2017 International Conference on Emerging Trends & Innovation in ICT (ICEI)

Pune Institute of Computer Technology, Pune, India, Feb 3-5, 2017

TrafficIntel
Smart Traffic Management for Smart Cities

Anurag Saikar1 Mihir Parulekar1 Aditya Badve1

[email protected] [email protected] [email protected]

Sagar Thakkar1 Aaradhana Deshmukh2

[email protected] [email protected]

1
Student, Department of Computer Science, S.K.N.C.O.E, Pune
2
Ass. Professor, Department of Computer Science, S.K.N.C.O.E, Pune

Abstract— In India the unbalanced growth in the freight entirely inefficient in long term, there is need to utilize the
volume as compared to the road length has led to severe issues of existing infrastructure to its maximum capacity. This can be
traffic congestion, road safety and health hazards, which, as done by applying intelligence to the existing infrastructure, i.e.
estimated by recent studies will surely escalate very rapidly switching from static road traffic management system to a
within coming years. The use of a real time dynamic traffic
management system to intelligently navigate the vehicles can
more intelligent and dynamic real time traffic management
optimize the traffic flow on the roads, thus utilizing the system. considering the high dynamicity of the traffic flow
infrastructure efficiently and providing an environment along with polynomial increase of number of vehicles on the
conducive to solving gridlock in cities. The authors propose a real road networks makes management of the said traffic flow
time traffic management system (RTMS) consisting of real time through a RTMS much more complex [6]. Furthermore the
traffic monitoring system formed by small network of road side solution to a traffic congestion usually tends to be a series of
units, junction units and mobile units to dynamically decide the small intricately correlated solutions, applying only one of
time of traffic lights to discourage formation of gridlock, coupled which, while no doubt useful, generates insubstantial results.
with a web based application for vehicle drivers that will derive Here the authors propose a system consisting of real time
the data from real time traffic analysis to indicate the local traffic
traffic analysis, the data reported by which will further be
flow and suggest the incoming vehicles to make use of alternative
routes in order to further alleviate growth of the gridlock. derived by a web based application to prevent, contain and
disseminate the traffic congestion in real time. The system
Keywords — traffic congestion, real time traffic management consisting of a RSU (Road Side Unit), JU (Junction Unit) and
system, smart traffic lights MU (Mobile Unit) works on various levels to solve the
congestion. A part of the traffic analysis system i.e. the JU
I. INTRODUCTION will work at an intersection by communicating with the MU
(vehicles) to gauge the traffic density and dynamically set the
The road traffic, being versatile and economical is considered
time for traffic lights depending on it so as to maintain a
to be the most common mode of transportation in India.
constant flow of vehicles and if necessary, disperse the traffic
However, according to a study [18] by The Transport
quickly without letting it accumulate and form a congestion.
Corporation of India (TCIL) and IIM (Calcutta) on operational
While RSU’s will, by, operating on longer roads without
efficiencies of freight transportation by roads the traffic
intersections and communicating with local MU’s, try to
congestion has become such a dire situation that on some key
ascertain the traffic flow and formation of congestion. On
corridors such as Mumbai-Chennai, Delhi-Chennai and Delhi-
confirmation of a congestion the RSU’s will transfer this
Guwahati the vehicles crawl at an average speed of less than
information to central server via internet. The web application
20 kmph while it's only 21.35 kmph on Delhi-Mumbai stretch.
will, on suspicion and confirmation of a congestion by RSU’s
Consequently the Indian economy faces losses of around
in a particular area, inform the incoming vehicles and suggest
60,000 crore every year, which translates roughly to a
usage of another route thereby inhibiting the further growth of
staggering amount of $10.82 billion each year not accounting
the congestion, additionally the vehicles already within the
the subsequently engendered fatal environmental damages,
congestion can be notified to move synchronously to optimize
losses due to delay in receiving emergency services such as
the traffic flow by minimizing the progressive delay that
medical help etc. The study reports that while India's freight
occurs due to slower human reaction times when the stagnant
volume was increasing at a compounded annual growth rate of
traffic stream starts moving.
9.08% and vehicles were growing at 10.76%, the road length
In developing countries like India where the growth rate of
was increasing at only 4.01%. This has resulted paucity of vehicles outruns the growth of road length, the traffic
road space to accommodate vehicles and to increase the speed management is vital, as without it the increasing severity of the
ensuing the traffic congestions that lead to the huge gridlocks. congestion and its subsequent adverse effects are imminent. In
To overcome the gridlocks without modifying the existing this paper the authors have proposed a system as an attempt for
infrastructure, which proves to be cumbersome, expensive and overcoming the infrastructural and other issues to and

