ISSN: 2277-3754

ISO 9001:2008 Certified

International Journal of Engineering and Innovative Technology (IJEIT)
Volume 5, Issue 6, December 2015

Traffic Management Using Swarm Intelligence

and Route Selection Using Android Application
Santwana Panda, Anjali M. Patki, Kedar Hushing
updates about the traffic congestion on each route, so that we
Abstract— With the ever increasing traffic demand, congestion can select the best route for our travel.
has become a serious problem in many major cities around the
world. This paper deals with the traffic congestion and also II. RELATED WORK
provides the citizen an alternate, less congestion route for their
travel, thus in turn saving a lot of time, fuel and also control Some of the optimization technique for the road network
pollution . Swarm intelligence is implemented in this paper. The management proposed by the researchers is grouped below,
embedded system using swarm intelligence proposed in this paper along with their strength and limitations.
captures images of the traffic and analysis their behavior
according to the surrounding and redirects the traffic in different A. Graph Theory Based
routes so that congestion in each routes will decrease. One more The classical minimum shortest route algorithm such as
new feature is added that is android application designed through Dijkstra algorithm, Priority queues, bidirectional search etc.
App-Inventor that helps to locate the least congestion route for a
are used by many research for road traffic management [1].
traveler based on his or her destination.
Appert et al. [2] utilized graph theory for the measuring urban
Index Terms—App-Inventor, Embedded System, Swarm road network vulnerability. Baruah and Baruah [3] proposed
Intelligence, Traffic Congestion. cut-set of graph for the traffic control problem. As the
complexity of traffic control on network expansions it
I. INTRODUCTION becomes more complicated to coordinate the actions of the
large number of heterogeneous traffic management
Swarm Intelligence (SI) describes the collective behavior instruments that are available in the network. One way of
of decentralized, self-organized systems. It is the technique handling this complexity is to divide the coordination
wherein several decentralized intelligent independent units problem into smaller coherent sub-problems that can be
function and aid each other in task. SI provides a basis with solved with a minimum of interaction. Multi agent systems
which it is possible to explore collective (or distributed) can aid in the distribution of the problem (over the various
problems solving without centralized control or the provision agents that comprise the multi agent system) and facilitate the
of a global model. Many units can function in tandem, coordination of the activities of these agents when required. In
reducing load and increasing efficiency. This also removes the literature no consensus exists about the best configuration
the need for a central server. In SI there are several units of the traffic managing multi agent system and how the
working together, helping each other without any central activities of the agents that comprise the multi agent system
control over them. should be coordinated [4] . Katwijk et al. [4] reported a test
For every square there will be four signals and each will be bed for multi agent control systems in road traffic
equipped with a camera. Cameras will take snapshots of the management that can deals the traffic managing multi agent
traffic and directly send to cloud for storage. From cloud the system can be configured, evaluated in a realistic simulated
images will be used to detect the intensity of traffic at each traffic environment, easily transferred to a real world
junction. Through edge detection process intensity on application. Raza and Rao [5] proposed agent based urban
incoming traffic can be judged and the flow of traffic from traffic and transportation control. This paper gives a
each direction also can be estimated. Then through ARM7 theoretical foundation of an intelligent traffic clouds.
microcontroller will direct the traffic lights according to B. Genetic Algorithm based
traffic flow, for a limited number of vehicles the green light
will glow for lesser amount of time and for a large amount of Genetic Algorithms (GAs) have been demonstrated to be a
vehicles rushing the green light will glow for a more amount promising search and optimization technique. It has been
of time. This was the scenario for a single traffic square but successfully applied to system identification and a wide range
for a the entire city swarm intelligence is used all traffic of applications including filter design, scheduling, routing,
signals are interconnected through a network and through control, and others. For applying GAs to complex problems
swarm intelligence flow of traffic is control adaptively and has been the high computational cost due to their slow
directed to different routes in order to minimize congestion at convergence rate is one of the main obstacles. Han and Tabata
every signal. So for a smart city this traffic management can [6] combined a genetic algorithm and controlling lethal gene
be applicable. for solving of the vehicle routine problem but the
Through App-Inventor using block programming an performance for the practical example was not investigated.
android application is designed through which we can get Meshkat and Vrancken [7] used multi objective technique for
the road network partitioning. This study fast and elitist
Non-dominated Sorting Genetic Algorithm (NSGA-II) and

