Internet of Things based Adaptive Traffic

Management System as a part of Intelligent

Transportation System (ITS)
Ankit Dubey, Mayuri Lakhani Shivansh Dave, Jignesh J. Patoliya
Dept. of Electronics & Communication Engineering, Dept of Electronics & Communication Engineering,
CSPIT, CHARUSAT, Changa, CSPIT, CHARUSAT, Changa,
Anand, Gujarat, INDIA Anand, Gujarat, INDIA
[email protected] [email protected]

Abstract— Traffic congestion and higher average waiting time population in some regions is increasing. For these reasons we
has been a problem for a very long time. The purpose of this need to implement Intelligent Traffic Management System that
project is to design and implement a traffic system that is is adaptive to the traffic conditions on a particular lane, in a
adaptive to nature of the traffic in respective lanes. Most of particular region. Also, considering a bigger picture of this, all
traffic signals are having counters according to which the traffic lanes and intersections must be connected to each other and to
lights of different lanes get changed one by one. To solve this Traffic Controllers and Developers so that analysis and study
problem of fixed wait time, counter for any traffic, we proposed about traffic patterns become easier for statisticians and
this adaptive traffic system which is connected to internet so that researchers. This can help to determine more hidden and real
different lanes can be monitored constantly. The data obtained time traffic problems and a most probable efficient solution to
from different lanes are examined and controlled by Central
Traffic Control Office from one place. Data obtained thus gives
it. So, we can say that to have a smart city, the very first step
value of traffic congestion in particular lane, according to which
must be an intelligent and adaptive traffic management system.
traffic lights are programmed to work. If the first lane is having B. Internet of Things and Big Data:
less traffic than other lane, then the signal lights will be decided
on the basis of less wait time and less pollution. This system also In this past decade the big data and IoT has been a trending
gives idea to drivers to choose the path with less congestion. This hot technology being used all over the world. Data is
system is also useful in emergency and VIP clearance and in omnipresent. These data can be used for analytics. Data can be
traffic survey. This increases the efficiency of traffic clearance. gathered from many sources such as trains, wind turbines,
This also reduces pollution and traffic congestion, thus being an farms, roads and more. Since these systems are real time
Adaptive Traffic Control System using Internet of Things. systems, the data associated with it would be very large. Well
this is a very general sense about Big Data. Researchers and
Keywords—Internet Of Things; Traffic Management System; statisticians use these data for research and development of
Intelligent Transportation System; present system. But how do we get these data and how do we
use these data to provide better solutions? Well the answer to
I. INTRODUCTION this is Internet of Things. So what is Internet of Things?
A. Intelligent Traffic Management- Backbone of Smart city: Internet of Things, more commonly referred to as IoT is an
environment and a platform that connects people, devices and
The technology around us is changing at a tremendous rate. computers by letting them share data with each other by having
We all hear about smart cities coming up. What is a smart city? machine to machine or machine to human interaction. Several
Well, a very simple answer to that is that it is a collective term, companies and organisations are working in this domain as it
i.e. it is an integration or merger of information technology, has been predicted as the future of electronics and information
huge infrastructure, digital framework, security along with technology. Some of these are IBM, Intel, Hitachi, TCS and
economic and social feasibility. Traffic Management System of more. Since IoT makes analysis, development, research and
a city is its backbone. To have smart city you need to have transfer data easier, providing specific solutions to specific
Intelligent Traffic Management System. A very general sense problems becomes possible.
about traffic management about some time ago would be that
traffic management is a method that is implemented in a city or C. Intelligent Transportation System:
a region for clearance of traffic in a particular area. Does it ITS refers to variety of tools such as traffic engineering
hold up true in modern day world too? Well, not precisely. In concepts, software, hardware and communication technologies
present world, clearance of traffic is no more the only priority that can be applied in an integrated manner to the
of traffic management. Many problems of modern day such as transportation system to improve its efficiency and safety [1].
pollution, congestion, average waiting time, etc. have become Different countries in different parts of the world are working
major factors to be eliminated so as to have cleaner air and on providing infrastructure and environment for ITS. Several
hassle free transport system. Day by day, the number of vehicle countries like United States and countries of Europe have
commuters is increasing. This is due to the fact that fuel prices started developing and implementing ITS at ground level [2].
are lowering and employment rates are increasing. Also the
 Our model is an integration of Intelligent traffic D. ITS- India and rest world:
management system and Internet of Things as a part of India is trying to match shoulders with developed countries
Intelligent Transportation System. of the world. In an attempt of so, India is one of the country in
II. MOTIVATION the world that is aimed at developing and implementing ITS at
ground level for future smart cities. According to the
A. Present traffic scenario in India: Department of Electronics and Information Technology,
The present traffic scenario in India is very concerning. (DeiTY) India, ITS can bring sustainable transportation
With a very large population and emerging workforce in the solution [7]. In this very approach an embedded Centre of
country, India exhibits huge traffic congestion problems. The Excellence for ITS will be created at CDAC,
main reason to this problem is that the country is emerging and Thiruvananthapuram, Kerala, India. This will also be the
developing at a very rapid speeds. The GDP annual growth rate national resource centre of the country [7].
of India has evidently increased from 6.9% in 2013 to 7.3% in III. ADAPTIVE TRAFFIC MANAGEMENT SYSTEM
2014 according to the World Bank Data [3]. This indirectly
emphasize that economy of India is increasing, and hence both A. Ideas:
Private vehicles and Freight Vehicles have increased The basic idea behind having an adaptive traffic controlling
sufficiently. But a very serious threat to this is the present system is to design the traffic management system in a way
traffic system. Mostly the traffic management system in India that it works in response to the real time traffic conditions. We
consists of traffic lights that change periodically for fixed time. also come across scenarios of present traffic light system where
At peak hours the congestion due to traffic is very high. The even though there is no vehicle in a particular lane and heavy
