Windowed MAPF with Completeness Guarantees
Rishi Veerapaneni, Muhammad Suhail Saleem, Jiaoyang Li, Maxim Likhachev
TL;DR
This work tackles windowed multi-agent path finding (MAPF) with theoretical completeness guarantees. It introduces WinC-MAPF, a general framework that enforces completeness by applying real-time single-agent heuristic updates to the joint configuration space and by updating heuristics for disjoint groups of agents, using an Action Generator that minimizes $c(\\mathcal{C},\\mathcal{C}^W) + h(\\mathcal{C}^W)$. As a concrete instantiation, it presents Single-Step CBS (SS-CBS), a CBS-based AG that handles heuristic penalties via heuristic conflicts and reports disjoint groups, enabling complete planning with $W=1$. Empirically, SS-CBS outperforms windowed baselines across congested scenarios, showing that windowed MAPF can be both complete and practically effective for fast replanning in dense environments.
Abstract
Traditional multi-agent path finding (MAPF) methods try to compute entire start-goal paths which are collision free. However, computing an entire path can take too long for MAPF systems where agents need to replan fast. Methods that address this typically employ a "windowed" approach and only try to find collision free paths for a small windowed timestep horizon. This adaptation comes at the cost of incompleteness; all current windowed approaches can become stuck in deadlock or livelock. Our main contribution is to introduce our framework, WinC-MAPF, for Windowed MAPF that enables completeness. Our framework uses heuristic update insights from single-agent real-time heuristic search algorithms as well as agent independence ideas from MAPF algorithms. We also develop Single-Step CBS (SS-CBS), an instantiation of this framework using a novel modification to CBS. We show how SS-CBS, which only plans a single step and updates heuristics, can effectively solve tough scenarios where existing windowed approaches fail.