59
 ISSN: 2277-3754
ISO 9001:2008 Certified
International Journal of Engineering and Innovative Technology (IJEIT)
Volume 5, Issue 6, December 2015
Pareto Archived Evolution Strategy (PAES) were hierarchical fuzzy model. This system allows adjusting
implemented. Jiang et al. [8] proposed an agent model with efficiently the road traffic according to the real time changes
adaptive weight based multi-objective algorithm to manage in road networks by the integration of adaptive vehicle route
road-network congestion problem. The aim of this study was guidance system. This system was implemented and
to construct a quantitative index series to describe the road simulated under a multi agent platform in order to discuss the
network congestion distribution, and use such indexes as improvement of the global road traffic quality in terms of
weights in the multi-objective algorithm to shunt vehicles on time, fluidity and adaptively.
those congested links. In the first phase, a multi-agent system
D. Swarm Intelligence Based
was built, where each agent stands for a vehicle that adapts its
route to real-time road network congestion status by a The swarm intelligence has been used to model complex
two-objective optimization process: the shortest path and the traffic and transportation processes. In fact, the
minimal congested degree of the target link. The agent-based self-organization of the social insects is based on relatively
approach captures the nonlinear feedback between vehicle simple rules of individual insect‟s behavior. Among the
routing behaviors and road-network congestion states. Next, a different colony insects, the ant colony succeeds to find food
series of quantitative indexes was constructed to describe the by following the path with highest pheromone quantity
congested degree of nodes, and such indexes were used as deposited by other ants [12]. The pheromone signal
weights in the two objective functions which were employed represents the communication tool between individual ants. It
by the agents for routing decisions and congestion avoidance. contributes to the formation of collective intelligence of social
ant colonies that can be considered as multi-agent systems.
C. Fuzzy Logic Based Bertelle et al. [27] proposed road traffic management by using
The fuzzy logic appeared in 1965 by Zadeh introducing the ant system for shortest path in weighted dynamic graph. This
concept of fuzzy sets. It was shown as a very capable method utilized neural networks for traffic flow regulation
mathematical approach for dealing with subjectivity, and it simulated using multi agent platform. Yang et al. [28]
ambiguity, uncertainty, and imprecision [13]. Fuzzy logic was proposed optimization model based on coarse-grain parallel
used as a framework to solve transportation problems such as ant colony algorithm for the bus network optimization. It was
traffic assignment problem, accident analysis and prevention, demonstrated on data of Dalian city, China but it did not
traffic control at roads intersection, and traffic light control. consider the real time traffic management. Deng et al.[29]
During the last decade, some developments in information proposed hybrid particle swarm optimization algorithm by
acquisition technologies through advanced traveler combining fluid neural network. This method is influenced by
information systems have been done. However, many search best path in stochastic traffic networks and this method
contextual factors (such as departure time, travel distance, was simulated with only 20 nodes road network. D‟Acierno et
usual driving speed of the driver, weather information, al. [30] proposed swarm intelligence algorithm to optimize
personal preferences, roadwork information, and other the signal setting of each intersection for the asymmetric
information which could be available to the guidance systems traffic assignment and it lacks on real time management.
in real-time) increase the uncertainty of the itinerary choice. Garcia-Nietoa et al. [31] used particle swarm intelligence to
Ridwan [22] used choice function based fuzzy preferences find cycle programs of traffic lights and implemented for 2
relations and considered the spatial knowledge of individual cities in Spain. Mostly, Ant Colony Optimization was used to
drivers. This method strengthened the travel decision by fuzzy solve transportation problems such as Travelling Salesman
preference relations but it utilized small number of influence Problem (TSP) and Vehicle Routing Problem (VRP), only
factors and in real scenario there are multiple influence few works based on swarm intelligence are developed to solve
factors. Hawas [23] estimated the route utility by using road traffic management problem [13]. In fact, the problem
neurofuzzy data training with a hidden neuron in each fuzzy cannot be solved using the classic versions: artificial ants are
process. This method used adaptive to the variation of able only to generate successively shorter feasible tours by
perceptions from drivers but there is no fuzzification training using information accumulated in the form of a pheromone
not exist. Arslan and Khisty [24] developed route choice trail deposited on the graph edges.
model. They utilized hybrid model based on fuzzy logic and
analytical hierarchy process. The preference was extracted III. WORKING OF THE PROPOSED SYSTEM
from driver‟s psychology. Ghatee and Hashemi [25] proposed Our aim is to develop the system at signals; this system will
quasi logist formula based algorithm for traffic assignment. It have multifunctional operations. Initially the system will
maximizes the level of certainty and minimizes the perceived measure the traffic density at different signals using image
travel delays. The limitation of this study is no results for real processing using edge detection and accordingly change the
networks. Balaji and Srinivasan proposed multi agent system time delays for traffic lights viz. the side at which the traffic is
based on type-2 fuzzy decision module for urban traffic high the signal will remain green for more time. Secondly it
management. This method reduces the total delay of vehicles will also communicate with the adjacent junction signal. Both
and it was simulated on real traffic of Singapore. The the signals will collectively manage the traffic depending on
limitation of this method is unavailability of vehicle route the density. So in the same way all signals of the city will
guidance. Kammoun et al. [9] [10] proposed an adaptive multi communicate with each other forming a SWARM dedicated
agent system based on the ant colony behavior and the