present system is not adaptive to the traffic conditions and traffic on another lane, the traffic light on the latter lane is
number of vehicles in a particular lane. This makes the traffic green for a fixed time while the other remains red for the same
conditions more hectic and congested at times. time. Well, reasons like these are responsible for the increase in
B. Time, Pollution and emergency: congestion and pollution at traffic intersections. So what can be
solution to these problems? The answer to these problems is to
The average wait time is a very important parameter when have an adaptive traffic light system which works in response
it comes to traffic congestion monitoring. The average wasted to number of vehicles in particular lanes. This traffic system
time in traffic in 2015 according to INRIX 2015 was as high as counts the number of vehicles in a lane, compares the data
81 hours in United States (Los Angeles, CA) and 101 hours in from other lanes and depending upon the number of vehicles it
Europe (London Commute Zone, UK) [4]. Well, these data decides whether to give priority to reducing average waiting
show us a mirror to how inefficient the traffic management time or decreasing congestion /pollution in a lane.
system is around the world. Another important parameter that
describes the traffic congestion problem is Pollution. CO2 is a B. Smart and Real-time controlling:
major byproduct of vehicular emissions. According to World The system that we propose uses two controllers. It has
Bank Data Center (IBRD; IDA) the CO2 emissions in India been designed and implemented in such a way so that it can
was 1.66 metric tons per capita [3]. This is tremendously high. work in real time. In other words, there are two dedicated
One more statistic that shows a major flaw in present traffic controllers one responsible for counting the no. of vehicles and
system is high emergency ambulance response time. As per the another for using that data to light traffic signals to clear the
Report of the Working Group on Emergency Care in India traffic. To make it possible we use a request/respond technique
Ministry of Road Transport & Highways, Govt. of India, in between the two controllers so that the two controllers work
India, the average ambulance response time is 30 minutes and independently. Also since the processing goes in a parallel
is required to be reduced to as low as 10 minutes [5]. manner, real time calculations can be made. This ensures that
C. IoT Development in India: both the controllers work in real time and do not miss any kind
of data.
In past years the trends in technology change rapidly. One
of the new emergent in these changing trends of technology is C. Requirements:
Internet of Things. The Government of India has identified this To implement adaptive traffic controlling along with real
and introduced policies related to IoT. The Department of time calculations we require to have a framework that is
Electronics and Information Technology, (DeiTY), India has capable of counting the number of vehicles in a lane with as
come out with a draft IOT Policy document which focuses on precision as possible, irrespective of what task is being carried
following objectives: out by the microcontrollers. We need ultrasonic sensors that
• To create an IoT industry in India of USD 15 billion by can detect any obstacle in form of vehicles and count them.
2020. It has been assumed that India would have a share Also this framework is required to be supported by internet
of 5-6% of global IoT industry. along with network security support. This helps in making the
Adaptive Traffic Management System connected to different
• To undertake capacity development (Human & Internet of Things- IoT using systems, so that all these systems
Technology) for IoT specific skill-sets for domestic can be merged to work with each other providing better
and international markets. solutions. The code is dumped in this controller from internet.
• To undertake Research & development for all the This access can be given to city traffic planner and developers
assisting technologies. to control the traffic of an intersection from a remote distant
place. This enables traffic control unit at distant place to
• To develop IoT products specific to Indian needs in all monitor the status of traffic on a particular lane and use these
possible domains [6]. data for later analysis.
IV. IMPLEMENTATION We experimented the idea with two lanes and two traffic
density monitoring sensors. These sensors were used in the
A. Prototype: respective two lanes for counting the vehicles. There were two
The system that we intend to implement is very huge. We traffic lights for each lane and also two congestion indicator
developed a small prototype that works on the principle of lights in exactly previous intersections of respective lanes. This
Adaptive Traffic Management System. We here used two give an indication of congestion ahead to vehicle users/drivers
controller boards. One of them was responsible for calculating to choose the lane on which he/she wants to continue. The
the traffic density data and send the same to another controller traffic lights change in such a manner that the there is hardly
that is responsible for using that data against the code dumped need for a car to stop. This ensures an approach towards zero
in it through internet for controlling traffic lights and clearing average waiting time. It’s also been taken care of that if same
the traffic. This is done using Request/Response signals which logic is repeated for more than 4 times, the priority to next
allows parallel and real time operation of the two controllers. logic to execute. We used an algorithm in which data about the
traffic density from different lanes in a traffic intersection is
compared to each other. There are two parameters that was to
be decided. One was Lane Selection: here, whether lane 1 or
lane 2. Another parameter is traffic density: Low, Medium or
High. These parameters are send in form of signals to another
controller connected to Traffic Controller or developer via
internet.
1) Hardware used for prototype: To implement this
system, we require the following 4 main hardware and
computational tools:
a) Intel Galileo Gen 2 Board
b) Arduino Duemilanove Board
c) Wyliodrin IDE Studio
d) Arduino IDE 1.6.8
Also we require ultrasonic sensor HC-SR04 for counting
the number of vehicles in the respective lanes. These
sensors are connected to Arduino Duemilanove board that
communicates with Intel Galileo Gen 2 Board as shown in
Fig 4.1. The code is dumped using Arduino IDE 1.6.8. To
connect the Intel Galileo board to internet via Ethernet
cable we need to first create our account at Wyliodrin
website. Then it is required to add Intel Galileo Gen 2
board and download the Wyliodrin.json file. We need to
download the “iot-devkit-latest-mmcblkp0.direct.img” file
and then write this image to a SD Card of minimum 4GB.
Also copy the Wyliodrin.json file to it. This file contains an
IP which will be assigned to the Galileo Board and hence
when this SD Card will be inserted the board will be
registered to Wyliodrin IDE and get online [8]
2) Test conditions and Results of Prototype: The results of
the implemented system are as shown in table below. These
Fig. 4.1. Block Diagram of Prototype of Proposed are results for green on time set as 3s for low traffic, 5s for
Traffic Management System
medium traffic and 8s for high traffic.

