Article:
Journal Version: Zijian Zhao, Sen Li*, "Discriminatory order assignment and payment-setting of on-demand food-delivery platforms: A multi-action and multi-agent reinforcement learning framework"，Transportation Research Part E: Logistics and Transportation Review (TR_E), 2025
Conference Version: Zijian Zhao, Sen Li*, "Multi-Agent Reinforcement Learning for Order Assignment and Payment Setting on Food-Delivery Platforms: The Implicit Algorithmic Biases", ISTDM 2025
Notice: We have updated the code in line with our recent work, "The Impacts of Data Privacy Regulations on Food-Delivery Platforms" (Transportation Research Part C 2025). You can access the new-version code and find instructions on how to run it at the GDPR-Food-Delivery. The latest version will be provided at AV-Food-Delivery
Acknowledgement: Some parts of the code (simulator part) is based on the work of Yulong Hu.
Due to copyright restrictions, we cannot provide the data used in this paper. However, we offer a brief introduction to the data format so you can utilize our code with your own dataset.
Our dataset consists of one hour of food delivery data in Hong Kong, China, containing approximately 10,000 records. It is saved in a CSV file, where each column represents an attribute and each row corresponds to an order. The relevant attributes include:
- dlat: Latitude of the destination
- dlon: Longitude of the destination
- plat: Latitude of the origin
- plon: Longitude of the origin
- minute: The minute at which the order is placed
As you can see, there is no ground truth for salary information. Therefore, we simply set the reservation value to range from 0.85 to 1.15, without a specific unit.
Notice: We have provided the synthetic food delivery data generated by a GAN, thanks to Yitong Shang.
For route planning, we utilize the Project-OSRM/osrm-backend: Open Source Routing Machine - C++ backend.
@article{ZHAO2026104653,
title = {Discriminatory order assignment and payment-setting of on-demand food-delivery platforms: A multi-action and multi-agent reinforcement learning framework},
journal = {Transportation Research Part E: Logistics and Transportation Review},
volume = {208},
pages = {104653},
year = {2026},
issn = {1366-5545},
doi = {https://doi.org/10.1016/j.tre.2025.104653},
url = {https://www.sciencedirect.com/science/article/pii/S1366554525006751},
author = {Zijian Zhao and Sen Li}