60
 ISSN: 2277-3754
ISO 9001:2008 Certified
International Journal of Engineering and Innovative Technology (IJEIT)
Volume 5, Issue 6, December 2015
system. Also send message to next signal when panic key is serially between two devices, that devices must be TTL
pressed using Wi-Fi module. compatible. TTL compatibility means TTL logic levels (logic
1/ logic 0) must be same for both the devices. MAX232
A. Edge Detection
converts TTL of 5v in to RS232 standard or RS232 standard
Edge detection is an image processing technique for in to TTL of 5v. Whereas MAX3232 converts TTL of 3.3v in
finding the boundaries of objects within images. It works by to RS232 standard or RS232 standard in to TTL of 3.3v.
detecting discontinuities in brightness. Edge detection is used
for image segmentation and data extraction in areas such as C. ARM7 Microcontroller
image processing, computer vision, and machine vision.[32] The original ARM7 was based on the earlier ARM6 design
There are many methods for edge detection but most of them and used the same ARMv3 instruction set. The ARM710
can be grouped into two categories, search based and variant was used in a CPU module for the Acorn Risc PC.
zero-crossing based [33]. Common edge detection algorithms This forms the core of our system, where decisions based on
include Sobel, Canny, Prewitt, Roberts, and fuzzy logic the inputs from the computer are made and respective LEDs
methods. are made to glow for desired time intervals. It also controls the
Wi-Fi module which is responsible for the communication
between the signals.
D. Wi-Fi Module (ESP8266)
ESP8266 offers a complete and self-contained Wi-Fi
networking solution, allowing it to either host the application
or to offload all Wi-Fi networking functions from another
Fig.1 Image segmentation using Sobel method [32] application processor. When ESP8266 hosts the application,
and when it is the only application processor in the device, it
is able to boot up directly from an external flash. It has
integrated cache to improve the performance of the system in
such applications, and to minimize the memory requirements.
Alternately, serving as a Wi-Fi adapter, wireless internet
access can be added to any microcontroller-based design with
simple connectivity through UART interface or the CPU
AHB bridge interface.

