Improving intersection safety through variable speed limits
Research on connected and automated vehicle (CAV) technology is an emerging field with a variety of applications. Researchers from the University of Minnesota are combining CAVs with another emerging technology—variable speed limits (VSLs)—to improve driving safety and efficiency at intersections.
In a project sponsored by the Roadway Safety Institute, the researchers aimed to model how CAVs behave at intersections if they are told to obey VSLs—which can be changed from minute to minute as traffic conditions fluctuate.
“There are a lot of vehicle interactions around intersections,” says Michael Levin, assistant professor in the U’s Department of Civil, Environmental, and Geo- Engineering and the project’s principal investigator. “Intersections force vehicles to come to a stop, [and] there are vehicles moving in conflicting paths that inherently cause safety issues,” Levin says.
With CAVs and VSLs, however, it might be possible to mitigate some of those safety issues. Levin, together with research assistant Rongsheng Chen and senior research associate Chen-Fu Liao from the Department of Mechanical Engineering, have been running computer simulations to determine if VSLs could be used to reduce the amount of deceleration or acceleration of cars approaching a stoplight. If a CAV knows that an upcoming light is about to turn green, for example, it might slow down slightly to ensure that it never has to come to a complete stop before the light turns. If there’s a crash on the road, CAVs could be warned to slow down more gradually, ahead of time, so that no vehicles have to brake suddenly. Not only is this safer, but it’s also more fuel-efficient.
CAVs, however, are not widespread and probably won’t be for a long time. VSLs also change so frequently (often every 200 to 500 feet) and by such small amounts that it’s unreasonable to expect human drivers to follow them exactly. To account for this, the computer models assumed that a small number of CAVs would be mixed in with a larger number of human-driven vehicles. What the models have shown so far is that even a small number of CAVs could have a large impact on traffic flow; a CAV that is following the speed limit creates a barrier for speeding human drivers. This forms a moving “bottleneck” and slows the overall flow of traffic. The idea is based on the standard kinematic wave model of traffic flow, which assumes that traffic moves like water—fast when the banks are wide, slower when it is constricted.
The research is currently preliminary; since there’s no roadside infrastructure in place to broadcast VSLs, the models cannot yet be implemented directly. However, they give an idea of how effective VSLs could be and how best to use them.
“Ultimately,” Levin says, “I hope that state, county, or city departments of transportation could implement this on their arterial roads.”
So far, the research team has finished creating the model of traffic behavior, and from that model they’ve determined what speed limits should be used to maximize fuel efficiency. Next, the team will analyze how VSLs affect safety on the road.