GRAND: Guidance, Rebalancing, and Assignment for Networked Dispatch in Multi-Agent Path Finding

TL;DR

GRAND is a hierarchical multi-agent dispatch algorithm that combines learned guidance with optimization to improve throughput and scalability in large robot fleets within warehouse environments.

Contribution

It introduces a novel hybrid approach integrating reinforcement learning-based guidance with lightweight optimization for efficient multi-agent task scheduling.

Findings

Improves throughput by up to 10% over the 2024 winning scheduler.

Maintains real-time execution within a 1-second compute budget.

Scalable to large fleets of up to 500 agents in congested warehouse scenarios.

Abstract

Large robot fleets are now common in warehouses and other logistics settings, where small control gains translate into large operational impacts. In this article, we address task scheduling for lifelong Multi-Agent Pickup-and-Delivery (MAPD) and propose a hybrid method that couples learning-based global guidance with lightweight optimization. A graph neural network policy trained via reinforcement learning outputs a desired distribution of free agents over an aggregated warehouse graph. This signal is converted into region-to-region rebalancing through a minimum-cost flow, and finalized by small, local assignment problems, preserving accuracy while keeping per-step latency within a 1 s compute budget. We call this approach GRAND: a hierarchical algorithm that relies on Guidance, Rebalancing, and Assignment to explicitly leverage the workspace Network structure and Dispatch agents to tasks. On congested warehouse benchmarks from the League of Robot Runners (LoRR) with up to 500 agents, our approach improves throughput by up to 10% over the 2024 winning scheduler while maintaining real-time execution. The results indicate that coupling graph-structured learned guidance with tractable solvers reduces congestion and yields a practical, scalable blueprint for high-throughput scheduling in large fleets.