978-1-5090-3404-8/17/$31.00 ©2017 IEEE 46

providing the panacea to traffic congestion and its subsequent their proposed device is installed is monitored using a camera.
effects. The camera takes photos of traffic situations after set interval
of time and stores them in database for further viewing. The
II. LITERATURE REVIEW officer at that intersection can then take a look at these photos
and decide on the traffic light timing pattern from his
The recent increase in urbanization and motorization has been
smartphone.
perpetuating for the frequent traffic congestion and
development in the field of intelligent transportation system.
In [12] Petrovska, N., & Stevanovic, A. propose use of image
While a substantial amount of solutions have been proposed to
processing to determine traffic density. They then display the
assuage the issue, a lot of them put forth the solution in form
information on traffic congestion in various locations using
of static traffic management perspective [16], which proves to
their application for smartphones. For doing this they use
be capricious at times, considering the dynamicity of the
Google’s API for interfacing with the traffic layer of the
traffic flow. The real time traffic management model, while
Google maps. The traffic density in areas is then displayed
considerably more lucrative, the design of which, however
using color codes each denoting very high congestion to no
tends to be increasing complex on a large scale owing to the
congestion.
variable and fluid nature of the traffic stream [16]. For eg. In
[6] Kammoun, H. M., Kallel, I., Alimi, A. M., & Casillas, J.
In [3] a simple approach is used to detect traffic congestion in
propose a hybrid method that makes use of adaptive vehicular
any area in a developing region. Here, an application on the
guidance systems to analyze the road traffic network, the
user’s devices will pull their GPS and speed information
traffic is adjusted intelligently by suggesting alternative path
further sending this data to the server. Once the data reaches
to the destination based on ant colony behavior and
the servers, the data is analyzed and presence of congestion is
hierarchical fuzzy system, while in [13] Shashikiran, V.,
confirmed algorithmically. The user is then made aware of the
Kumar, T. S., Kumar, N. S., Venkateswaran, V., & Balaji, S.
congestion on the respective route, via the application, so that
propose a solution that makes use of Kruskal’s algorithm for
the user may choose to take a different route. This system is
suggesting optimal path by mapping the traffic junctions to
highly efficient and optimized as it requires very less
nodes and traffic flow as link weight by using Google maps.
infrastructure development and instead relies on
The selection of optimal path is based on various parameters
crowdsourcing for the data.
such as traffic rate, speed of the vehicle, shortest path etc. a
traffic in-charge personnel has to update the traffic
Sowmiya, A., & Ram, N. P. in [15] propose a tracking device
information periodically through their mobile in order to
placed every 250 meters apart on roads. A device placed
accomplish this.
within the vehicles, is capable of communicating the vehicle’s
speed and some other relevant information with the tracking
Nafi, N. S., Khan, R. H., Khan, J. Y., & Gregory, M. in [9]
device. The tracking device after receiving this information
propose a system using IEEE 802.11p based vehicle to
about vehicles within its range, will calculate custom speed for
infrastructure communications system (VANET) to predict the
every vehicle with which it should travel, in order to avoid all
future traffic intensities at intersections. The vehicles are
red lights on its route. Thus saving fuel as well as reducing
rerouted based on this prediction to reduce the traffic
congestion along various routes.
congestion and minimize the driving time of the individual.
In [13] a design for traffic management system has been
In [5] Jayapal, C., & Roy, S. S. suggest use of RSU (Road side
proposed for intersections, which makes use of long range
units) and mobile phones to detect traffic congestion in a
photoelectric sensors placed strategically to monitor traffic
certain area. The mobile phone and RSU communicate with
and decide the time of traffic lights accordingly. The increase
each other wirelessly via a vehicular ad hoc network
or decrease in time of traffic light is made only in case of
(VANET). The distance travelled by the vehicle in particular
suspicion of large number of vehicles i.e. emergency situation,
interval of time is detected. If this value is below a set
elsewise the system switches to the default system.
threshold, congestion is suspected. If other vehicles in the
same area show similar signs of congestion, then congestion is
confirmed. Once congestion is confirmed vehicles
Reference [9] proposes a system for cognitive traffic
approaching the congested route are notified about the same
management that primarily focuses on smart traffic lights. The
using the mobile app or via a display near the traffic signals.
dynamic timing of these smart traffic lights is based on the
conclusions drawn from the analysis of data obtained from
Reference [7] proposes a system which dynamically lets the
cars and road sensors, as well as alternative sources of user
officer in a particular intersection decide the traffic lights
data, such as social networks, the results of opinion polls and
timing pattern after he has taken a look at the traffic using his
others.
smartphone. The system also allows emergency vehicles to
claim their right of way by allowing the authorized personnel
From the systems discussed above it can be discerned that
inside the emergency vehicle to change the traffic light pattern
most of them, while making use of real time data, only suggest
and status. The traffic flow at a particular intersection where

