On Performability Theory and the Inverse Sliding Problem

Abstract

For a given robot,we would like to know its capabilities and limitations. The question of what tasks a robot is capable of executing has been asked with increasing frequency as researchers explore the central issues of robotics. I shall refer to the theory that would answer this question as Performability Theory . One approach to Performability Theory is to examine the mechanics of the robot and the task. However, a complete model of the mechanics of a real robot and its environment quickly becomes intractable. The Inverse Sliding Problem involves simple mechanics; it deals with a planar world in which objects are free to translate and rotate but are subject to the forces of Coulomb friction. The problem is to determine the initial translational and rotational velocities required for an object to slide to a given final configuration. In this paper, I present a solution to a portion of the Inverse Sliding Problem, which I propose as the basis for a physically realistic yet simple world in which we can start to formulate the basic ideas of Performability Theory.