Skill-based instruction of collaborative robots in industrial settings

During the past decades, the increasing need for more flexible and agile manufacturing equipment has spawned a growing interest in collaborative robots. Contrary to traditional industrial robots, collaborative robots are intended for operating alongside the production personnel in dynamic or semi-structured human environments. To cope with the environment and workflow of humans, new programming and control methods are needed compared to those of traditional industrial robots. This paper presents a task-level programming software tool allowing robotic novices to program industrial tasks on a collaborative robot. The tool called Skill Based System (SBS) is founded on the concept of robot skills, which are parameterizable and task-related actions of the robot. Task programming is conducted by first sequencing skills followed by an online parameterization performed using kinesthetic teaching. Through several user studies, SBS is found to enable robotic novices to program industrial tasks. SBS has further been deployed and tested in two manufacturing settings demonstrating its applicability in real industrial scenarios.