July 7, 2014
MIT System Puts Citywide Traffic Lights In Unison
Brett Smith for redOrbit.com - Your Universe Online
Traffic can be the bane of a modern urban lifestyle and cities have used computer algorithms to time a single traffic light or a collection of lights in the most efficient way possible.
Now, a team of researchers from MIT has devised a system that times traffic on a citywide scale, according to an upcoming report in the journal Transportation Science.
“Usually in practice, when you want to time traffic lights, traditionally it’s been done in a local way,” said study author Carolina Osorio, an assistant professor of civil and environmental engineering at MIT. “You define one intersection, or maybe a set of intersections along an arterial, and you fine-tune or optimize the traffic lights there.”
The new system was based on a traffic study of Lausanne, Switzerland – a city of over 120,000 on the shores of Lake Geneva.
“What is less done, and is more difficult to do, is when you look at a broader scale, in this case the city of Lausanne, and you want to change signal times at intersections distributed across the entire city, with the objective of trying to improve conditions across the entire city,” Osorio added.
A citywide system is more difficult because a shift at one intersection can create ripples across the surrounding region, or variations in driver behavior after changes in traffic-light patterns. For instance, if wait times on a selected route rise, drivers may seek other ways known to have fewer red lights.
The new system can time lights in large city areas whilst accounting for the complex and diverse responses of specific drivers. The team's approach utilizes high-resolution simulators that describe the detailed behavior of drivers in reaction to variations in travel conditions.
In comprehensive simulations of Lausanne’s traffic, the team discovered that the timings generated by their approach decreased the average travel time for commuters by 22 percent, as opposed to timings generated by commercial traffic-light timing methods.
Some cities are making use of similar high-quality simulators, but the complexity of these models means they are computationally rigorous. The new MIT method allows these models to be utilized in a useful and computationally efficient way. While other citywide models set up proposed timings, they treat traffic flow simplistically and homogeneously – as opposed to a collection of individual travelers with distinct and complex behavior.
The new system combined a detailed vehicle-level analysis and a city-scale optimization.
“What if we combine information from these microscopic simulations with [citywide] information from these simple traffic models that are very computationally efficient and run instantly, but have very low resolution?” Osorio asked.
She added that the new system could not only be used to minimize commuting times, but also to minimize fuel consumption. It could also inform determining locations for services such as car-sharing hubs.
Mohamad Talas, a deputy director of system engineering for New York City’s Department of Transportation, said his department is currently working with the MIT researchers to test the new method.
“Such a model can validate our active traffic-management system in Manhattan, and allow us to fine-tune our processes and improve the network operation,” Talas said. “I believe that this approach is economically viable, with cost savings for any jurisdiction that needs to assess and improve traffic conditions for a large area of the transportation network.”