Machine learning for biped robot locomotion

Abstract

This paper presents an overview of the state of the art methods in machine learning and the potential of application for biped robot locomotion. Adopting a bottom-up structure, the paper starts by introducing the state of machine learning, the fundamental questionsit addresses and the current research topics. Second, machine learning methods that have been successfully applied in robotic systems are briefly discussed. Finally, the paper focuses on problems of biped locomotion that can be solved by using learning (e.g., reinforcement learning). Case-studies highlighting the implemented solutions help to understand how a locomotion robot system can improveits control strategy through experience.