IoT-Based Smart Office System

Architecture Using Smartphones

and Smart Wears with MQTT
and Razberry

R. Padma Priya, J. Marietta, D. Rekha, B. Chandra Mohan

and Akshay Amolik

Abstract Internet of things due to its profound capability has been responsible for
the innovative transformations in lifestyles of human. We propose a system for the
smart office environment using the Internet of things. Our developed system con-
figuration provides a comfort of handling smart office environment using Google
Voice Over Protocol and Android-based application. Use of lightweight MQTT pro-
tocols, high-frequency RF trans-receivers, Raspberry 2 Z-wave-based long distance
controllers makes this system feasibly suitable for wider office areas. For low power
consumption, high performance and immediate time-to-time response QoS param-
eters, the system is equipped with PIR, gas, flame, sound, light, temperature and
humidity sensors. We ensure reliable communication with a separate channel for
every sensor signals, secured with AES encryption and plug-and-play provisions.
Further, mail-based anomaly or emergency alert notification is also incorporated.
Finally, we have performed the experiment and also tested the developed system in
Android phones and smart wear watch, namely Moto360 Watch.

Keywords Internet of things · Smart office · MQTT · Raspberry · Smart city

R. Padma Priya · J. Marietta (B) · D. Rekha · B. Chandra Mohan · A. Amolik

Vellore Institute of Technology, Vellore, India
e-mail: [email protected]
R. Padma Priya
e-mail: [email protected]
D. Rekha
e-mail: [email protected]
B. Chandra Mohan
e-mail: [email protected]
A. Amolik
e-mail: [email protected]

© Springer Nature Singapore Pte Ltd. 2019 623

A. J. Kulkarni et al. (eds.), Proceedings of the 2nd International Conference on Data
Engineering and Communication Technology, Advances in Intelligent Systems
and Computing 828, https://doi.org/10.1007/978-981-13-1610-4_63
624 R. Padma Priya et al.

1 Introduction

Internet of things has numerous applications. Many problems related to the environ-
ment and smart city could be addressed by various technologies in IoT. The concept
of smart office, as an application belonging to Internet of things (IoT) domain, pro-
vides features related to the intelligent behaviour of entire work environment such
as office rooms, office entrances, office parking.
A network of dedicated sensors, actuators and various specialized devices is used
to control temperature, light intensity, humidity, vibration, noise level and other in-
room environment parameters related to office employees. Nowadays, the need to log,
an underlying business process that helps operate the daily tasks, which are regularly
done in the offices, is becoming an important artefact. Smart office-based solutions
have to ensure the need to communicate with the various project management systems
and information systems to control the regular office work. Thus, Internet of things
(IoT) not only is about improving business processes but has also the potential to
profoundly impact the lives of many citizens.
This paper is divided into the following sections where Sect. 2 presents a survey
of work done in this field. Section 3 gives the Proposed Work, while Sect. 4 shows
Experimental Results and Analysis and Sect. 5 gives Conclusion and Future Work.

2 Literature Survey

Various literatures have been studied related to the smart office environment in the
smart city. The idea of smart office environment by using the Internet of things was
proposed in [1]. They have used a methodology which is based on communication
between various sensors and server running remotely.
Domaszewicz et al. [2] have done the whole automation system with low size and
automated hints and performed action on nearby object. Olivieri et al. [3] proposed
the publish–subscribe approach to make integrated IoT-based system. They have
proposed their research work with smart office as a use case. Foster et al. [4] have
given intelligent methods which control the air conditioners which are provided by
sensor data. Sanoob et al. [5] proposed the integrated system which uses the camera to
capture the motion and sends SMS to the user’s smartphone about any security theft.
Karol et al. [6] have given adaptable network architecture of smart office solution
based on service-oriented technology which focuses on the continuous enhancement
of user experience and monitoring of evaluation of adjustment of the integrated
system to user requirements which are continuously changing according to social
and business parameters.
Nati et al. [7] proposed a user-centric approach for IoT in smart campus. Prasad
et al. [8] proposed an innovative Hadoop-based cloud framework for observing and
tracking tangible assets in smart campus. In their research paper, they experimented
on the displacement or missing assets in laboratories and library by integrating cloud
IoT-Based Smart Office System Architecture Using Smartphones … 625