TABLE I. Results and Test Conditions

No. of Vehicles Actual no. of
Sr. No. Status Priority
detected vehicles
Lane 1:- 5 Lane 1:- 5 Lane 2 Green 3s ; Lane 1 Green 8s
1. Low Average Waiting Time
Lane 2:- 1 Lane 2:- 1 Congestion message on Lane 1 displayed

Lane 1 Green 8s ; Lane 2 Green 8s

Lane 1:- 6 Lane 1:- 6
2. Congestion message on Lane 1 and Lane 2 Low Traffic Congestion
Lane 2:- 5 Lane 2:- 5
displayed

Lane 1:- 0 Lane 1:- 0 Lane 1 Green 3s (4 times and then priority shifts) Changing priority to not miss
3. any vehicle if same logic
Lane 2:- 0 Lane 2:- 1 Lane 1 Green 3s (4 times and then priority shifts) repeats itself more than 4 times
B. Sensing Traffic Density:
For our prototype we have used ultrasonic ping sensors
HC-SR04 for calculating traffic density. The ultrasonic ping
sensors are triggered by high pulse signal and they give result
in form of distance. We have set a logic to increase the count
by 1 when the distance changes from a non-zero value to zero
value. When there is no vehicle the distance sensed will be
more than maximum set distance value, hence the distance
calculate will be zero. This way vehicles are counted
continuously by ultrasonic sensors on respective lanes.
C. Algorithm:
Since there are two boards controlling the entire operation,
there are two algorithms for them. The algorithm that we have
shown here is generalised for 2, 3, 4, …..., n number of lanes
where n is a positive integer. The algorithm for working of
controller connected to traffic density monitoring sensors, here
HC-SR04 sensor and decide lane priority and traffic level is
shown in fig 4.2. Also the algorithm for controller connected to
internet and the latter controller is shown in fig 4.3.
D. Scalability and Flexibility:
These logic and implementation is not limited to 2 or 4
lanes. There are certain parameters that can be scaled as per the
traffic situations. These are maximum distance of ultrasonic
sensors, number of ultrasonic sensors, number of traffic lights
and traffic signal: Green or Yellow or Red, On Time
respectively. The number of ultrasonic sensors and traffic lights
can be scaled to n times where n is any positive real number.
The maximum distance can also be set by traffic controller as
desired. There can be more than 2 traffic levels to increase the
dynamic range of the system. For the prototype we have used
term lane, i.e. roads with single one way partition. Although
this can be scaled for partition having multiple lanes for fast-
moving to slow-moving vehicles. In case of multiple lanes,
each lane will have a sensor for counting of vehicles and the
sum of count of number of vehicles in each lane will be
parameter for taking decision on decides whether to give
priority to reducing average waiting time or decreasing
congestion /pollution. Integrating this system with Central
Computing System can help us to get data from various traffic
junctions. These data can be used to identify region of high
congestion and peak traffic hours, and government can take
necessary steps to counter the same. Also different parameters
for different traffic junctions can be set from this central
computing system in case of malfunction of any traffic
junction. The sensing methodology can be changed and thus
system is more flexible.
V. FUTURE SCOPE:
A. Integration with ITS:
This system that we tried to implement is adaptive to the
real time traffic conditions. The algorithm works in such a way
that the average wait time is reduced considerably low. In case
of low or medium traffic, this system can work as zero wait
time traffic system. This system possess immense potential to
be a part of an Intelligent Transportation System ITS. ITS is a
very huge system. Actually it is in general a system of many
smaller traffic management system. This adaptive traffic
management system can be one of such small system. Thus in
future the proposed traffic management system can be
integrated with ITS.
Fig. 4.2. Logic diagram of priority and traffic level
 B. Simplified / innovative method for sensing
(antenna based sensor):
Traffic generally is not just about vehicles in a lane, its also
