Towards a Behavior Tree Based Robotic Skill Execution Framework for Human Robot Collaboration in Industrial Assembly

This paper delves into the application of behavior tree-based robotic skill execution engines, with a particular emphasis on their utilization in industrial contexts. The study presents a thorough exploration of a behavior tree-based robot skill execution framework, highlighting its implementation in the execution of various robotic skills within a real-world industrial pilot line. The novelty of this framework is evident in its successful execution of robot skills on a unique mobile manipulator featuring an 8-degree-of-freedom robotic arm. Notably, the framework operates efficiently on the limited onboard hardware of the robot. This research addresses the practical challenges encountered in the industrial assembly process, effectively showcasing the versatility of the behavior tree-based skill execution framework. The capabilities of the framework extend beyond the execution of standalone robotic skills to encompass the seamless execution of collaborative actions in a human-robot partnership. The study provides valuable insights into the inherent modularity and reactivity of behavior trees, presenting a promising approach to address challenges in human-robot collaboration within industrial settings, even when confronted with computational constraints. This is demonstrated by applying the behavior tree-based task execution framework to the novel mobile manipulator in a real-world pilot line scenario involving car door assembly.