A mathematical framework for cooperative collision avoidance of human-driven vehicles at intersections

Active safety systems for collision avoidance of road vehicles are a promising solution to the enormous human and monetary cost of road crashes. In this paper, we describe some new mathematical approaches for the construction of collision avoidance algorithms for human-driven vehicles. These algorithms exploit the sensing and communication capabilities of the vehicles to share information on the current state of the system, and based on this information check the safety of the human decisions and correct them when they would lead to a collision.