Decoupled MPPI-Based Multi-Arm Motion Planning

TL;DR

This paper introduces MR-STORM, a distributed multi-robot motion planning algorithm that extends STORM to efficiently handle multiple high-DOF arms with dynamic obstacles, showing empirical improvements over state-of-the-art methods.

Contribution

We extend the STORM algorithm to a multi-robot setting with dynamic obstacle handling and a dynamic priority scheme, enabling scalable and efficient multi-arm motion planning.

Findings

MR-STORM outperforms SOTA algorithms in dynamic environments.

The distributed approach improves scalability for multiple arms.

Handling dynamic obstacles enhances real-world applicability.

Abstract

Recent advances in sampling-based motion planning algorithms for high DOF arms leverage GPUs to provide SOTA performance. These algorithms can be used to control multiple arms jointly, but this approach scales poorly. To address this, we extend STORM, a sampling-based model-predictive-control (MPC) motion planning algorithm, to handle multiple robots in a distributed fashion. First, we modify STORM to handle dynamic obstacles. Then, we let each arm compute its own motion plan prefix, which it shares with the other arms, which treat it as a dynamic obstacle. Finally, we add a dynamic priority scheme. The new algorithm, MR-STORM, demonstrates clear empirical advantages over SOTA algorithms when operating with both static and dynamic obstacles.