Real-Time Skill Providing System Using Human Haptic Information

Since the advance of the aging society with a falling birthrate, technical inheritance has been a serious problem. This paper proposes a method to provide a skillful motion. First, haptic data (i.e., position and force information) is acquired using a bilateral control system. Second, the acquired haptic data is divided into action components by using modal transformation. Third, position and force information in each action component are compared with recorded data concerning the skillful motion. Finally, if an operator performs motion similar to skillful motion, the recorded data is loaded to the bilateral control system. Similarity between the performed motion and the skillful motion is calculated based on dynamic programming matching algorithms. The validity of the proposed method was experimentally verified. In the experiments, the proposed method was applied for grasping motion, and the recorded skillful motion was reproduced when the operator tried to perform the skillful motion. The maximum position error was 0.072 cm.