BIOLOGICALLY INSPIRED ROBOT CONTROL

Abstract Limitations of conventional control methods for robots operating in unstructured environments are discussed, and contrasted with control techniques used by living systems. Three specific examples of robot systems designed with inspiration from biology are discussed: (1) multifingered robot hands with shape-adaptive grasping, (2) a walking machine which uses genetic algorithms to evolve a gait pattern, and (3) a behavior-based control system for a robot flying machine.