47
alternate path on its basis, which has very little effect on traffic for further analysis of vehicular density for gauging the
traffic congestion, that too on a local level. Also if all the formation of congestion. The MU’s i.e. the tags placed inside
vehicles make use of the same guidance system it leads to them continuously broadcast RF beacons, once the MU arrives
concentration on other routes [17]. with the range the signal is sensed by the RSU.
The RSU’s are to be placed on the on the length of the roads
III. PROPOSED SYSTEM with an appreciable distance between them (about 200 meters)
The proposed system consists of three real time components [8]. The RSU (router) forwards this data to the appropriate
the RSU, the JU and the MU. The system makes use of active wireless coordinator, which relays the data to the server for
RFID (based on IEEE 802.15.4 protocol, 2.4 GHz ISM band), further analysis. Ideally the MU or the vehicle travels from
the specific RFID tags are to be kept in the vehicle, both these router to junction coordinator, i.e. the vehicle passes through a
coupled together form the MU i.e. the mobile unit. The RSU’s number of RSU’s towards the JU.
contain, a RFID reader together termed as a wireless router, The JU’s work at the intersections, acting partly as RFID
with an assigned gateway (i.e. a wireless coordinator) for a readers to measure the vehicle density at each side of the
particular set of RSUs. While the JU’s contain, within them, a intersection while simultaneously reading the data updated on
processor for the required computation, a controller for traffic the server by the RSU’s to estimate the incoming traffic along
lights and a RFID reader. The Tags emit radio signals that can each direction, this data, coupled with the wait time of the
be captured by devices like routers or coordinators. Routers vehicles at the junction and the distance of each of the vehicle
are capable of capturing tag data and relay the captured data to from that particular junction are the factors based on which the
either coordinator or another router in its range. Coordinators weight for a route of the road on the intersection will be
have a serial interface through which external GSM/ GPRS decided. These weights, calculated by the JU, for each
devices can be interfaced, thus enabling the further segment of the road on the intersection are measured against a
communication with the centrally located server. threshold to decide the time of the traffic light for the same.
The distance plays a major role in deciding the contribution of
an individual MU to the total effective weight of the segment
of the road, leading to incremental increase in contribution of
MU towards the effective weight of the road segment as it
nears the particular junction, this allows for easier
coordination between the JU’s to minimize the overall idle
time. Furthermore this enables the whole network of JU’s in
synchronicity to minimize, and in case it occurs, quickly
prevent the congestion from growing further.
The data updated by the RSU’s is also used to calculate the
lane velocity i.e. the average velocity of the vehicles on the
specific road segment. This can be done with using the
location of RSU’s and the timestamps for when the MU
arrives in its range, this helps in confirmation and later with
gauging the intensity of the congestion. This data can further
Figure 1 : General Architecture of the System be relayed by the web base application to the users to allow
selection optimal path towards the destination, and in case of
The figure 1 shows the general architecture of the system, congestion, the active users headed to the congested route can
depicting the data flow between the MU, the RSU, the JU and take an alternative route thereby reducing the glut of vehicles
further the server. The RSU's and the JU's sense the data contributing to the traffic congestion and disseminating the
associated with the specific RFID tags, identifying the MU's traffic over various available routes.
in their locality and further the vehicular density of the same. The system aims at synchronizing the traffic lights to
This data is further reported to the centrally locate server via minimize the possibility of formation of congestion, and in
internet, to be used by the web based application, amongst case of congestion, the system works on multiple levels to
other things. The data reported and updated by all the other rectify the traffic congestion. Firstly it tries to resolve the
RSUs and the JUs is read by a specific JU from the centrally congestion owing to the dynamicity of traffic lights time and
located server so as to ensure proper synchronization, thereby secondly prunes its further growth by alerting the incoming
minimizing the overall idle time of the vehicles and vehicles of the congestion and suggesting the possible
furthermore discouraging formation of congestion or alternative routes. Furthermore, it reduces the overall idle time
gridlocks. of the vehicles during the travel, thereby minimizing the fuel
wastage.
The RFID tags placed within the MU’s are specifically used to
identify the type of the vehicle. This helps in categorization of