Fig.2 Image segmentation using Canny method [32]

E. LED Panel
The LED panel forms the visual part of the system which
will direct the traffic flow. The LEDs are controlled by the
micro-controller and based on its control action.
F. LCD Display
LCD display is used to display messages to the public in
case of emergencies or suggestions for alternate routes. It can
also be used to display the general traffic conditions of the
next signal so that the drivers can decide their routes wisely.
Fig.3 Image segmentation using Fuzzy Logic method [33]
G. Android Phone
Android phone is used to receive the traffic updates so that
we can select the best route for our faster travel. Block
programming is written using App-inventor through which
our android phone can measures the less congestion routes
and display them on our screen.
App-Inventor is a open source web application designed by
Google. It uses a graphical interface, very similar to Scratch
and the StarLogo TNG user interface, which allows users to
drag-and-drop visual objects to create an application that can
run on Android devices. In creating App Inventor, Google
drew upon significant prior research in educational
computing, as well as work done within Google on online
Fig.4 Block diagram of the system at unit signal 1and 2 development environments.
B. MAX232
MAX 232 is the name of IC which is used for TTL to
RS232 or RS232 to TTL conversion. To communicate

61
 ISSN: 2277-3754
ISO 9001:2008 Certified
International Journal of Engineering and Innovative Technology (IJEIT)
Volume 5, Issue 6, December 2015
IV. APPLICATION V. FUTURE SCOPE
A. Smart City The major application is the traffic management of a smart
A smart city uses information and communication city. Further the image processing could be replaced by video
technologies (ICT) to enhance quality, performance and processing for more accurate results. Data on cloud server
interactivity of urban services, to reduce costs and resource could be used for analysis of behavioral pattern followed at
consumption and to improve contact between citizens and different signals so help in reducing traffic congestion for
government. Sectors that have been developing smart city effectively and within stipulated time.
technology include government services, transport and traffic
management, energy, health care, water and waste. Smart city VI. CONCLUSION
applications are developed with the goal of improving the
management of urban flows and allowing for real time This paper has fairly tried to project the concept of dealing
responses to challenges. A smart city may therefore be more with traffic congestion in a smart city. It introduces us to the
prepared to respond to challenges than one with a simple vast world of image processing, which is used in several
„transactional‟ relationship with its citizens. Other terms that modern day applications. Endeavour has been made to touch
have been used for similar concepts include „cyberville‟, every field the device can be related to like designing,
„digital city‟, „electronic communities‟, „flexicity‟, programming involved, circuitry, construction, working and
„information city‟, „intelligent city‟, „knowledge-based city‟, application. It highly reduces the modern day problem of
„MESH city‟, „telecity‟, „teletopia‟, „Ubiquitous city‟, „wired traffic congestion that we are facing today.
city‟.
REFERENCES
B. Direction of Management of Traffic [1] D. Schultes, “Route Planning in Road Networks” Ph.D. Thesis,
In addition to the earlier method of traffic congestion University of Fridericiana, Karlsruhe, 2008.
detection, one more method can be used. A server can be [2] M. Appert, and C. Laurent. “Measuring urban road network
maintained which can receive certain crucial data calculated vulnerability using graph theory : the case of Montpellier's road
by the Controller of the signals. The main aim is to implement network”, Frederic Leone, Freddy Vinet, 2008,
a system that would trace the travel time of individual cars as
[3] N. Baruah and A. K. Baruah, “On a Traffic Control Problem
they pass the roadside controllers and compute an average trip using Cut-Set of Graph”, Int. J. Advanced Networking and
time using a rule-based system to decide whether the area is Applications, 3(4), pp. 1240-1244, 2012.
congested or uncongested. If congestion is sensed then system
would control traffic signals / generate automatic re-routing [4] R.T. van Katwijk et al., “Test bed for multiagent control
systems in road traffic management,” Transportation Research
messages to selected approaching vehicles.
Record, no. 1910, pp. 108–115, 2005.
C. Automatic Detection of Speed [5] M. M. Raza and R. R. Rao, “Agent Based Urban Traffic and
We can use this technique to calculate the speed of a Transportation Control with Cloud Computing”, International
motorist and to detect if he violates the prescribed/set speed Journal of Computer Trends and Technology, Vol. 3, Issue 1,
limit. If the motorist violates the rule, a warning message will pp. 120-123, 2012.
be sent to the motorist via audio and/or video interface and [6] S. Han and Y. Tabata “A Hybrid Genetic Algorithm for the
penalty will be calculated in the server and billed monthly to Vehicle Routing Problem with Controlling Lethal Gene”, Asia
the vehicle owner. Pacific Management Review, 7(3),pp. 405-426, 2002.

