Path planning of grippers modeled after human grasping motions

This paper describes path planning of a robot grippers emulating human grasping motions with respect to the thumb and index finger. When the human transports the hand to the target object, the velocity patterns of the tips are bell-shaped regardless of the initial hand position. Path planning for the two tips of digits is similar to the obstacle avoidance against the target object, and opening or closing between the two tips tend to be conducted by moving the index finger. Utilizing the obtained human characteristics, the path planning model is constructed in the complex potential field. The results of simulations show that the two tips separately go round the target object so as to approach the grasped part due to the artificial potential. Near the target object, the tips move slowly just like a human since their motions are based on the bell-shaped velocity pattern