Posterior Sampling for Anytime Motion Planning on Graphs with Expensive-to-Evaluate Edges
This addresses the need for real-time motion planning in robotics by providing strong anytime performance, though it is incremental as it builds on existing lazy planning methods.
The paper tackles the problem of motion planning with expensive collision checks by introducing PSMP, an anytime algorithm that quickly finds an initial feasible path and progressively yields shorter ones, achieving an expected regret bound of $ ilde{O}(\sqrt{\mathcal{S} \mathcal{A} T})$ and outperforming baselines in 2D and 7D planning tasks.
Collision checking is a computational bottleneck in motion planning, requiring lazy algorithms that explicitly reason about when to perform this computation. Optimism in the face of collision uncertainty minimizes the number of checks before finding the shortest path. However, this may take a prohibitively long time to compute, with no other feasible paths discovered during this period. For many real-time applications, we instead demand strong anytime performance, defined as minimizing the cumulative lengths of the feasible paths yielded over time. We introduce Posterior Sampling for Motion Planning (PSMP), an anytime lazy motion planning algorithm that leverages learned posteriors on edge collisions to quickly discover an initial feasible path and progressively yield shorter paths. PSMP obtains an expected regret bound of $\tilde{O}(\sqrt{\mathcal{S} \mathcal{A} T})$ and outperforms comparative baselines on a set of 2D and 7D planning problems.