing Automation for Mass Personalization 31

[49] Valckenaers, P., & Van Brussel, H. (2005). Holonic manufacturing execution systems. CIRP annals,
54(1), 427-432.
[50] Tharumarajah, A. (1996). Comparison of the bionic, fractal and holonic manufacturing system
concepts. International Journal of Computer Integrated Manufacturing, 9(3), 217-226.
[51] Gu, W., Tang, D., Zheng, K., & Wang, L. (2011). A neuroendocrine-inspired bionic manufacturing
system. Journal of Systems Science and Systems Engineering, 20(3), 275.
[52] Ouelhadj, D., Hanach, C., & Bouzouia, B. (1998, May). Multi-agent system for dynamic scheduling
and control in manufacturing cells. In Proceedings. 1998 IEEE International Conference on Robotics and
Automation (Cat. No. 98CH36146) (Vol. 3, pp. 2128-2133). IEEE.
[53] Kouiss, K., Pierreval, H., & Mebarki, N. (1997). Using multi-agent architecture in FMS for dynamic
scheduling. Journal of intelligent Manufacturing, 8(1), 41-47.
[54] Guo, Q., & Zhang, M. (2009). A novel approach for multi-agent-based intelligent manufacturing
system. Information Sciences, 179(18), 3079-3090.
[55] Balta, Efe C., Lin,Y., Barton,K., Tilbury, D.M., Mao, Z.M., (2018). Production as a Service: A
Digital Manufacturing Framework for Optimizing Utilization. IEEE Transactions on Automation Science
and Engineering, 15(4), 1483-1493
[56] Shpilevoy, V., Shishov, A., Skobelev, P., Kolbova, E., Kazanskaia, D., Shepilov, Y. and Tsarev, A.,
(2013). Multi-agent system “Smart Factory” for real-time workshop management in aircraft jet engines
production. IFAC Proceedings Volumes, 46(7), pp.204-209.
[57] Luo, H., Fang, J. and Huang, G.Q., (2015). Real-time scheduling for hybrid flowshop in ubiquitous
manufacturing environment. Computers & Industrial Engineering, 84, pp.12-23.
[58] Dal, B., Tugwell, P., & Greatbanks, R. (2000). Overall equipment effectiveness as a measure of
operational improvement–a practical analysis. International Journal of Operations & Production
Management.
[59] Nakajima, S. (1988). Introduction to Total Productive Maintenance (TPM), Productivity Press,
Cambridge. MA.
[60] Bertrand, J.W.M and Muntslag D.R., (1993), “Production Control in engineer-to-order firms”,
International Journal of Production Economics,30-31, 3-22
[61] Hicks, C. and Braiden, P. M., (2000), “Computer-aided production management issues in the
Engineer-to-order Production of Complex Capital goods explored using a simulation approach.”,
International Journal of Production Research, 38, 4783–4810

Digital Thread Enabled Manufacturing Automation for Mass Personalization 32