MRGRP: Empowering Courier Route Prediction in Food Delivery Service with Multi-Relational Graph

TL;DR

MRGRP tackles the challenge of accurately predicting courier routes in rapid food delivery by modeling fine-grained task correlations with a Multi-Relational Graph (MR-graph) and a GraphFormer-based encoder. It couples this encoder with a reference-guided route decoder that leverages a competitive heuristic and dynamic context (distance and time) to prune the solution space and adapt to temporal changes, while an auxiliary ETA task provides timing cues. Offline tests show state-of-the-art improvements in LSD and SR across multiple cities, and online deployment on Meituan’s Turing platform yields substantial reductions in route deviation and better arrival-time predictions, translating to improved courier and user satisfaction and platform profitability. The results demonstrate that explicitly modeling task relations and using reference guidance are effective for scalable, dynamic route prediction in large-scale food delivery systems.

Abstract

Instant food delivery has become one of the most popular web services worldwide due to its convenience in daily life. A fundamental challenge is accurately predicting courier routes to optimize task dispatch and improve delivery efficiency. This enhances satisfaction for couriers and users and increases platform profitability. The current heuristic prediction method uses only limited human-selected task features and ignores couriers preferences, causing suboptimal results. Additionally, existing learning-based methods do not fully capture the diverse factors influencing courier decisions or the complex relationships among them. To address this, we propose a Multi-Relational Graph-based Route Prediction (MRGRP) method that models fine-grained correlations among tasks affecting courier decisions for accurate prediction. We encode spatial and temporal proximity, along with pickup-delivery relationships, into a multi-relational graph and design a GraphFormer architecture to capture these complex connections. We also introduce a route decoder that leverages courier information and dynamic distance and time contexts for prediction, using existing route solutions as references to improve outcomes. Experiments show our model achieves state-of-the-art route prediction on offline data from cities of various sizes. Deployed on the Meituan Turing platform, it surpasses the current heuristic algorithm, reaching a high route prediction accuracy of 0.819, essential for courier and user satisfaction in instant food delivery.

Figures (4)

• Figure 1: The overall framework of MRGRP.
• Figure 2: The schematic of MRGC layer.
• Figure 3: Ablation study. w/o Ref/Dyn/MRGC: without reference route solutions/dynamics features during decoding/MRGC layers.
• Figure 4: Case study. Orange, blue, and green lines correspond to real routes, routes predicted by Graph2Route, and MRGRP.