A causal discovery and inference framework for on-demand food delivery delays

Read full text Summarise with VIRAT
Authors: Miaojia Lu, Rui Liu, Zhicheng Jin, Quan Yuan
Cited by: 0
References: 42

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The on-demand food delivery (OFD) industry has seen growth yet grapples with arrival delays. For major platforms like DoorDash and Uber Eats, over one-third of orders arrive late, highlighting the severity of this challenge. Currently, few studies distinguish true drivers of these delays from mere correlations. This study addresses this gap by developing an innovative framework combining Bayesian causal discovery with double machine learning. From 405,180 OFD records in China, we found that 16.7% of orders experienced delays. Pickup and transport durations exhibited the strongest causal effects to these delays. In addition, delay propagation was first identified within OFD services, where delays in preceding orders significantly increase the length of subsequent delays. The findings offer practical insights for OFD platforms to reduce order delays, such as optimizing courier pickup processes and mitigating delay propagation. By targeting these root causes, platforms can enhance operational efficiency and make their services more sustainable.

Fields of Study & Subjects

Urban and Freight Transport Logistics Digital Economy and Work Transformation Vehicle Routing Optimization Methods

Publication details

Journal / source
npj Sustainable Mobility and Transport
Volume
3
Issue
1
Publication date
25 March 2026
Language
EN