New Approach to Improve Traffic Light Timings
Nobody enjoys sitting at red lights, waiting for the signal to change, but finding the optimal timing is difficult and complex. Models do exist that can achieve amazing resolution, but these come at the cost of efficiency and sometimes larger accuracy. Researchers at MIT though are proposing a new method that could optimize timings better than these models, cutting down on wait times significantly.
Typically cities will time their traffic lights by focusing on the primary arteries and just optimizing the times along those routes. This approach has the limitation of not considering the ripple effects on other roads, such as from drivers taking an alternate route, to avoid lights. As these models will consider individual driver behavior, it becomes prohibitively complex to grow them to consider all of these effects. The MIT researchers however found an efficient way to arrive at the optimal timings while avoiding the complexity. It starts with the high resolution technique to propose timings and then uses a lower resolution model to look at traffic flow.
To test this new approach, the researchers considered the traffic of Lausanne, Switzerland. The timings it produced resulted in a 22% reduction in average wait time for the city, according to simulations. Next the researchers want to expand the model to be able to adapt to changing traffic conditions.