A machine learning approach for marginal fulfillment cost estimation in last mile delivery

Abstract

Determining marginal fulfillment costs (MFC) is crucial for effective decision-making in online grocery retail, a sector struggling with small profit margins and arduous service requirements of attended home deliveries. Paramount to improving operational efficiency, e-grocers need accurate real-time MFC estimations to optimize their service offers and prices for online customers. Traditional methods for estimating MFC are either too slow for online decision-making or inaccurate. This paper introduces a novel machine learning (ML) approach that provides fast and accurate MFC estimations with the help of carefully engineered features (predictors) that can capture complex routing dynamics. Experiments with real-world data demonstrate the superiority of the proposed approach over state-of-the-art MFC estimation methods. Our analysis of more than 2000 potential predictors, from which 20 are curated for practical applicability, reveals critical insights into the use of network-level, neighborhood-based, and node-level features in capturing complex VRP dynamics to develop ML-based approaches to address problems that arise in different transportation applications.