Potential Based Diffusion Motion Planning

TL;DR — You present a way to parameterize potential based motion planning with diffusion models and illustrate is compositionality.

Abstract

Effective motion planning in high dimensional spaces is a long-standing open problem in robotics. One class of traditional motion planning algorithms corresponds to potential-based motion planning. An advantage of potential based motion planning is composability -- different motion constraints can easily combined by adding corresponding potentials. However, constructing motion paths from potentials requires solving a global optimization across configuration space potential landscape, which is often prone to local minima, causing these approaches to fall out of favor in recent years. We propose a new approach towards learning potential based motion planning, where we train a neural networks to capture and learn an easily optimizable potentials over motion planning trajectories. We illustrate the effectiveness of such approach, significantly outperforming both classical and recent learned motion planning approaches, and illustrate its inherent composability, enabling us to generalize to a multitude of different motion constraints.