The Combined Problem of Online Task Assignment and Lifelong Path Finding in Logistics Warehouses: A Case Study

2025-02-11

AI Summary

This paper formalizes and solves, for the first time, the joint online task assignment and lifelong path planning problem in logistics warehousing settings, aiming to maximize system throughput. Method: Departing from conventional sequential paradigms, we propose an end-to-end collaborative framework comprising (i) a rule-driven, congestion-aware lifelong path planner incorporating robot kinematic modeling and online heuristic task assignment, and (ii) a simulation-driven automatic parameter search mechanism enabling adaptive optimization of the task allocator with respect to the path planner. Contribution/Results: Evaluated on Meituan’s real-world warehouse simulation, our algorithm reduces execution time to 83.77% of the current production system and improves throughput by 8.09% over state-of-the-art methods. Moreover, it achieves equivalent throughput using 40% fewer robots.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We study the combined problem of online task assignment and lifelong path finding, which is crucial for the logistics industries. However, most literature either (1) focuses on lifelong path finding assuming a given task assigner, or (2) studies the offline version of this problem where tasks are known in advance. We argue that, to maximize the system throughput, the online version that integrates these two components should be tackled directly. To this end, we introduce a formal framework of the combined problem and its solution concept. Then, we design a rule-based lifelong planner under a practical robot model that works well even in environments with severe local congestion. Upon that, we automate the search for the task assigner with respect to the underlying path planner. Simulation experiments conducted in warehouse scenarios at extit{Meituan}, one of the largest shopping platforms in China, demonstrate that (a)~ extit{in terms of time efficiency}, our system requires only 83.77% of the execution time needed for the currently deployed system at Meituan, outperforming other SOTA algorithms by 8.09%; (b)~ extit{in terms of economic efficiency}, ours can achieve the same throughput with only 60% of the agents currently in use.

Problem

• Online task assignment integration
• Lifelong path finding optimization
• Warehouse logistics efficiency improvement

Innovation

• Integrated online task assignment
• Rule-based lifelong path planner
• Automated task assigner search