Training-Free Diffusion Planning with Analytical Local Scores

2026-10-01

AI Summary

This study addresses the reliance of diffusion-based planners on large-scale training data and their limited generalizability to novel maps by proposing a training-free diffusion motion planning method. The approach substitutes the learning of global scores with analytically derived local scores, leveraging local interactions to reconstruct trajectory evaluation. By deeply integrating classical optimization structures with the iterative refinement properties inherent in diffusion models, it eliminates the need for prior training. Experimental results demonstrate that the proposed method rapidly generates smooth, collision-free paths for over 300 agents in complex environments without any training phase. Overall, it significantly outperforms existing baseline methods in terms of comprehensive performance, offering a scalable and efficient solution for multi-agent motion planning in unseen scenarios.

Abstract

Path finding and multi-robot motion planning require trajectories that are smooth, goal-directed, and collision-free in environments with complex geometric constraints. Recent diffusion-based planners have shown that trajectory generation can be cast as iterative denoising which has opened the doors to learning-based approaches that can handle multi-modal trajectory distributions and refine entire trajectories. However, a key limitation is that diffusion planners require training on large collections of feasible trajectories, rendering them map-specific, and difficult to deploy when high-quality demonstrations are unavailable. This paper introduces a training-free diffusion-based motion planner that replaces learned global trajectory scores with analytical local scores derived from obstacle, smoothness, velocity, and inter-agent feasibility terms. The proposed idea relies on a key observation: the score of a trajectory can be reconstructed by considering only local interactions between neighboring waypoints and nearby constraints. This structure exploitation yields a decomposed denoising procedure that retains the optimization structure of classical trajectory methods while inheriting the iterative refinement behavior of diffusion models. Experiments on a large collection of complex environments and large multi-agent planning tasks show that the proposed analytical score produces smooth and feasible trajectories within limited computational costs, for example in generating feasible paths for 300+ agents in environments containing 100+ obstacles in under 6 seconds on a GPU, outperforming strong learning-based and optimization baselines, while avoiding the data requirements of learned diffusion planners.

Problem

• diffusion planner
• motion planning
• training-free
• multi-robot planning
• path finding

Innovation

• Training-Free Diffusion Planning
• Analytical Local Scores
• Multi-Robot Motion Planning
• Decomposed Denoising
• Trajectory Optimization