🤖 AI Summary
Collaborative robots face safety-critical obstacle avoidance challenges in dynamic environments due to tool changes and partial observability. Method: This paper proposes a tool-aware, real-time adaptive collision avoidance framework integrating a learning-based point cloud perception model—featuring tool-aware filtering and occlusion-aware collision prediction—with a constrained reinforcement learning controller, enabling millisecond-level response and online decision-making under dynamic task switching. Contribution/Results: It is the first approach to explicitly embed tool geometry and interaction patterns into the perception–control closed loop, achieving sub-millimeter-accurate safe navigation under partial observability. Experiments demonstrate a 60% reduction in computational overhead compared to Artificial Potential Field (APF) and Model Predictive Path Integral (MPPI) methods. The framework has been successfully deployed on a real collaborative robot system, demonstrating strong safety guarantees, motion smoothness, and modular extensibility.
📝 Abstract
Ensuring safe and efficient operation of collaborative robots in human environments is challenging, especially in dynamic settings where both obstacle motion and tasks change over time. Current robot controllers typically assume full visibility and fixed tools, which can lead to collisions or overly conservative behavior. In our work, we introduce a tool-aware collision avoidance system that adjusts in real time to different tool sizes and modes of tool-environment interaction. Using a learned perception model, our system filters out robot and tool components from the point cloud, reasons about occluded area, and predicts collision under partial observability. We then use a control policy trained via constrained reinforcement learning to produce smooth avoidance maneuvers in under 10 milliseconds. In simulated and real-world tests, our approach outperforms traditional approaches (APF, MPPI) in dynamic environments, while maintaining sub-millimeter accuracy. Moreover, our system operates with approximately 60% lower computational cost compared to a state-of-the-art GPU-based planner. Our approach provides modular, efficient, and effective collision avoidance for robots operating in dynamic environments. We integrate our method into a collaborative robot application and demonstrate its practical use for safe and responsive operation.