48
 V. CONCLUSION
Rapid urbanization in India has led to an increase in the
number of personal transport vehicles as well as public and
commercial transport vehicles while development of road
infrastructure however has not been commensurate. This fuels
an ever growing problem of traffic congestion, which despite
enforcement of new regulations about freight volume has been
a massive concern. This problem can be resolved by either
improving the road infrastructure or increasing the efficiency
of traffic on the current roads. The latter seems the apparent
choice as it requires comparatively negligible investment of
extraneous resources. Here the authors propose simple system
that, by making use of the RFID technology, aims at enabling
reduction in traffic congestion and also saving a lot of time and
fuel for the growing economy of India.

REFERENCES
[1] Burton, R. S., Thomas, N. C., & Wright, D. (2000). Wireless
Figure 2 : Schematic of the system communications to road users and within urban traffic management and
control systems. In Road Transport Information and Control, 2000.
The above image represents a schematic of the system spread Tenth International Conference on (Conf. Publ. No. 472) (pp. 152-157).
IET.
over several blocks. A JU is placed at each intersection which [2] Chorlton, C., & Fraser, I. (1998, April). Traffic monitoring on all-
senses the vehicles within their range, thereby monitoring the purpose roads for strategic network management. In Road Transport
traffic at the intersections. The overall traffic i.e. the traffic Information and Control, 1998. 9th International Conference on (Conf.
density on the roads and at the intersections can be monitored Publ. No. 454) (pp. 125-129). IET.
[3] Dhakad, R., & Jain, M. (2014, December). GPS based road traffic
and reported, with the help of network of RSUs and JUs, to congestion reporting system. In Computational Intelligence and
the centrally located server for further use by the web based Computing Research (ICCIC), 2014 IEEE International Conference
application and for further processing. Each JU consists its on (pp. 1-6). IEEE.
own processor for faster, efficient and real time computation, [4] Gregoriades, A. (2007, December). Towards a user-centred road safety
management method based on road traffic simulation. In 2007 Winter
furthermore local computation enables the system to compute Simulation Conference (pp. 1905-1914). IEEE.
in real time without introducing considerable lag. [5] Jayapal, C., & Roy, S. S. (2016, March). Road traffic congestion
management using VANET. In 2016 International Conference on
IV. RESULTS Advances in Human Machine Interaction (HMI) (pp. 1-7). IEEE.
[6] Kammoun, H. M., Kallel, I., Alimi, A. M., & Casillas, J. (2011, April).
Improvement of the road traffic management by an ant-hierarchical
fuzzy system. In Computational Intelligence in Vehicles and
Transportation Systems (CIVTS), 2011 IEEE Symposium on (pp. 38-
45). IEEE.
[7] Malla, A. M., & Sahu, R. K. (2013). A Review on Vehicle to Vehicle
Communication Protocols in VANETs. International Journal of
Advanced Research in Computer Science and Software
engineering, 3(2).
[8] Mandal, Koushik, et al. "Road traffic congestion monitoring and
measurement using active RFID and GSM technology." 2011 14th
International IEEE Conference on Intelligent Transportation Systems
