Behavior Tree Based Robotic Skill Execution for Human Robot Collaboration in Industrial Settings

This paper extensively explores the utilization of behavior tree-based robotic skill execution engines, focusing specifically on their application in industrial settings. By integrating behavior trees into the robotic framework, this research significantly contributes to enhancing the adaptability of robots in dynamic environments. The modularity and reactivity offered by behavior trees play a pivotal role in enabling robots to dynamically adjust their behaviors in response to unforeseen circumstances, especially in the context of extensive human-robot collaboration in industrial scenarios. The demonstrated application of this approach in a real-world assembly scenario utilizes a novel mobile collaborative manipulator with reduced computational power. This real-world implementation not only underscores the practical relevance of behavior tree-based execution engines but also highlights their applicability in human robot collaborative handling of objects. The behavior tree ticking mechanism, even on a less powerful PC, showcases the robustness of the proposed methodology. This work offers insights into the modularity and reactivity inherent in behavior trees, providing a promising avenue for addressing challenges in human-robot collaboration within industrial settings, even when operating under computational constraints.