Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Lifelong Multi-Agent Path Finding in Large-Scale Warehouses

About

Multi-Agent Path Finding (MAPF) is the problem of moving a team of agents to their goal locations without collisions. In this paper, we study the lifelong variant of MAPF, where agents are constantly engaged with new goal locations, such as in large-scale automated warehouses. We propose a new framework Rolling-Horizon Collision Resolution (RHCR) for solving lifelong MAPF by decomposing the problem into a sequence of Windowed MAPF instances, where a Windowed MAPF solver resolves collisions among the paths of the agents only within a bounded time horizon and ignores collisions beyond it. RHCR is particularly well suited to generating pliable plans that adapt to continually arriving new goal locations. We empirically evaluate RHCR with a variety of MAPF solvers and show that it can produce high-quality solutions for up to 1,000 agents (= 38.9\% of the empty cells on the map) for simulated warehouse instances, significantly outperforming existing work.

Jiaoyang Li, Andrew Tinka, Scott Kiesel, Joseph W. Durham, T. K. Satish Kumar, Sven Koenig• 2020

Related benchmarks

TaskDatasetResultRank
Multi-Agent Path FindingAmazon warehouse map
Total TP512
6
Multi-Agent Path Findingmaze 32-32-2 (# agents: 30)
UA Conflicts7.95
6
Multi-Agent Path FindingSymbotic warehouse map
Total TP1.67e+3
6
Multi-Agent Path Findingempty-32-32 # agents: 300
UA Conflicts9.5
6
Multi-Agent Path Findingrandom-32-32-10 # agents: 300
UA Conflicts10.85
6
Multi-Agent Path Findingrandom-32-32-20 # agents: 200
UA Conflicts9.15
6
Multi-Agent Path Findingden312d agents: 200
UA Conflicts19.97
6
Multi-Agent Path Findinght_chantry # agents: 500
UA Conflicts21.15
6
Multi-Agent Path Findingmaze-32-32-4 (# agents: 30)
UA Conflicts5.69
6
Lifelong Multi-Agent Path FindingAmazon map N=80 agents
TPA4.94
5
Showing 10 of 15 rows

Other info

Follow for update