Crossroads: Autonomous Intersection Management

Researchers: Edward Andert and Mohammad Khayatian

For autonomous vehicles, intelligent autonomous intersection management will be required for safe and efficient operation of the intersection. In order to achieve safe operation despite uncertainty in vehicle’s trajectory, intersection management techniques must consider a safety buffer around the vehicles. For truly safe operation, the safety buffer should also account for the network and computational delay caused by communication with the IM (Intersection Manager) to request entry and receive a reply. However, modeling the worst-case computation and network delay as additional safety buffer degrades the throughput of the intersection. To avoid this problem, AIM a popular state-of-the-art IM adopts a query-based approach, in which the IM only provides a yes/no answer to vehicle’s queries of speed and arrival time. Although this solution does not degrade the position uncertainty, but it increases the network traffic, the amount of computation on both the car and the IM, and ultimately results in poor intersection throughput. We present Crossroads – a time-sensitive programming method to program the interface of a vehicle and IM, without requiring additional buffer to account for the effect of network and computational delay, and enables efficient intersection management. Our results on 1/10 scale model of intersection using TRAXXAS RC cars demonstrate that time-sensitive programming based approach Crossroads obviates the need for large buffers to accommodate for the network and computation delay, and can reduce the average wait time for the vehicles at a single-lane intersection by 24%. To compare Crossroads with previous approaches, we perform extensive Matlab simulations, and find that Crossroads achieves on average 1.62X higher throughput than simple VT-IM with extra safety buffers, and 1.36X better than AIM.

Link to Edward’s thesis.