Simulation-Informed Diffusion for Decentralized Multi-robot Motion Planning

2026-05-26

AI Summary

This work addresses the challenge in decentralized multi-robot systems operating in dense, complex environments, where reliance on local static observations alone hinders accurate prediction of neighboring robots’ behaviors and often leads to collisions. To overcome this limitation, the authors propose the Simulation-Informed Diffusion (SID) framework, which introduces a constraint-aware diffusion model to predict future trajectories of nearby robots. Leveraging simulation-derived insights, each robot plans its own path under explicit safety constraints while employing an event-triggered lightweight communication mechanism to coordinate collision avoidance. By unifying trajectory prediction and motion planning within a single framework, SID achieves superior performance in large-scale scenarios involving 108 robots and 160 obstacles, demonstrating high planning success rates, strong constraint satisfaction, and excellent scalability compared to existing approaches.

Abstract

Decentralized multi-robot motion planning requires each robot to generate collision-free trajectories from local observations, without global sensing or reliable communication. However, most existing planners, whether classical or learning-based, generate trajectories from a static snapshot of the local observation, which limits their ability to anticipate the future behavior of neighboring robots. This limitation is critical as the number of robots increases and the environment becomes more cluttered. To overcome this challenge, this paper introduces Simulation-Informed Diffusion (SID), a decentralized framework built on constraint-aware diffusion models (CADM). SID first uses CADM to simulate the future trajectories of neighboring robots from their currently observed states, and then uses the same CADM to plan each robot's own trajectory under safety constraints informed by these simulations. Crucially, the accurate simulation of neighbors enables a minimal communication scheme that triggers coordination only when necessary in highly congested scenarios. Experiments across diverse environments show that SID consistently outperforms baseline methods in terms of planning effectiveness and constraint satisfaction, and scales to scenarios with 108 robots and 160 obstacles.

Problem

• decentralized multi-robot motion planning
• collision-free trajectories
• local observations
• future behavior prediction
• scalability

Innovation

• Simulation-Informed Diffusion
• constraint-aware diffusion models
• decentralized motion planning
• multi-robot coordination
• minimal communication