KNOWLEDGE-BASED CONTROL OF GRASPING IN ROBOT HANDS USING HEURISTICS FROM HUMAN MOTOR-SKILLS

The development of a grasp planner for multifingered robot hands is de scribed. The planner is knowledge-based, selecting grasp postures by r easoning from symbolic information on target object geometry and the n ature of the task. The ability of the planner to utilize task informat ion is based on an attempt to mimic human grasping behavior. Several t ask attributes and a set of heuristics derived from observation of hum an motor skills are included in the system. The paper gives several ex amples of the reasoning of the system in selecting the appropriate gra sp mode for spherical and cylindrical objects for different tasks.