Overcoming Challenges in Programming Human-Robot Collaboration in Flexible Manufacturing

Abstract

Advanced collaborative robots (cobots) are revolutionizing flexible manufacturing. But designing and programming the control software for human-robot collaboration can be challenging. This paper identifies difficulties in existing approaches and tries to find pragmatic solutions to these challenges. A central issue is the extensive domain knowledge that is required, especially when working with the most advanced systems. A multidisciplinary team may have difficulties working with such a system. This research aims to reduce the need for extensive domain knowledge, streamline system development, and thereby make the implementation of intelligent cobots more affordable for high-mix low-volume applications. As an example, we introduce our work-in-progress approach XRob2, and discuss how a skill-based, distributed and human-centric architecture can simplify the integration of intelligent collaborative robots into industrial environments.