computing and RFID transceivers. Zanella et al. [9] presented and discussed var-
ious guidelines for best practice and technical solutions in IoT for smart cities.
Their research paper provides comprehensive knowledge and survey about mod-
ern architectures, protocols and technologies used in urban IoT. Muralidhara et al.
[10] proposed and developed secured and smart healthcare monitoring system.
Lucke et al. [11] developed cyber-physical system which can be effective and can
be adapted by governments for their administration. This will help the government to
execute public tasks effectively and efficiently. Ghazal et al. [12] proposed traffic light
control system with the help of PIC microcontroller which continuously analyses
traffic density using infrared sensors and provides dynamic monitoring. Hayajneh
et al. [13] proposed an adaptable and resilient disaster recovery network for smart
cities.

3 Proposed Work

3.1 System Architecture

The detailed design of the fully automotive system consists of IoT-based configu-
ration and mobile application-based user interface. Scalability and flexibility have
been given special importance while designing the proposed method. This allows
heterogeneous devices to be connected to the system. Cloud storage provides the
options to store the huge data. Wireless sensor nodes are connected through the
suitable protocols. This provides wide area coverage and accuracy in sensor data.
The flow of the total system is shown in Fig. 1 where all operations are clearly
specified. The following flow chart provides how exact sensor data is processed and
how communication between phone and sensor happens. In the following flow chart,
each operation represents a function which is performed in this system. The flow of
operation is simple and easy to understand.
All sensors are controlled by the Arduino Controller, which is having RFM69HW
wireless transceiver mounted over it. We can call this whole set-up as sensor node.
All the sensors send data to another Arduino which has Arduino W5100 ethernet
shield mounted on it. We can call this ethernet gateway.

3.2 Face Detection Algorithm

Algorithm for face detection is as follows:

626 R. Padma Priya et al.

Fig. 1 Flow chart for the proposed methodology

Algorithm
1. Start.
2. Pass the input image or take a photo to detect face.
3. Click on Detect button to start detecting the human faces in the photograph.
IoT-Based Smart Office System Architecture Using Smartphones … 627

4. Extract the face parameters from the detected face (parameters such as the
distance between left eye and right eye, the distance and position of lip).
5. Compare the extracted parameters with the parameters within the database
(mobile database, i.e. SQLite).
6. If the parameters are matched, then the result (i.e. the person’s information
is passed to the user).
7. Otherwise, the person is unknown and notified to the user.
8. Stop.

4 Experimental Results and Analysis

We were able to achieve the reliability and accuracy of the system. Due to use of
MQTT protocol which is lightweight protocol, we have reduced the time of commu-
nication between sensor and mobile device.
Below scenarios are experimentally tested in our proposed model which includes
fire emergency condition monitoring, theft detection, surveillance and security pro-
vision with the camera (Fig. 2). The accuracy of results is very high, and the fully
adaptable system can be easily integrated or scaled with a larger system.
Making an adaptable, scalable smart system with IoT is a challenging Task. The
cost is also an important factor while making these systems. But a combination of
cheap sensors, adaptable server with uniprocessor Raspberry Pi promotes researchers

0 5 10 15 20 25

Temperature Sensor Gas Sensor Sound Sensor Light

Sensor Flame Sensor PIR sensor

X-Axis:- Sensor Values (Analogy and Digital Values).

Y-Axis:- Time in Seconds.
Fig. 2 Graph of sensor values over time
628 R. Padma Priya et al.

