Adaptive routing of autonomous vehicles using neighborhood traffic

As research on autonomous vehicles increases, automotive manufacturers and researchers are developing coordination techniques to enable safe passage of vehicles through intersections. These techniques are called Autonomous Intersection Management (AIM). Even though AIM techniques improve intersection throughput, they do not effectively reduce congestion. Real life urban roads comprise of a networks of multiple intersections. In such scenarios, communicating traffic information between intersections is essential for reducing congestion on the roads. To achieve this, we propose an adaptive routing algorithm that incorporates a fusion of vehicle-to-intersection (V2I) communication and intersection-to-intersection (I2I) communication in order to bring about significant reductions in congestion. To implement this algorithm, we constructed the Enhanced AIM simulation framework as an extension of AIM simulator (University of Texas, Austin). We demonstrate with simulation experiments that our proposed routing algorithm shows reduced congestion and wait-time, and improved user experience.