Flow-Based Task Assignment for Large-Scale Online Multi-Agent Pickup and Delivery

2025-08-07

AI Summary

This paper addresses the dynamic task assignment problem in online Multi-Agent Pickup and Delivery (MAPD). To overcome limitations of existing heuristic or computationally expensive approaches, we propose an integrated task assignment framework based on Minimum-Cost Flow (MCF) modeling. Specifically, we formulate a congestion-aware MCF model on the environment graph and design two edge cost functions that eliminate the need for precomputed pairwise distances, enabling simultaneous task assignment and path guidance. Our method incorporates graph-network-based optimization, real-time traffic estimation, and guided path extraction. Empirically, it processes over 20,000 agents and 30,000 tasks within one second—demonstrating unprecedented scalability and assignment quality. The approach achieves breakthroughs in both computational efficiency and solution optimality, outperforming prior methods in large-scale, dynamic MAPD settings.

Technology Category

• Planning, Routing, and Scheduling: Planning with Markov Models (MDPs, POMDPs)
• Multiagent Systems: Multiagent Planning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We study the problem of online Multi-Agent Pickup and Delivery (MAPD), where a team of agents must repeatedly serve dynamically appearing tasks on a shared map. Existing online methods either rely on simple heuristics, which result in poor decisions, or employ complex reasoning, which suffers from limited scalability under real-time constraints. In this work, we focus on the task assignment subproblem and formulate it as a minimum-cost flow over the environment graph. This eliminates the need for pairwise distance computations and allows agents to be simultaneously assigned to tasks and routed toward them. The resulting flow network also supports efficient guide path extraction to integrate with the planner and accelerates planning under real-time constraints. To improve solution quality, we introduce two congestion-aware edge cost models that incorporate real-time traffic estimates. This approach supports real-time execution and scales to over 20000 agents and 30000 tasks within 1-second planning time, outperforming existing baselines in both computational efficiency and assignment quality.

Problem

• Online task assignment for large-scale multi-agent pickup and delivery
• Scalable flow-based method avoiding pairwise distance computations
• Congestion-aware edge costs for real-time traffic optimization

Innovation

• Flow-based task assignment on environment graph
• Congestion-aware edge cost models
• Efficient guide path extraction for planning