MAPF-LNS2: Repairing Multi-Agent Path Finding via Large Neighborhood Search - Robotics Institute Carnegie Mellon University

MAPF-LNS2: Repairing Multi-Agent Path Finding via Large Neighborhood Search

Jiaoyang Li, Zhe Chen, Daniel Harabor, Peter J. Stuckey, and Sven Koenig
Conference Paper, Proceedings of 36th AAAI Conference on Artificial Intelligence (AAAI '22), February, 2022

Abstract

Multi-Agent Path Finding (MAPF) is the problem of planning collision-free paths for multiple agents in a shared environment. In this paper, we propose a novel algorithm LNS2 based on large neighborhood search for solving MAPF efficiently. Starting from a set of paths that contain collisions, LNS2 repeatedly selects a subset of colliding agents and replans their paths to reduce the number of collisions until the paths become collision-free. We compare LNS2 against a variety of state-of-the-art MAPF algorithms, including Prioritized Planning with random restarts, EECBS, and PPS, and show that LNS2 runs significantly faster than them while still providing near-optimal solutions in most cases. With a runtime limit of just 5 minutes, LNS2 solves 80 percent of the random-scenario instances with the largest number of agents from the MAPF benchmark suite with a runtime limit of just 5 minutes, which, to our knowledge, has not been achieved by any existing algorithms.

BibTeX

@conference{Li-2022-131388,
author = {Jiaoyang Li and Zhe Chen and Daniel Harabor and Peter J. Stuckey and Sven Koenig},
title = {MAPF-LNS2: Repairing Multi-Agent Path Finding via Large Neighborhood Search},
booktitle = {Proceedings of 36th AAAI Conference on Artificial Intelligence (AAAI '22)},
year = {2022},
month = {February},
}