Enabling Safe Autonomous Driving in Real-World City Traffic Using Multiple Criteria Decision Making

Abstract

This paper addresses the topic of real-time decision making for autonomous city vehicles, i.e. the autonomous vehicles' ability to make appropriate driving decisions in city road traffic situations. The paper explains the overall controls system architecture, the decision making task decomposition, and focuses on how Multiple Criteria Decision Making (MCDM) is used in the process of selecting the most appropriate driving maneuver from the set of feasible ones. Experimental tests show that MCDM is suitable for this new application area.