Synthesis of provably correct controllers for autonomous vehicles in urban environments

Abstract

This paper considers automatic synthesis of provably correct controllers for autonomous vehicles operating in an urban environment populated with static obstacles and live traffic. We express traffic rules such as collision avoidance, vehicle separation, speed limit, lane following, passing, merging and intersection precedence requirements in a formal specification language. Embedded control software synthesis is then applied to generate a controller that ensures that the vehicle obeys this set of traffic rules in any road and traffic conditions that satisfy certain assumptions.