You are here

Scientists propose new method to manage traffic signals better

IISc scientists propose a new method to manage traffic lights, and this can potentially reduce your waiting times at traffic signals.

The new rule is simple: the lane with maximum vehicles will get the green signal first. Of course, you may have had this idea all along; but teaching the signal lights to do this on their own, without any human intervention, is a great technical challenge.

Throughout the world, the number of vehicles is increasing exponentially, while the space available for them, the width of the road, has remained more or less the same. According to the Road Transport Year Book (2007-2009), prepared by the Ministry of Road Transport and Highways, Government of India, about 96 lakh vehicles were added to the Indian streets between 2008 and 2009. A direct impact of this is that the time that a person would wish to spend with his family or in the office is now being frittered away at the chaos on the road.

Prabuchandran, Hemanth Kumar and Prof Shalabh Bhatnagar from Indian Institute of Science, Bangalore have been working on the smart ways to manage traffic signals better. Their idea aims to intelligently control the traffic lights at junctions for minimizing the average delay experienced by the road users. The objective is to pick the most congested line, give it the green signal first to clear the overall congestion quicker.

Broadly, there are two approaches to manage a busy junction. In one approach, the green signal is given in an order, but the signal duration is varied. For example, green signal may be given in clockwise manner, but lanes with more vehicles remain green for a longer time. In the other approach, the green signal duration is fixed, but the order in which the signal is given is varied depending on traffic volume. For example, the lane with more vehicles will get a green signal earlier than a less crowded lane; but green lights up for a fixed number of seconds. IISc researchers have taken the second approach.

The traffic congestion level across various lanes is available from sensors. Instead of controlling the complete road network as a single entity, each traffic junction is considered as a point of decision-making. By doing so, this approach can be used for any road network irrespective of how big the city is. At the start, the initial order at a junction is set randomly. At every junction, the priority of the lanes is determined based on the traffic congestion level information on the lanes on its junction and its neighbors. Based on this, the high traffic lanes are given more opportunities to stay green while the low traffic lanes are skipped from getting a green signal. This way the various junctions organize themselves by talking to their neighbors, thus clearing the congestion quicker. Smarter the traffic signals, happier are the commuters.

Speaking of the future, the authors said that there is plan to look at more attributes like exogenous events, driver behavior, etc.

About the authors:

Shalabh Bhatnagar is a Professor in the Department of Computer Science and Automation, Indian Institute of Science, Bangalore. Prabuchandran is a senior PhD student in the same department. Hemanth Kumar was a masters student in the department and now in industry.


Link to abstract: