By: Katie Byl
Robot manipulation tasks for a mobile, personal robot will often be importantly distinct from those of a traditional, factory robot arm; correspondingly, appropriate motion planning solutions may be notably different, as well. This paper introduces novel definitions of and solution methods for optimal kinodynamic planning for mobile robot manipulation applications.
In particular, we consider a generalized, real-world scenario where (1) robots must interact safely with humans and with objects in home environments, (2) there may be significant uncertainty about the impedance properties of objects to be manipulated, and (3) there may also be significant environmental and sensory noise. Unlike traditional kinodynamic planning, where minimum time trajectories are generally considered optimal, we suggest that the paramount goal in planning for personal robots should instead be safety; that is, we should minimize the probability of “failure” or at least ensure it is below some threshold. We discuss the appropriateness of this definition of optimality, suggest a generalized methodology for achieving optimal or near-optimal solutions, and present a simple, 2D manipulation problem to demonstrate the approach.