Dependable grasping strategy for service robots using Fine Approaching Positions and adaptive hands

Grasping strategies to grasp various shapes and postures of objects for service robots are presented in this paper. Because the performance of algorithms for the grasp problem is closely connected with the structure and functions of hands, both of these should be considered simultaneously. It is our procedure that searching the optimal approaching position using objects and shapes database in the first place, and then grasping the real objects using adaptive hands. The optimal FAPs (Fine Approaching Positions) generated by solving the problems of collision and variety of postures are verified by simulation.