Sensory-Motor Fusion Architecture Based on High-Speed Sensory Feedback and Its Application to Grasping and Manipulation

In this paper we propose a new hierarchical parallel processing model based on high-speed realtime sensory feedback to realize manipulation in a dynamically changing environment. In this model high-speed sensory feedback is used at hierarchical three stages: (1) servo control, (2) motion planning, and (3) adaptation of the motion planner and the motion controller. Because the structure of the motion planner and the motion controller is changed according to an external environment at a high rate, responsiveness and flexibility to dynamic changes of the environment is achieved. And grasping and handling of a dynamically moving object is achieved based on this model. Experimental results are shown with a dextrous hand-arm and a high-speed active vision.