D. Automatic Billing [7] Meshkat and J. L. M. Vrancken, “Multi-Objective Road

Network Partitioning”, Procedia - Social and Behavioral
Automatic toll collection and automatic ―core area
Science, 2014.
charge‖ collections are also done using the same framework.
Controller unit will be placed at toll-booth and along the [8] Jiang et al., “An Agent Model with Adaptive Weight-based
motor able roads around the core area which will detect each Multi-objective Algorithm for Road-network Congestion
individual vehicle uniquely within its zone by capturing their Management”, International Journal of Computer and
Information Technology, 3(6), pp. 1188-1198, 2014.
device ids and will keep records of the time during which the
vehicle was seen by those Controllers within its reading zone. [9] H M Kammoun et al. “Adapt-Traf: an adaptive multiagent road
This information will be sent to a main server. Accordingly traffic management system based on hybrid ant-hierarchical
the main server will calculate the charges and raise bills fuzzy model”, Elsevier Transportation Research Part C Vol.
against the vehicle ids. 42, pp. 147-167, 2014.
[10] Vagia Kaltsa, “Swarm Intelligence for Detecting Interesting
E. Selection of Less Congestion Route
Events in Crowded Environments”, IEEE Transactions On
Through our android phone we can select the less Image Processing, Vol. 24, No. 7, July 2015.
congestion routes for our travels. This feature has been
[11] Joseph Mathew, P. M. Xavier2, “A Survey On Using Wireless
included in the above proposed system. Signals For Road Traffic Detection”, Volume: 03 Special
Issue: 01 | NC-WiCOMET-2014.
[12] Ali Ziya Alka, Mehmet Atif Karaca, “An Internet-Based
Interactive Embedded Data-Acquisition System for Real-Time

