Model-Based Robot Control for Human-Robot Flexible Material Co-Manipulation

Abstract

Despite market importance and growth, manufacturing systems involving flexible materials, textiles and composites remain manual. Challenges related to flexible material deformation highlight limitations of robot cognition during fabric handling. This manuscript presents a model-based closed-loop control framework for seamless human-robot or multi-robot fabric co-manipulation. A mass-spring model is used for simulating ply distortion and generating optimal grasping points' spatial localization. The model is enhanced with real-time operator's handling actions, as captured from the implemented perception system. The proposed sensor and model-based controlling framework incorporates robot motion planners either for operator support, through non-rigid object co-manipulation, or synchronization of cooperative robots within fully automated tasks. An experimental setup is used for validating system's handling cognition during collaborative manipulation.