Harvesting Efficient On-Demand Order Pooling from Skilled Couriers: Enhancing Graph Representation Learning for Refining Real-time Many-to-One Assignments

Yile Liang, Jiuxia Zhao, Donghui Li, Jie Feng, Chen Zhang, Xuetao Ding, Jinghua Hao, Renqing He·June 20, 2024

Summary

The paper explores the optimization of on-demand food delivery systems, particularly focusing on Meituan Waimai, a prominent Chinese platform. The authors propose the Skilled Courier Delivery Network (SCDN), a novel approach that leverages graph representation learning and rich data from skilled couriers to improve order pooling, consolidate requests in real-time, and enhance courier efficiency during peak hours without compromising delivery times. SCDN addresses the challenge of efficient order assignment by considering non-additive MD scores and the computational complexity that increases with the number of orders and couriers. By deploying SCDN in Meituan's dispatch system, a 45-55% increase in courier efficiency during noon peaks was observed, demonstrating the potential of using courier knowledge and advanced algorithms to optimize the last-mile delivery process. The study highlights the importance of incorporating real-world data and efficient methods to streamline on-demand food delivery operations.

Key findings

7
  • header
  • header
  • header
  • header
  • header
  • header
  • header

Advanced features