Planner-Conditioned Diffusion for Coordinated Multi-Agent Exploration
2026-08-17 • Robotics
Robotics
AI summaryⓘ
The authors developed a new way for multiple robots to explore an area together without bumping into each other too much. Instead of each robot deciding step-by-step, their method plans whole paths ahead using a model that learns from many different planning styles. They combine these paths and pick the best set to reduce overlap, helping robots cover more ground efficiently. Tests showed their method works better than older approaches and even helped two real robots avoid each other naturally without special rules.
multi-agent explorationdiffusion-based plannerstrajectory planningplanner-conditioned policymulti-modal learningcoordinationlocal rerankingrobot navigationmulti-agent systems
Authors
Marcus Yu Siong Teo, Jeric Lew, Tanishq Duhan, Guillaume Sartoretti
Abstract
Coordinated multi-agent exploration requires not only efficient individual coverage but also non-redundant coverage across agents over extended planning horizons. Conventional approaches rely on hand-crafted coordination rules, while end-to-end multi-agent learning methods are difficult to scale and train. Diffusion-based planners such as DARE offer a promising alternative by generating long-horizon trajectories instead of single-step actions, but existing methods are trained on a narrow planner distribution, limiting behavioral diversity and inference-time controllability. We propose a Planner-Conditioned Diffusion Policy (PCDP) for graph-based multi-agent exploration. PCDP is trained on demonstrations from multiple planner styles with planner identity as an explicit conditioning input, enabling a single shared model to learn a multimodal trajectory distribution and generate diverse, controllable trajectory candidates from the same observation. Rather than learning coordination end-to-end, we reuse this multimodal single-agent policy across all agents and introduce coordination through local reranking, in which nearby agents jointly select the trajectory combination with minimal predicted overlap. We evaluate PCDP against classical and diffusion-based baselines on 100 held-out maps in a four-agent simulation setting. PCDP matches the perfect success rate of the diffusion-based baselines while improving mean max-agent travel, total team travel, and agent imbalance. Crucially, reranking alone over a single-planner baseline yields only marginal gains, indicating that planner-conditioned multimodality is the main contributor to improved coordination. Qualitative simulation results and real-robot experiments with two agents further validate that diverse long-horizon trajectory generation produces emergent spatial separation between agents without any explicit repulsion mechanism.