RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Tang, Yimin, Xiong, Xiao, Xi, Jingyi, Li, Jiaoyang, Bıyık, Erdem, Koenig, Sven
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913976804179968
author Tang, Yimin
Xiong, Xiao
Xi, Jingyi
Li, Jiaoyang
Bıyık, Erdem
Koenig, Sven
author_facet Tang, Yimin
Xiong, Xiao
Xi, Jingyi
Li, Jiaoyang
Bıyık, Erdem
Koenig, Sven
contents Multi-Agent Path Finding (MAPF), which focuses on finding collision-free paths for multiple robots, is crucial for applications ranging from aerial swarms to warehouse automation. Solving MAPF is NP-hard so learning-based approaches for MAPF have gained attention, particularly those leveraging deep neural networks. Nonetheless, despite the community's continued efforts, all learning-based MAPF planners still rely on decentralized planning due to variability in the number of agents and map sizes. We have developed the first centralized learning-based policy for MAPF problem called RAILGUN. RAILGUN is not an agent-based policy but a map-based policy. By leveraging a CNN-based architecture, RAILGUN can generalize across different maps and handle any number of agents. We collect trajectories from rule-based methods to train our model in a supervised way. In experiments, RAILGUN outperforms most baseline methods and demonstrates great zero-shot generalization capabilities on various tasks, maps and agent numbers that were not seen in the training dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2503_02992
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks
Tang, Yimin
Xiong, Xiao
Xi, Jingyi
Li, Jiaoyang
Bıyık, Erdem
Koenig, Sven
Robotics
Artificial Intelligence
Multi-Agent Path Finding (MAPF), which focuses on finding collision-free paths for multiple robots, is crucial for applications ranging from aerial swarms to warehouse automation. Solving MAPF is NP-hard so learning-based approaches for MAPF have gained attention, particularly those leveraging deep neural networks. Nonetheless, despite the community's continued efforts, all learning-based MAPF planners still rely on decentralized planning due to variability in the number of agents and map sizes. We have developed the first centralized learning-based policy for MAPF problem called RAILGUN. RAILGUN is not an agent-based policy but a map-based policy. By leveraging a CNN-based architecture, RAILGUN can generalize across different maps and handle any number of agents. We collect trajectories from rule-based methods to train our model in a supervised way. In experiments, RAILGUN outperforms most baseline methods and demonstrates great zero-shot generalization capabilities on various tasks, maps and agent numbers that were not seen in the training dataset.
title RAILGUN: A Unified Convolutional Policy for Multi-Agent Path Finding Across Different Environments and Tasks
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2503.02992