Robust Traffic Merging Strategies for Sensor-Enabled Cars Using Time Geography

Abstract

We present two novel merging algorithms that optimize traffic flow on highways, particularly at intersections of ramps and main roads. In our work, cars are equipped with sensors that can detect distance to neighboring cars, and communicate their velocity and acceleration readings with one another. Sensor-enabled cars can locally exchange sensed information about traffic and adapt their behavior much earlier than regular cars. However, the accuracy level of sensors is a major challenge for merging algorithms, because inaccuracies can potentially lead to unsafe merging behaviors. In this paper, we investigate how the accuracy of sensors impacts merging algorithms, and design robust merging algorithms that tolerate sensor errors. Experimental results show that our main proposed merging algorithm, which is based on concepts from time geography, is able to guarantee safe merging while tolerating four times more imprecise positioning information, and can double the road capacity and increase the traffic flow by 25%.