Combining Planning and Diffusion for Mobility with Unknown Dynamics

TL;DR

This paper introduces a hierarchical approach combining planning and diffusion-based policies for long-horizon mobile manipulation tasks with unknown dynamics, demonstrating improved success and generalization on a robot pushing chairs.

Contribution

It proposes a hierarchical algorithm that decomposes complex manipulation into high-level planning and low-level diffusion policies, enabling better performance and generalization in unknown dynamic environments.

Findings

Hierarchical policy outperforms baseline methods on success rate.

Policy generalizes to new environments without retraining.

Diffusion-based short-horizon policies are effective for unknown dynamics.

Abstract

Manipulation of large objects over long horizons (such as carts in a warehouse) is an essential skill for deployable robotic systems. Large objects require mobile manipulation which involves simultaneous manipulation, navigation, and movement with the object in tow. In many real-world situations, object dynamics are incredibly complex, such as the interaction of an office chair (with a rotating base and five caster wheels) and the ground. We present a hierarchical algorithm for long-horizon robot manipulation problems in which the dynamics are partially unknown. We observe that diffusion-based behavior cloning is highly effective for short-horizon problems with unknown dynamics, so we decompose the problem into an abstract high-level, obstacle-aware motion-planning problem that produces a waypoint sequence. We use a short-horizon, relative-motion diffusion policy to achieve the waypoints in sequence. We train mobile manipulation policies on a Spot robot that has to push and pull an office chair. Our hierarchical manipulation policy performs consistently better, especially when the horizon increases, compared to a diffusion policy trained on long-horizon demonstrations or motion planning assuming a rigidly-attached object (success rate of 8 (versus 0 and 5 respectively) out of 10 runs). Importantly, our learned policy generalizes to new layouts, grasps, chairs, and flooring that induces more friction, without any further training, showing promise for other complex mobile manipulation problems. Project Page: https://yravan.github.io/plannerorderedpolicy/