Fig. 3 Hardware setup and notification in the watch

towards testing smart applications. Due to MQTT protocol and openHAB server, we
have achieved good results. We have assigned each sensor data a unique key so that
we can get accurate results with less amount of time. This avoids the mismatching
of data and false results as observed in (Fig. 2). For each sensor, the threshold value
is checked and thus it helps us to process only valid data and in reducing the time
of transmission. Using encryption key, all sensor data are secured, which solves the
major problem of security. Many receivers are connected to the system which can
be scaled. So, the user can connect his phone, laptop, smartwatch simultaneously to
the system.
The above graph shows the sensor values over time. The sensor continuously sends
the valid data to the server only after some time interval. This interval is different for
various sensors so by doing this space on the server is utilized properly. The graph in
the following shows the time when the data is sent to server. This also ensures proper
utilization of resources. The blue rectangle box shows an emergency condition where
all sensor values are above the threshold, and at this time, the user is notified with
notification or through mail. All these things make the whole system scalable and
adaptable with less resource utilization. The server logs can be saved for getting
history. Past days’ logs are saved, and the older logs are deleted automatically to
do proper utilization of space. PIR sensor and flame sensor send digital values. So
when digital output is 1, then only sensor sends data to the server. This is shown in
the emergency condition when the entire sensor sends high values. So this will also
reduce the time and server space. Only valid data is processed so this reduces extra
overhead on the server (Fig. 3 and Table 1).
IoT-Based Smart Office System Architecture Using Smartphones … 629

Table 1 Hardware Specifications

Device Name Hardware Specification/Features
Raspberry pi 1GB Ram, 10/100 Ethernet, 2.4GHz 802.11n
wireless LAN, Bluetooth 4.1 Classic Bluetooth
Low Energy, CPU: – 4× ARM Cortex-A53,
1.2GHz
Arduino UNO R3 Digital I/O Pins: – 14
Flash Memory: – 32 KB
SRAM: – 2KB
Operating Voltage: – 5V
W5100 Ethernet Shield For Arduino Digital I/O Pins: – 14
6 Analog Inputs, 16 MHz crystal oscillator,
Rj45 connection, power jack, ICSP header
RFM69HW Transceiver +20 dBm –100 mW power Capacity.
RF is constant over voltage.
FSK, MSK, OOK, GMSK, GFSK modulations.
Automatic RF Sense with Ultra-Fast AFC
PIR Sensor Sensitivity Range: - up to 20 Feet (6 Meters),
110° X 70° detection range
Power Range: – 5V–12V
MQ2 Gas Sensor Detect Combustible Gas, Smoke
Power Supply:- 5v
Wide Detecting scope
High Sensitivity and Fast Response
Sound Sensor Easy to use
Microphone Sensitivity 52–48 db
Microphone Frequency 16–20 kHz

5 Comparison Results of AMQP and MQTT

MQTT is the IoT connectivity protocol. It is described as the machine-to-machine

protocol. Since it supports smallest measuring and the monitoring devices. Hence
termed as the lightweight protocol. The data could be transmitted in the intermit-
tent network too. It is a publish/subscribe model. The real-world objects could be
connected to the servers and consumers. The scalability issue faced by the sensors,
actuators, mobiles and the tablets could be a problem, and to overcome this issue,
MQTT protocol is designed. The battery power and the bandwidth are very essential
for the connected devices in IoT. The developed principles and the designed MQTT
protocol are helpful for the connected IoT devices. This is a protocol for collecting
device data and communicating it to servers. It is TCP/IP-based messaging protocol.
It is light on network bandwidth and has a low power draw.
AMQP is designed as a replacement for existing proprietary messaging middle-
ware. The advantage of using AMQP is reliability and interoperability. The features
included by AMQP are reliable queuing, topic-based publish-and-subscribe messag-
ing, flexible routing, transactions and security. The message formats such as classic
630 R. Padma Priya et al.

