MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at Scale

2024-08-29

arXiv.org

AI Summary

This work addresses Multi-Agent Path Finding (MAPF), an NP-hard combinatorial optimization problem. We propose MAPF-GPT, the first end-to-end foundation model for MAPF built solely on imitation learning and a Transformer architecture. Trained via behavioral cloning on expert trajectory data, it requires no hand-crafted heuristics, inter-agent communication, or online search. Its key contribution is the first demonstration of zero-shot generalization in MAPF: given unseen maps, obstacle configurations, and agent counts, it directly generates collision-free paths without fine-tuning. Extensive experiments on diverse MAPF benchmarks show that MAPF-GPT significantly outperforms existing learnable solvers in solution quality and success rate, while achieving high inference efficiency and strong scalability—enabling real-time deployment on large-scale instances.

Technology Category

• Multiagent Systems: Multiagent Planning
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Multi-agent pathfinding (MAPF) is a problem that generally requires finding collision-free paths for multiple agents in a shared environment. Solving MAPF optimally, even under restrictive assumptions, is NP-hard, yet efficient solutions for this problem are critical for numerous applications, such as automated warehouses and transportation systems. Recently, learning-based approaches to MAPF have gained attention, particularly those leveraging deep reinforcement learning. Typically, such learning-based MAPF solvers are augmented with additional components like single-agent planning or communication. Orthogonally, in this work we rely solely on imitation learning that leverages a large dataset of expert MAPF solutions and transformer-based neural network to create a foundation model for MAPF called MAPF-GPT. The latter is capable of generating actions without additional heuristics or communication. MAPF-GPT demonstrates zero-shot learning abilities when solving the MAPF problems that are not present in the training dataset. We show that MAPF-GPT notably outperforms the current best-performing learnable MAPF solvers on a diverse range of problem instances and is computationally efficient during inference.

Problem

• Develops imitation learning for multi-agent pathfinding.
• Uses transformer-based model without heuristics.
• Achieves zero-shot learning in diverse problem instances.

Innovation

• Imitation learning for MAPF
• Transformer-based neural network
• Zero-shot learning capabilities