BaB-ND: Long-Horizon Motion Planning with Branch-and-Bound and Neural Dynamics

TL;DR

This paper introduces BaB-ND, a GPU-accelerated branch-and-bound framework for long-horizon motion planning using neural dynamics models, effectively handling complex contact-rich manipulation tasks.

Contribution

It presents a novel branch-and-bound approach with specialized heuristics and bound propagation for neural dynamics, improving planning efficiency and quality in manipulation tasks.

Findings

Outperforms existing methods in contact-rich manipulation tasks

Generates high-quality trajectories in simulated and real-world settings

Supports various neural network architectures and scales efficiently

Abstract

Neural-network-based dynamics models learned from observational data have shown strong predictive capabilities for scene dynamics in robotic manipulation tasks. However, their inherent non-linearity presents significant challenges for effective planning. Current planning methods, often dependent on extensive sampling or local gradient descent, struggle with long-horizon motion planning tasks involving complex contact events. In this paper, we present a GPU-accelerated branch-and-bound (BaB) framework for motion planning in manipulation tasks that require trajectory optimization over neural dynamics models. Our approach employs a specialized branching heuristics to divide the search space into subdomains, and applies a modified bound propagation method, inspired by the state-of-the-art neural network verifier alpha-beta-CROWN, to efficiently estimate objective bounds within these subdomains. The branching process guides planning effectively, while the bounding process strategically reduces the search space. Our framework achieves superior planning performance, generating high-quality state-action trajectories and surpassing existing methods in challenging, contact-rich manipulation tasks such as non-prehensile planar pushing with obstacles, object sorting, and rope routing in both simulated and real-world settings. Furthermore, our framework supports various neural network architectures, ranging from simple multilayer perceptrons to advanced graph neural dynamics models, and scales efficiently with different model sizes.