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

Learning-guided Prioritized Planning for Lifelong Multi-Agent Path Finding in Warehouse Automation

About

Lifelong Multi-Agent Path Finding (MAPF) is critical for modern warehouse automation, which requires multiple robots to continuously navigate conflict-free paths to optimize the overall system throughput. However, the complexity of warehouse environments and the long-term dynamics of lifelong MAPF often demand costly adaptations to classical search-based solvers. While machine learning methods have been explored, their superiority over search-based methods remains inconclusive. In this paper, we introduce Reinforcement Learning (RL) guided Rolling Horizon Prioritized Planning (RL-RH-PP), the first framework integrating RL with search-based planning for lifelong MAPF. Specifically, we leverage classical Prioritized Planning (PP) as a backbone for its simplicity and flexibility in integrating with a learning-based priority assignment policy. By formulating dynamic priority assignment as a Partially Observable Markov Decision Process (POMDP), RL-RH-PP exploits the sequential decision-making nature of lifelong planning while delegating complex spatial-temporal interactions among agents to reinforcement learning. An attention-based neural network autoregressively decodes priority orders on-the-fly, enabling efficient sequential single-agent planning by the PP planner. Evaluations in realistic warehouse simulations show that RL-RH-PP achieves the highest total throughput among baselines and generalizes effectively across agent densities, planning horizons, and warehouse layouts. Our interpretive analyses reveal that RL-RH-PP proactively prioritizes congested agents and strategically redirects agents from congestion, easing traffic flow and boosting throughput. These findings highlight the potential of learning-guided approaches to augment traditional heuristics in modern warehouse automation.

Han Zheng, Yining Ma, Brandon Araki, Jingkai Chen, Cathy Wu• 2026

Related benchmarks

TaskDatasetResultRank
Multi-Agent Path FindingSymbotic warehouse map
Total TP1.74e+3
6
Multi-Agent Path FindingAmazon warehouse map
Total TP3.17e+3
6
Lifelong Multi-Agent Path FindingSymbotic map N=80 agents
TPA18.38
5
Lifelong Multi-Agent Path FindingAmazon map N=80 agents
TPA31.8
5
Lifelong Multi-Agent Path FindingAmazon map N=100 agents
TPA28.84
5
Lifelong Multi-Agent Path FindingAmazon map N=120 agents
TPA25.56
5
Lifelong Multi-Agent Path FindingSymbotic map N=100 agents
TPA15.38
5
Lifelong Multi-Agent Path FindingSymbotic map N=120 agents
TPA11.31
5
Showing 8 of 8 rows

Other info

Follow for update