2018 2nd Int. Conf.
on Innovations in Science, Engineering and Technology (ICISET)
27-28 October 2018, Chittagong, Bangladesh
An Internet of Things (IoT) based Smart Traffic
Management System: A Context of Bangladesh
Abdul Kadar Muhammad Masum Md. Kalim Amzad Chy
Department of Computer Science and Department of Computer Science and
Engineering
International Islamic University Chittagong, International Islamic University
Bangladesh Chittagong, Bangladesh
akmmasum@yahoo.com kalim.amzad.chy@gmail.com
Mohammad Nazim Uddin
Department of Computer Science and Engineering
International Islamic University Chittagong, Bangladesh
nazimuddinasifiiuc@gmail.com
Abstract— With the fast growth of population, traffic
congestion monitoring and control has become a great challenge.
Increasing vehicles creates lots of problem like time wastage, fuel
wastage, air and sound pollution, even death by getting stuck
emergency vehicles. This paper proposes a real-time traffic
management system (TMS) using the Internet of Things (IoT)
and data analytics. Ultrasonic sensors are used to measure the
traffic density. After analysing the sensor data, system controller
sets traffic signal time by traffic management algorithm and also
sends data to a cloud server through a Wi-Fi module. The
proposed system can predict probable traffic congestion in the
intersection point. If an emergency vehicle is detected, it gives
priority, i.e. high signal duration to pass the intersection. In case
of the signal violation, the system can identify the vehicle and
charge a fine that is paid through Traffic Wallet mobile app. This
proposed system is cost-effective, very simple to install and easy
to maintain.
Keywords - Data Analytics; Internet of Things (IoT); smart
traffic system; Sensors; Wi-Fi module
I. INTRODUCTION
A recent statistic shows that 30% of the air pollution is
caused by the fuel that is used in the vehicle. It has been
reported that congestion in the Dhaka – capital of Bangladesh
eats up around TK 200 billion in a year [1]. Researchers also
stated that this traffic jam lost 32 lakh business hours per day
[2]. Centre for Economics and Business Research reported that
it would rise to nearly TK 300 billion by 2030 [2]. The
congestion for hours causes different severe types of pollution.
Consequently, people are suffering from any kind of physical
and mental discomfort [3]. To get rid of this unwanted
congestion, Bangladesh needs a system that will overcome the
problems of the existing system.
So, a smart traffic system can be a solution to these
problems. Controlling the vehicle with the help of the
traditional traffic light with the combination of the sensors and
artificial intelligence is known as the smart traffic system [4].
The proposed smart TMS measures traffic density by analysing
978-1-5386-8524-2/18/$31.00 ©2018 IEEE
Iaamanur Rahman
Department of Computer
Engineering Science and Engineering
International Islamic University
Chittagong, Bangladesh
rafidrahman38@gmail.com
Khairul Islam Azam
Department of Computer Science and Engineering
International Islamic University Chittagong, Bangladesh
khairulislam.azam@gmail.com
sensor data. It sets the timing for traffic signal light by traffic
management algorithm. Traffic density is visualised
graphically at Thingspeak by sending data through ESP8266
Wi-Fi module. In case of emergency vehicles like ambulance,
fire brigade, the system stops it’s as usual task and changes the
signal as green until that emergency vehicle passes to that
intersection. In violation of traffic signal, it can detect the
vehicle through RFID module and the system can charge fine,
and a notification will send to that person through Traffic
Wallet Mobile App.
The prime objective of this paper is to develop a smart
traffic management system (TMS) using IoT. To the best of
our knowledge, smart traffic system is not implemented in
Bangladesh yet. This whole approach will cost less than other
approaches. Moreover, operating the proposed system is more
comfortable than other existing systems.
The rest of this paper is structured as follows. Section II
deliberates the literature reviews. The architecture of the
proposed system is described in Section III while Section IV
describes the methodology of the system. An experimental
result is presented on in Section V. Finally; the last section
includes the limitations, future directions and implications of
the study.
II. LITERATURE REVIEW
The urban population is incredibly increasing in this
modern era, and that affects everyday life very badly,
especially in transportation. Cities like Delhi, Dhaka and many
more developing countries are still using the traditional way of
managing the vehicles for the intensely increasing population.
According to United Nations report, in 2018, about 55%
population of the world resides in an urban area that is
anticipated to be 68%, and increasing rate of Asia and Africa
closed to be 90% by 2050 [5]. For this growing urban
population, an effective smart traffic congestion avoiding
system is a crying need for managing the significant number
of vehicles.
