Graph Attention-Guided Search for Dense Multi-Agent Pathfinding
This work addresses real-time coordination problems for multi-agent systems in dense environments, representing an incremental improvement over prior hybrid methods.
The paper tackled the challenge of finding near-optimal solutions for dense multi-agent pathfinding (MAPF) in real-time by developing LaGAT, a hybrid framework that integrates a learned heuristic from MAGAT into the LaCAM search algorithm, outperforming both purely search-based and learning-based methods in dense scenarios.
Finding near-optimal solutions for dense multi-agent pathfinding (MAPF) problems in real-time remains challenging even for state-of-the-art planners. To this end, we develop a hybrid framework that integrates a learned heuristic derived from MAGAT, a neural MAPF policy with a graph attention scheme, into a leading search-based algorithm, LaCAM. While prior work has explored learning-guided search in MAPF, such methods have historically underperformed. In contrast, our approach, termed LaGAT, outperforms both purely search-based and purely learning-based methods in dense scenarios. This is achieved through an enhanced MAGAT architecture, a pre-train-then-fine-tune strategy on maps of interest, and a deadlock detection scheme to account for imperfect neural guidance. Our results demonstrate that, when carefully designed, hybrid search offers a powerful solution for tightly coupled, challenging multi-agent coordination problems.