Fig. 4 Time consumed when sending message using MQTT

message queues, round-robin, store-and-forward are supported by AMQP. In order

to support idempotent messages and message grouping, meta-data support used by
AMQP is a queuing system designed to connect servers to each other.
CoAP runs on UDP. This can be used on extremely resource-constrained envi-
ronments. It uses semantic, parallel to HTTP which is a best mechanism for local
network communication. Both MQTT and CoAP are used for the device-to-service
connection. CoAP is Application protocol, which requires a broker (server).
The MQTT statistic graphs provide statistics about message published from and
received by the MQTT broker during a simulation run. The hardware platform for
testing has two Raspberry Pi boards with the same hardware configuration to investi-
gate rate of packet transmission and performance based on time. The measurement of
performance is run by sending messages between publisher and subscriber through
broker.
Graph Figs. 4 and 5 represents as simulation system model by sending messages
between Raspberry Pi 1 and Raspberry Pi 2 either simulation with AMQP or MQTT.
Performance is measured by sending message between publisher and subscriber
through broker. Below graph represents the simulation model by sending messages
through the kit for AMQP (or) MQTT. AMQP could send message up to 3000 with
acknowledgement MQTT could send message up to 4000 with acknowledgement.

6 Conclusion and Future Work

The observations and experimental tests, prove that the system is designed to fulfil the
smart office environment, removing all the barriers of theft and to protect the office
from theft. This system can be made as hub in future work where we can use various
IoT-Based Smart Office System Architecture Using Smartphones … 631

Fig. 5 Time consumed when sending message using AMQP

machine learning techniques. This will help the system to recognize speech or voice
of the user and will help ON the user command. All the sensors can communicate
with this hub where fuzzy logic is used to make prediction. Interfaces such as Amazon
Echo (Alexa Voice Service), Google Speech API can also be integrated with Google
Assistant. The voice commands to the system could be given by the device such as
Amazon Echo dot. We can also add Bluetooth devices and GPS tracking to improve
performance. Use of routing algorithm can help to enhance the system.

References

1. Durand M (1999) Smart office: an intelligent and interactive environment. In: Proceedings of 1st
international workshop on managing interaction in smart environments, Dublin, pp 104–113,
R.I, 997
2. Domaszewicz J, Lalis S, Pruszkowski A, Koutsoubelias M (2016) Soft actuation: smart home
and office with human-in-the-loop. IEEE Pervasive Comput 15(1)
3. Olivieri AC, Rizzo G, Morard F (2015) A publish-subscribe approach to IoT integration:
the smart office use case. In: INSPEC Accession Number-15107080, Advanced Information
Networking and Applications Workshops (WAINA) IEEE Conference on Networking
4. Foster TW, Bhatt DV, Hancke GP, Silva B (2016) A web-based office climate control system
using wireless sensors. IEEE Sens J 16(15):6104–6113
5. Sanoob AH, Roselin J, Latha P (2016) Smartphone enabled intelligent surveillance system.
IEEE Sens J 16(5) 1361–1367
6. Furdik K, Lukac G, Sabol T, Kostelnik P (2013) The network architecture designed for an
adaptable IoT-based smart office solution. Int J Comput Netw Commun Secur 1(6):216–224
7. Nati M, Gluhak A, Domaszewicz J, Lalis S, Moessner K (2014) Lessons from smartcampus:
external experimenting with user-centric internet-of-things testbed. Wireless Pers Commun
93(3):709–723
632 R. Padma Priya et al.

8. Prasad DVV, Priya RP, Jaganathan S (2016) An innovative cloud framework for tracking and
monitoring assets in smarter campus using RFID. Asian J Technol 15(11):1713–1722, ISSN:
1682-3915
9. Zanella A, Bui N, Castellani A, Vangelista L, Zorzi M (2014) Internet of things for smart cities.
IEEE Internet Things J 1(1):22–32
10. Muralidhara KN, Bhoomika BK (2015) Secured smart healthcare monitoring system based on
Iot. Int J Recent Innov Trends Comput Commun 3(7):4958–4961. ISSN: 2321-8169
11. Lucke JV (2016) Smart government—the potential of intelligent networking in government and
public administration. In: 2016 conference for e-democracy and open government (CeDEM)
12. Ghazal B, Elkhatib K, Chahine K, Kherfan M (2016) Smart traffic light control system. In:
2016 third international conference on electrical, electronics, computer engineering and their
applications (EECEA)
13. Hayajneh AM, Zaidi SAR, Mclernon DC, Ghogho M (2016) Drone empowered small cellular
disaster recovery networks for resilient smart cities. In: 2016 IEEE international conference
on sensing, communication and networking (SECON Workshops)