Fostering Robust Human-Robot Collaboration through AI Task Planning

Abstract

Recent advances in Artificial Intelligence (AI) are facilitating the deployment of intelligent systems in manufacturing. In Human-Robot Collaboration (HRC), industrial robots offer accuracy and efficiency while humans guarantee both experience and specialized and not replaceable skills. The seamless coordination of such different abilities constitutes one of the current challenges. This paper presents a dynamic task sequencing system for robust HRC developed within a EU-funded project. The proposed solution uses AI techniques to deal with the temporal variance entailed by the active presence of humans as well as to dynamically adapt task plans according to actual behavior of the pair human-worker/robot. The tool has been deployed in a real pilot plant.