T (ITSC). IEEE, 2011.
Figure 3 : Overview of the Web based Application he [9] Misbahuddin, S., Zubairi, J. A., Saggaf, A., Basuni, J., Sulaiman, A., &
The information updated by the network of JUs and RSUs in Al-Sofi, A. (2015, December). IoT based dynamic road traffic
the central server are reflected onto the web based application, management for smart cities. In 2015 12th International Conference on
relaying the status of the vehicular flow on the route to the High-capacity Optical Networks and Enabling/Emerging Technologies
(HONET) (pp. 1-5). IEEE.
users. This allows the vehicle drivers to choose an appropriate [10] Miz, V., & Hahanov, V. (2014, September). Smart traffic light in terms
path to the destination and in case of a significant congestions, of the Cognitive road traffic management system (CTMS) based on the
to make use of an alternate route thus helping not only to rein Internet of Things. In Design & Test Symposium (EWDTS), 2014 East-
in the growth of the congestion on the road in question but also West (pp. 1-5). IEEE
to curb wastage of resources such as fuel, while ensuring a [11] Nafi, N. S., Khan, R. H., Khan, J. Y., & Gregory, M. (2014, November).
A predictive road traffic management system based on vehicular ad-hoc
smooth and faster travel with less idle time. This allows for the network. In Telecommunication Networks and Applications Conference
traffic in already congested area to be disseminated fast. (ATNAC), 2014 Australasian (pp. 135-140). IEEE.
[12] Patel, H. N., Desai, K., & Panchal, T. (2015, December). An algorithm
for automatic license plate detection from video using corner features.
In 2015 International Conference on Information Processing (ICIP) (pp.
301-306). IEEE

49
[13] Petrovska, N., & Stevanovic, A. (2015, September). Traffic Congestion [19] Yu, S., Mabu, S., Mainali, M. K., Shimada, K., & Hirasawa, K. (2011,
Analysis Visualisation Tool. In 2015 IEEE 18th International October). Dynamic traffic management model for real world road
Conference on Intelligent Transportation Systems (pp. 1489-1494). networks. InSystems, Man, and Cybernetics (SMC), 2011 IEEE
IEEE. International Conference on (pp. 2197-2202). IEEE.
[14] Report on Operational Efficiency of Freight Transportation by Road in
India http://www.tcil.com/tcil/pdf/TCI%20IIM%20Study%20Report-
2011-12.pdf (last accessed on 15/9/2016)
[15] Salama, A. S., Saleh, B. K., & Eassa, M. M. (2010, November).
Intelligent cross road traffic management system (ICRTMS).
In Computer Technology and Development (ICCTD), 2010 2nd
International Conference on (pp. 27-31). IEEE.
[16] Shashikiran, V., Kumar, T. S., Kumar, N. S., Venkateswaran, V., &
Balaji, S. (2011, June). Dynamic road traffic management based on
krushkal's algorithm. In Recent Trends in Information Technology
(ICRTIT), 2011 International Conference on (pp. 200-204). IEEE.
[17] Sowmiya, A., & Ram, N. P. (2015, March). An intelligent approach for
effective road traffic management. In Innovations in Information,
Embedded and Communication Systems (ICIIECS), 2015 International
Conference on(pp. 1-4). IEEE.
[18] Vrancken, J. L., & Kruse, O. C. (2006). Intelligent control in networks:
The case of road traffic management. In 2006 IEEE International
Conference on Networking, Sensing and Control (pp. 308-311). IEEE.

50