Toward a deeper understanding of motion alternatives via an equivalence relation on local paths

Abstract

Many problems in robot motion planning involve collision testing a set of local paths. In this paper we propose a novel solution to this problem by exploiting the structure of paths and the outcome of previous collision tests. Our approach circumvents expensive collision tests on a given path by detecting that the entire geometry of the path has effectively already been tested on a combination of other paths. We define a homotopy-like equivalence relation among local paths to detect this condition, and we provide algorithms that (1) classify paths based on equivalence, and (2) circumvent collision testing on up to 90% of them. We then prove both correctness and completeness of these algorithms and provide experimental results demonstrating a performance increase up to 300% in the rate of path tests. Additionally, we apply our equivalence relation to the navigation problem in a planning algorithm that takes advantage of information gained from equivalence relationships among collision-free paths. Finally, we explore applications of path equivalence to several other mechanisms, including kinematic chains and medical steerable needles.