418
 The contemporary research stated that Japan, one of the
leading positions in the Intelligent Transportation System
[ITS], uses two techniques to collect real-time data increasing
congestion control. One is embedding sensors in roadsides and
another via smartphones[6]. Similarly, South Korea collects
real-time traffic data by camera sensor equipped every 2-3 km
and Vehicle Detection System (VDS). The data is stored at
their National Transport Information Center (NTIC) [6].
Singapore also uses probe vehicles to collect real-time
information. They deliver the services to the commuters
through vehicle devices or ACMS (Advanced Congestion
Management System) [6]. Moreover, Research and
Innovative Technology Administration (RITA) wing of the
United States deals with the traffic congestion avoiding
system. They mainly concentrate on IntelliDrive, Telephonic
Data Dissemination, Congestion Initiative, Emergency
Transportation Operations and Clarus Initiative [7]. Likewise,
the United Kingdom uses MIRA that utilizes the self-healing
communications networks. It can confirm safeties of a very
complex system [8].
As a developing country, India, in some province, uses
MATLAB, KEIL (Microcontroller coding) based system and
surveillance camera to control traffic congestion [4]. The
system was cumbersome to install and costly so that some
province uses shortest route detections, IR sensors to measure
traffic density[9]. But, IR sensor is impacted by temperature
and humidity. Consequently, the result which was produced
by the IR sensor was not accurate. In Pakistan, traffic density
is measured using camera and sensors. Based upon the
sensors’ data, Pakistan controls the traffic signalling. They
also used a smoke sensor to detect the emergency situation,
i.e. fire accident [10]. Camera sensor may be affected by rain
and fog. Moreover, it’s not cost effective. In Nepal, wireless
traffic data and CCTV live video are used in 35 traffic
junctions. The system can modify signal policies and reroute
drivers to prevent congestion [11].
The research on smart traffic system is limited in the
context of Bangladesh. Recently, closed-circuit cameras and
car detection devices are used to handle the traffic congestion,
and the system can calculate the number of vehicles on the
road [12]. But, in rainy and foggy environment cameras can't
detect vehicle accurately. To the best of our knowledge,
Bangladesh doesn’t have any holistic model to monitor and
control the whole traffic system. This paper is the first attempt
to propose a smart traffic management model in the context of
Bangladesh.
III. SYSTEM ARCHITECTURE
The designed system, consists of IoT and wireless
technology, are equipped at roadsides to monitor traffic
density (Figure 1). To measure traffic density, a series of HC-
SR04 ultrasonic sensors were used which sends data to the
server through the ESP8266 Wi-Fi module. The system
dynamically sets the signal time based upon the density level
where time is measured by the traffic management algorithm
[10]. Also, the system deals with emergency transports such as
fire brigade, ambulance etc. Sensors data is stored in cloud
storage for further analysis. From those data, the system can
predict traffic density by using a machine learning
419
classification algorithm namely Support Vector Machine
(SVM).
If anyone violates the traffic signal, he will be identified by
scanning RFID tag and rational fine will be imposed on him.
Also, the system can locate either a vehicle licence and driver
licence is expired or not and can take steps like rule violation.
From a national database, the system will get all necessary
information regarding that people. Fine will be paid through
the mobile banking system. To implement this, we have
developed a mobile app named “Traffic Wallet” where the
owner of the vehicle will be notified about his mobile banking
transaction and many more. By analysing the sensors’ data,
micro-controller controls signals, processes and sends data to
the cloud server through the ESP8266 Wi-Fi module.
Fig. 1. System Architecture
IV. METHODOLOGY
Researchers have used several ways to detect traffic
density consisting of different kind of sensors like surveillance
cameras, ultrasonic sensors, RFIDs, the light beam that have
merits as well as demerits. Ultrasonic sensors and RFIDs are
suitable sources for our proposed system. The ultrasonic
sensor is most used sensor to identify the traffic density level
in TMS [9]. It can calculate the distance up to 400 cm [13].
