GRASPER: A Permissive Planning Robot

Execut ion of classical plans in the real world can be problematic. Small discrepancies between a planner's internal representations and the real world are unavoidable. These can conspire to cause real-world fai lure even though the planner is sound and, therefore, "proves" that a sequence of actions achieves the desired goal. The diff iculty, of course, is that the planner's proof is contingent on its internal world model precisely captur ing all relevant features of the external real wor ld . This is seldom the case, par t icu lar ly in robotics where uncertainties abound. Small but unavoidable sensor errors preclude accurate knowledge of the state of the world Worse, the planner's own characterization of the effects of its actions are themselves only approximations. Realworld execution of a sequence of actions can introduce and quickly magnify inconsistencies w i th the internal mi cro wor ld.