Date of Original Version
Abstract or Table of Contents
The widespread success of sampling-based plan- ning algorithms stems from their ability to rapidly discover the connectivity of a conﬁguration space. Past research has found that non-uniform sampling in the conﬁguration space can signiﬁcantly outperform uniform sampling; one important strategy is to bias the sampling distribution based on features present in the underlying workspace. In this paper, we unite several previous approaches to workspace biasing into a gen- eral framework for automatically discovering useful sampling distributions. We present a novel algorithm, based on the RE I NF ORCE family of stochastic policy gradient algorithms, which automatically discovers a locally-optimal weighting of workspace features to produce a distribution which performs well for a given class of sampling-based motion planning queries. We present as well a novel set of workspace features that our adaptive algorithm can leverage for improved conﬁguration space sampling. Experimental results show our algorithm to be effective across a variety of robotic platforms and high- dimensional conﬁguration spaces.