62
 ISSN: 2277-3754
ISO 9001:2008 Certified
International Journal of Engineering and Innovative Technology (IJEIT)
Volume 5, Issue 6, December 2015
Applications”, IEEE Transactions On Instrumentation And [26] P. G. Balaji and D. Srinivasan, “Type-2 fuzzy logic based
Measurement, Vol. 58, No. 3, March 2009. urban traffic management”, Eng. Appl. Artif. Intell. 24 (1), pp.
12–22, 2011.
[13] Phillip Walker, Saman Amirpour Amraii, Nilanjan
Chakraborty, Michael Lewis, and Katia Sycara, “Human [27] C. Bertelle et al., “Road traffic management based on ant
Control of Robot Swarms with Dynamic Leaders”, 2014 system and regulation model”, In: Proc. of the Int. Workshop
IEEE/RSJ International Conference on Intelligent Robots and on Modeling and Applied Simulation, pp. 35–43, 2003.
Systems (IROS 2014) September 14-18, 2014, Chicago, IL,
[28] Z. Yang, “A parallel ant colony algorithm for bus network
USA.
optimization”, Comput.-Aided Civil Infrastructure. Eng. 22
[14] Yuncan Xue, Haixia Zhang, and Juntao Fei, “Optimum Ship (1), pp. 44–55, 2007.
Pilot plan Model Based on the Modified Discrete Particle
[29] Y. Deng, “Dynamic shortest path in stochastic traffic networks
Swarm Optimization”, 2014 11th IEEE International
based on fluid neural network and particle swarm
Conference on Control & Automation (ICCA) June 18-20,
optimization”, In: Proc. of the 6th Int. Conf. on Natural
2014. Taichung, Taiwan.
Computation ICNC, IEEE, pp. 2325–2329, 2010.
[15] Matthias R. Brust, Bogdan M. Strimbu, “A Networked Swarm
[30] L. D‟Acierno, “An ant colony optimization algorithm for
Model for UAV Deployment in the Assessment of Forest
solving the asymmetric traffic assignment problem”. Eur. J.
Environments”, 2015 IEEE Tenth International Conference on
Oper. Res. 217 (2), pp. 459–469, 2012.
Intelligent Sensors, Sensor Networks and Information
Processing (ISSNIP) Singapore, 7-9 April 2015. [31] J. García-Nietoa, “Swarm intelligence for traffic light
scheduling: application to real urban areas”, Eng. Appl. Artif.
[16] Dipak K Dash, TNN May 31, 2012. ―India loses Rs 60,000
Intell. 25 (2), pp. 274–283, 2012.
crore due to traffic congestion: Study‖. Times Of India.
http://articles.timesofindia.indiatimes.com/2012-05-31/india/3 [32] www.matlab.com
1920307_1_toll-plazas-road-space-stoppage.
[33] www.wikipedia.com
[17] Bing-Fei Wu. Dept of Electr. And control eng, Nat. Chiao
Tung Univ, Hsinchu, Taiwan. A new Approach to Video-based
Traffic surveillance using fuzzy hybrid Information Inference AUTHOR’S PROFILE
Mechanism. March 2013.IEEE Intelligent Traffic Society.
Santwana Panda received her B.E under Electronics and
[18] Gomez, G. Diaz, and K. Boussetta. “ How Virtual Police Telecommunication from MIT in the year 2014. She is wowing towards her
Agents can help in the traffic guidance?” In Wireless
Communications and Networking Conference Workshops M.E degree under communication network from Indira College of
(WCNCW), 2012 IEEE, pages 360-364. IEEE, 2012. Engineering and Management. She attended National Level UG
[19] Arif A. Bookseller, Rupali R Jagtap, “Image processing based conference-2014 on Emerging Trends in E&TC engineering in Pune and
adaptive traffic control system” IOSR Journal of Electronics presented the paper entitled as “FPGA based design and implementation of
and Communication Engineering (IOSR-JECE), Pages: 33-37,
convolutional encoding and viterbi decoding techniques”.
2013
[20] L. Bieker and D. Krajzewicz. Evaluation of opening bus lanes
for private traffic triggered via V2X communication. In Anjali M. Patki, Prof & H.O.D. , E&TC Department, Indira College Of
Integrated and Sustainable Transportation System (FISTS), Engineering & Management has a teaching experience of more than 12 years
2011 IEEE Forum on, pages 48-53. IEEE, 2011.
and expert in the domain of communication engineering.
[21] HM Kammoun et al., “A Road Traffic Multi Agent Simulation
using Turtkit under Mdadkit”, Proc. of 9th International
Kedar Hushing founder and director of Axis Technologies- A high-end
Conference on Artificial Intelligence and Soft Computing, pp.
503-514, 2008. technology Solution Company has a working experience of more than 10

[22] M Ridwan, “Fuzzy preference based traffic assignment years in the field of embedded systems and microcontroller.
problem”, Elsevier Transportation Research Part C, Vol. 12
(3– 4), pp. 209–233, 2004.
[23] Y. E. Hawas, “Development and calibration of route choice
utility models: neuro-fuzzy approach”. J. Transport. Eng. 130
(2), 171–182, 2004.
[24] T. Arslan and C. J. Khisty, “A rational reasoning method from
fuzzy perceptions in route choice”, Fuzzy Sets Syst. 150 (3),
pp. 419–435, 2005.
[25] M. Ghatee, and S. M. Hashemi, “Traffic assignment model
with fuzzy level of travel demand: an efficient algorithm based
on quasi-Logit formulas”, Eur. J. Oper. Res. 194, pp. 432–451,
2009.

63