A Deep Reinforcement Learning Approach for Online Parcel Assignment

TL;DR

This paper presents a deep reinforcement learning approach for online parcel assignment, modeling the problem with a novel MDP framework and demonstrating competitive performance against traditional methods in minimizing delivery costs.

Contribution

The paper introduces PPO-OPA, a deep RL algorithm with attention networks for online parcel assignment, addressing stochastic challenges and outperforming traditional proportional assignment methods.

Findings

PPO-OPA achieves lower total costs than traditional methods.

The algorithm maintains constraint violations within acceptable limits.

Performance is comparable to the Primal-Dual algorithm without requiring known parcel volume.

Abstract

In this paper, we investigate the online parcel assignment (OPA) problem, in which each stochastically generated parcel needs to be assigned to a candidate route for delivery to minimize the total cost subject to certain business constraints. The OPA problem is challenging due to its stochastic nature: each parcel's candidate routes, which depends on the parcel's origin, destination, weight, etc., are unknown until its order is placed, and the total parcel volume is uncertain in advance. To tackle this challenge, we propose the PPO-OPA algorithm based on deep reinforcement learning that shows competitive performance. More specifically, we introduce a novel Markov Decision Process (MDP) framework to model the OPA problem, and develop a policy gradient algorithm that adopts attention networks for policy evaluation. By designing a dedicated reward function, our proposed algorithm can achieve a lower total cost with smaller violation of constraints, comparing to the traditional method which assigns parcels to candidate routes proportionally. In addition, the performances of our proposed algorithm and the Primal-Dual algorithm are comparable, while the later assumes a known total parcel volume in advance, which is unrealistic in practice.