MAPFAST: A Deep Algorithm Selector for Multi Agent Path Finding using Shortest Path Embeddings

TL;DR

MAPFAST addresses the lack of a dominating optimal MAPF solver by building a CNN-based algorithm selector that leverages single-agent shortest-path embeddings and additional loss terms to choose the fastest solver from a diverse portfolio (CBS, CBSH, BCP, SAT). The approach demonstrably outperforms all portfolio solvers and a state-of-the-art MAPF selector on a large, diverse dataset, with ablation studies confirming the benefits of shortest-path encoding and multi-terminal losses. The work provides both practical algorithm selection gains and insights into when different MAPF solvers excel, informing future heuristic designs and problem decomposition strategies. Overall, MAPFAST enables faster, scalable MAPF solving across varied environments and contributes a data-driven framework for per-instance solver selection.

Abstract

Solving the Multi-Agent Path Finding (MAPF) problem optimally is known to be NP-Hard for both make-span and total arrival time minimization. While many algorithms have been developed to solve MAPF problems, there is no dominating optimal MAPF algorithm that works well in all types of problems and no standard guidelines for when to use which algorithm. In this work, we develop the deep convolutional network MAPFAST (Multi-Agent Path Finding Algorithm SelecTor), which takes a MAPF problem instance and attempts to select the fastest algorithm to use from a portfolio of algorithms. We improve the performance of our model by including single-agent shortest paths in the instance embedding given to our model and by utilizing supplemental loss functions in addition to a classification loss. We evaluate our model on a large and diverse dataset of MAPF instances, showing that it outperforms all individual algorithms in its portfolio as well as the state-of-the-art optimal MAPF algorithm selector. We also provide an analysis of algorithm behavior in our dataset to gain a deeper understanding of optimal MAPF algorithms' strengths and weaknesses to help other researchers leverage different heuristics in algorithm designs.

Figures (6)

• Figure 1: Accuracy and coverage data for portfolio algorithms with respect to different map types. Random and Warehouse maps are labeled as Rand and Ware respectively.
• Figure 2: Maps in Table \ref{['tab:simu_game']}. Left to right, first row: brc202d, orz900d, den312d; second row: den520d, lak303d, ost003d.
• Figure 3: (a) MAPF instance marked only with start (green circles) and goal (blue squares) locations. (b)(c) Two different mappings of the start and goal locations with respect to the map in (a). Planned paths are marked in colored lines.
• Figure 4: Encoding an instance map with (a) start and goal locations and (b) single-agent shortest paths.
• Figure 5: The CNN architecture of MAPFAST.