about number of pedestrians, emergency vehicles, public
transport etc. To integrate the proposed adaptive traffic
management system with parameters as mentioned before, we
will require more simplified but innovative methods for
sensing the traffic. These can be antenna based sensors that can
transmit and receive signals about density of traffic in a lane or
a camera based sensing system that can identify images of
vehicles, pedestrians, emergency vehicles, etc. as well as
distinguish them from each other and use these data to take
decisions based on the algorithm much similar to the proposed
Adaptive Management System.
C. Implementations of system to real environment:
In reality traffic is not just about vehicles, its about
pedestrians, locality i.e. highly dense or low density region,
and also on traffic density relation with time (whether peak
time traffic or morning traffic). The system that we have tried
to implement can further be enhanced by sensing of even
pedestrian traffic. This will increase the efficiency of traffic
clearance. Also an algorithm can be designed that senses which
time it is, whether peak time or morning time and on its own
decide the time for which green light of a lane should remain
on. This will make the entire traffic management system more
reliable and easily acceptable.

VI. CONCLUSION
The system that we have tried to implement here is a part of
a larger system of Intelligent Transportation System (ITS) and
Internet of Things (IoT). As an approach to this we have made
an effort to make this system more reliable and provide very
low average waiting time. Also there is a provision to change
the priority of lane selection if same logic repeats itself more
than four times. This ensures that even if by chance a vehicle is
not detected, it is given the priority after four times. This shows
the robustness of the system in case if there is a system failure.
Since there are smart cities coming up all across the world, this
system provides a more economical and feasible approach
towards changing the present traffic management system and
making it adaptive and intelligent. Since the system remains
connected to the cloud via internet, the traffic controller and
developer sitting at far distant place can always monitor the
data related to the traffic situations of the region that he/she
desires to monitor. Also the proposed adaptive traffic
management system is flexible with regards to the type of
sensors to be used for sensing the traffic density. Thus the
robustness of the system, the adaptive nature of the proposed
system, real time processing of the data makes the system ideal
to be implemented in real time.

ACKNOWLEDGMENT
The authors acknowledge ECE department, Charusat, for
providing work environment and hardware-software resources
Fig. 4.3. Logic diagram of controller connected to internet while designing and testing this project.
 REFERENCES 5. Report of the “Working Group on Emergency Care in India”, Ministry
of Road Transport & Highways, Govt. of India.
1. M.A.Chowdhury, A.Sadek, “Fundamentals of Intelligent Transportation
Systems Planning, Artech House”, London 2003. 6. Revised Draft IoT Policy on “Internet of Things”, Department of
Electronics & Information Technology(DeitY), Ministry of
2. Office of the Assistant Secretary for Research and Technology (OST-R) Communication and Information Technology Government of India;
• U.S. Department of Transportation (US DOT); Online Available : Online Available : http://deity.gov.in/
http://www.rita.dot.gov/
7. Intelligent Transportation System (ITS); Research and Development in
3. World Bank national accounts data, and OECD National Accounts data electronics; Industrial and Electronics Application Development
files. Open License. Catalog Sources: World Development Indicators; Division, Department of Electronics & Information Technology(DeitY)
Online Available : http://www.worldbank.org/ Ministry of Communication and Information Technology Government of
4. INRIX 2015 Scorecard Website; Online Available : http://inrix.com/ India; Online Available : http://deity.gov.in/
scorecard/ 8. The IDE for IoT; Online Available : https://www.wyliodrin.com/