Data-Driven Last-Mile Shuttle Service Design: A Precision Matching and Service Optimization Model

Abstract

With the acceleration of urbanization and the rise of large-scale industrial parks, commuter traffic pressure is mounting, particularly the "last-mile" problem from public transport hubs to final destinations, which has become a critical bottleneck constraining urban mobility efficiency and sustainable development. Existing last-mile solutions, such as fixed-route shuttles, commonly suffer from service rigidity, resource wastage, and supply-demand mismatch. To address this challenge, this study proposes a new paradigm for last-mile shuttle design based on dynamic data, aiming to enhance service efficiency and user satisfaction through precision matching and service optimization. Taking the Shenzhen High-Tech Industrial Park as an empirical case, this research constructs a three-stage integrated design framework utilizing large-scale ride-hailing order data, Points of Interest (POI) data, and built environment data. First, an improved DBSCAN clustering algorithm and a spatio-temporal analysis model are employed to accurately identify and predict dynamically changing travel demands. Second, a Passenger-Stop-Vehicle (PSV) three-level precision matching model is proposed to achieve effective alignment between personalized demands and service resources. Finally, a multi-objective optimization model is formulated with the goals of minimizing operating costs, minimizing total passenger travel time, and maximizing system service coverage. An improved Genetic Algorithm (GA) is used to coordinately optimize the shuttle stop layout, routes, and schedules. Simulation experiments and comparative analysis demonstrate that the proposed model, compared to the traditional fixed-route model, can reduce average waiting times by approximately 36.3%, decrease vehicle deadheading rates by 25.4%, and improve overall operational profit by 96.3%. This research not only provides a data-driven, intelligent solution for the last-mile transportation problem in industrial parks but also offers theoretical support and practical reference for building more efficient and resilient urban public transport systems.