Traffic waves are phenomena that emerge when the vehicular density exceeds a critical threshold. This research demonstrates, experimentally, that intelligent control of an autonomous vehicle can dampen stop-and-go waves that can arise even in the absence of geometric or lane changing triggers.
Metrics for velocity, braking events, and fuel economy were compared across three experiments involving more than 20 vehicles on a circular track, one of which was a University of Arizona self-driving capable Cognitive and Autonomous Test (CAT) vehicle. The CAT vehicle was able to transition between manual velocity control and automated velocity control. Each experiment lasted between 7 and 10 minutes. The CAT vehicle began each experiment in manual mode and was switched into automated velocity control mode during the experiment. Traffic waves occurred in all experiments and the unsteady traffic condition was allowed to persist for at least 45 second before the controller was activated.
Experimental findings suggested stop-and-go freeway traffic flow can be controlled to improve fuel economy and throughput with as few as 5 percent autonomous vehicles in the traffic stream.
In each experiment, stop-and-go waves arose dynamically when all vehicles were under human control. Once one vehicle was activated to be autonomous, the traffic waves were dissipated.
Compared to when waves were present, automated control resulted in a throughput increase of up to 15 percent.