The scourge of careless drivers who run red lights may soon be a thing of the past, because the red light—a 150-year-old technology designed for horse carriages—may soon be a thing of the past. In a new paper, researchers at MIT are proposing what they call “Slot-based Intersections” (SIs) that smart autonomous vehicles can safely traverse using slot-based systems commonly used for aerial traffic. They note that such intersections have been simulated before; their advance is to compare performance with the performance of traffic-light-based intersections. SIs, they say, can double intersection capacity and significantly reduce delays.
They cite one intuitive observation: a traffic-light cycle is divided into phases in which certain vehicles are granted access to the intersection, which operates suboptimally during setup (amber light) phases of five to eight seconds. The SI approach reduces the setup phase by a factor of about five.
Kevin Hartnett in the Boston Globe quotes Carlo Ratti, director of the MIT Senseable City Lab and one of the paper’s authors, as saying, “When sensor-laden vehicles approach an intersection, they can communicate their presence and remain at a safe distance from each other, rather than grinding to a halt at traffic lights. By removing the waits caused by traffic lights, slot-based intersections create a system that is much more efficient.”
Optimal results may ultimately require more than car-to-car, or V2V, communication. Hartnett quotes Thomas Van Woensel, a professor of freight transport and logistics at Eindhoven University of Technology in the Netherlands, as saying, “You’ll need centralized decision-making that organizes flows and intersections. Car-to-car sharing of information is extremely local, while I think global information about the network itself is also extremely important.”
You can see a simulation of an SI and traffic-light intersection here.