Here, Density is calculated by the following formulae:
3
∑ (Pi) = Pi + Pi+1 + Pi+2 (1)
i=1
Where P is the pair of ultrasonic sensors. If the value of the
three pairs of sensors is high, that will be considered as high
density, if two pairs are high and another pair is low, that will
be medium and if one pair is high and another two pairs are
low that will be low density. The sensors continuously send
data to the microcontroller to detect traffic density. So, a
density-based and dynamic traffic signal timing is essential
instead of manual traffic signalling process. For that reason,
traffic management algorithm was used. In rush interval, for
high traffic density, a time scheduling formula is followed and
if not, then another time scheduling formula is developed that
is shown in the flow chart (Figure 3). Regression Tree
Algorithm is used to calculate Rush interval time, a span of 30
minutes. Here, α is the usual time that is pre-fixed, β is
additional time in case of traffic jam, γ is extra time in case of
rush interval closed to approach, X=0 and Ѳ=90. Also, in case
of an emergency vehicle like an ambulance, fire brigade it will
stop it’s as usual task and changes the signal as green until that 5) LEDs signal: LED's signals are used as a traffic
emergency vehicle passes to that intersection. Violating rule is signal
one of the most critical, and this system can detect the user light that is controlled by time scheduling algorithm shown in
through scanning RFID tag who break the rules. Also, the figure 2.
system can check to expire of a license. A proper amount of
fine will be imposed, and a notification will be sent to the B. Softwares:
owner through "Traffic Wallet". That fine will be paid through
the mobile banking system. All the data about traffic density, By analysing real-time traffic data, IoT analytics creates
date and time are sent to the cloud server for further IoT channels to plot real-time graphs. The uploaded code of the
analytics. To predict traffic density at an intersection point system is written in the Arduino language. ESP8266 Wi-Fi
machine learning algorithm named SVM is used. In figure 2, a module transfers sensors’ data to the server through micro-
block diagram of the system appears: controller. Server database stores sensors’ data for further
analysis. Uploaded system code is burnt through Arduino IDE
that is user reliable and easily understandable. Android studio
was used to develop Traffic Wallet mobile app.
C. Flowchart:
The flowchart of the working module is shown in figure 3.
The flowchart was followed to implement the proposed system.
D. System Design Implementation:
To describe our proposed system, a prototype was
Fig. 2. Block Diagram developed. All elements are tested before prototype
implementation. In this part, Arduino Mega 2560 is
In the Block diagram of the proposed framework, at the accommodated with HC-SR04 ultrasonic sensors, RFID
input side, Arduino Mega 2560 controller is accommodated module, LEDs signal and ESP8266 Wi-Fi module shown in
with ultrasonic sensors, RFID module and at output side figure 4.
ESP8266 Wi-Fi module to transfer data at cloud server, LEDs
as signal light.
The proposed system is classified into two main parts:
A. Hardwares: Following sections are used in our proposed
system:
1) Series of HC-SR04 Ultrasonic Sensors: Ultrasonic
Wave based, HC-SR04 ultrasonic sensor’s head emits an
ultrasonic wave and receives the reflected wave from the
target. Measuring the taken time between the emission and
reception it measures the distance to the target. In this system,
HCSR04 ultrasonic sensors are used to measure traffic density
at roads. At 40 KHz frequency, it can detect vehicles at range
from 2cm to 400 cm [14].
2) Arduino Mega 2560 controller: The Arduino Mega
2560 is ATmega2560 based microcontroller that has 16
analogue inputs, 54 digital input/output pins, 4 UARTs, a 16
MHz crystal oscillator, an ICSP header, a USB connection, a
reset button and a power jack [15] which is useful for this
system.
3) RFID module: MF RC-522 based RFID module uses
Electromagnetic fields to automatically identify and track tags
attached to objects. It is a low-cost, small size and non-contact
card reader chip that is used in this system to detect a signal
violation person by scanning his RFID tag.
4) ESP8266 Wi-Fi module: TCP/IP protocol based
ESP8266 Wi-Fi Module has a powerful enough processing
capability that allows it to be integrated with the sensors with
minimal loading during runtime. In this system, it is used to
transfer real-time data at the server side [16].
Fig. 3: Flow Diagram

420

Fig. 4: Proposed system prototype
V. EXPERIMENTS AND RESULTS
Ultrasonic sensors based real-time traffic monitoring
system can measure traffic density at roads. When the
frequency is high, that route is in green signal, and a low
dense route is in red signal. The system can also detect rule
violation user and check the licence is expired or not. For such
kind of guilty person, the system can take steps automatically
by imposing fine. The system sent and store the data at a cloud
server for further analytics. To show the authorities statistical
data graphically, a web interface is also developed. An
authorized person can access those data from anywhere
through internet. The real-time results of traffic density at
roads is shown in Figure 5 and Figure 6. Traffic Wallet
Mobile app interface is shown in figure 7 and figure 8.
