A study on al-based approaches for high-level decision making in highway autonomous driving

• Published in: 2017 IEEE International Conference on Systems, Man, and Cybernetics (SMC) pp. 3671 - 3676
• Main Authors: Claussmann, Laurene, Revilloud, Marc, Glaser, Sebastien, Gruyer, Dominique
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-10-2017
• Subjects: Approximation algorithms; artificial intelligence; autonomous vehicle technology; Autonomous vehicles; Cognition; Decision making; intelligent transportation systems; Planning; Road transportation
• DOI: 10.1109/SMC.2017.8123203

Autonomous driving relies on a wide range of domains of research. It faces rapid technological and theoretical advances, with various methods and process developments. For the interest of the high-level decision making subpart of autonomous vehicle architecture, the previous states of the art report a vast literature, from traditional mobile robotics to human-modelling approaches. The purpose of this paper is to survey the current and major algorithms in the specific field of artificial intelligence for autonomous vehicles. Such systems are particularly suited for high-level decision making since they must, by definition, be able to perceive and react to their environment in order to reach given objectives. The scope is reduced to highway driving applications, considering individual, collective, and cooperative decisions. Strengths and limitations of the reviewed methods are compared, with respect to the structure and constraints of the studied driving situations. Open questions are proposed as a reflection towards the next generation of decision-makers for autonomous vehicles.