Learning to Solve the Min-Max Mixed-Shelves Picker-Routing Problem via Hierarchical and Parallel Decoding

TL;DR

This paper introduces a hierarchical and parallel multi-agent reinforcement learning approach to efficiently solve the min-max mixed-shelves picker-routing problem, achieving superior solution quality and speed for large-scale warehouse logistics tasks.

Contribution

It presents a novel hierarchical and parallel decoding method that improves coordination and scalability in solving MSPRP using multi-agent reinforcement learning.

Findings

State-of-the-art solution quality

Faster inference speed

Effective for large-scale instances

Abstract

The Mixed-Shelves Picker Routing Problem (MSPRP) is a fundamental challenge in warehouse logistics, where pickers must navigate a mixed-shelves environment to retrieve SKUs efficiently. Traditional heuristics and optimization-based approaches struggle with scalability, while recent machine learning methods often rely on sequential decision-making, leading to high solution latency and suboptimal agent coordination. In this work, we propose a novel hierarchical and parallel decoding approach for solving the min-max variant of the MSPRP via multi-agent reinforcement learning. While our approach generates a joint distribution over agent actions, allowing for fast decoding and effective picker coordination, our method introduces a sequential action selection to avoid conflicts in the multi-dimensional action space. Experiments show state-of-the-art performance in both solution quality and inference speed, particularly for large-scale and out-of-distribution instances. Our code is publicly available at http://github.com/LTluttmann/marl4msprp.