Fig. 5: Firs Lane Traffic Level
Fig. 6: Second Lane Traffic Level
421
Fig. 7: Interface of Law Enforcement Authority
Fig. 8: Notification after Fine
VI. CONCLUSION
Traditional system has many limitations to manage present
increased traffic effectively. This paper proposed a smart TMS
to control traffic situation more effectively and efficiently. By
analysing sensor data, it sets traffic signal time dynamically
and sends the data to a cloud server through a Wi-Fi module
that is stored for further data analytics. It also deals with
emergency vehicle and in case of traffic signal violation can
detect and charge a fine that is paid through Traffic Wallet
mobile app. This is a unique concept in perspective of
developing countries especially Bangladesh. The whole
system is very cost effective than existing system in
developing countries. But, the security should be ensured for
sensors as these are equipped at roadsides. For further
research, some features can be initiated to handle Bigdata by
Hadoop HDFS, MapReduce as day by day server data will be
massive. The government can equip this proposed system
immediately to change the current terrible scenario of the
traffic congestion of Bangladesh.
REFERENCES
[1] S. Correspondent, "Traffic Congestion: Tk 20,000cr loss to
economy a year," in The daily star, ed. Dhaka -Bangladesh, 2016,
September 25.
[2] a. Samad, "'Traffic jam' silent killer," in The Asian Age, ed. Dhaka-
Bangladesh, 2016, January 28.
[3] A. Khan, "Problems caused by traffic jam in Dhaka," in Daily
Observer, ed. Dhaka-Bangladesh, 2016, Maech 30.
[4] P. Jadhav, P. Kelkar, K. Patil, and S. Thorat, "Smart traffic control
system using image processing," Int. Res. J. Eng. Technol, vol. 3,
2016.
[5] U. News, "68% of the world population projected to live in urban
areas by 2050, says UN," ed. New York - USA, 2018, May 16.
[6] G. Singh, D. Bansal, and S. Sofat, "Intelligent Transportation
System for Developing Countries-A Survey," International
Journal of Computer Applications, vol. 85, 2014.
[7] ITS. (2018). The Intelligent Transportations Systems Joint
Program Office (ITS JPO). [Online].Available:
http://www.its.dot.gov/Efm/. [Accessed: Sep. 23, 2018].
[8] MIRA. (2018) . [Online]. Available: https://www.horiba-
mira.com/. [Accessed: Sep. 23, 2018].
[9] K. Swathi, V. Sivanagaraju, A. Manikanta, and S. D. Kumar,
"Traffic Density Control and Accident Indicator Using WSN,"
Traffic, vol. 2, 2016.
[10] S. Javaid, A. Sufian, S. Pervaiz, and M. Tanveer, "Smart traffic
management system using Internet of Things," in Advanced
Communication Technology (ICACT), 2018 20th International
Conference on, 2018, pp. 393-398.
[11] T. H. Times. (April 29, 2016). Intelligent traffic management
system mooted to ease jams. [Online]. Available:
422
https://thehimalayantimes.com/kathmandu/intelligent-traffic-
management-system-mooted-ease-jams/. [Accessed: Sep. 23,
2018].
[12] S. Bosu, "Dhaka to get ‘intelligent’ traffic system, now " in The
Independent, ed. Dhaka -Bangladesh, 2018, February 19.
[13] S. N. Mahalank, K. B. Malagund, and R. Banakar, "Device to
device interaction analysis in IoT based Smart Traffic Management
System: An experimental approach," in Colossal Data Analysis
and Networking (CDAN), Symposium on, 2016, pp. 1-6.
[14] ElectronicWings. (2018) . Ultrasonic Module HC-SR04. [Online].
Available: http://www.electronicwings.com/sensors-
modules/ultrasonic-module-hc-sr04. [Accessed: Sep. 23, 2018].
[15] (2018). ARDUINO MEGA 2560 REV3. [Online]. Available:
https://store.arduino.cc/usa/arduino-mega-2560-rev3. [Accessed:
Sep. 23, 2018].
[16] S. Electronics. (2018). WiFi Module - ESP8266. [Online].
Available: https://www.sparkfun.com/products/13678. [Accessed:
Sep. 23, 2018].