Exploring Spatial Patterns of Intracity Heavy-Duty Truck Flows in Shanghai Using Large-Scale Trajectory Data: A VSTC-TPS Approach

Heavy-duty trucks (HDTs) are crucial to the logistics system, and understanding their activity patterns is key for improving logistics performance and addressing challenges in economic, environmental, and infrastructure planning. A critical first step is to pinpoint the specifics of HDT stops—when, where, how long, and why they occur. However, existing methods, whether based on vehicle trajectories or surveys, are limited in processing speed and/or accuracy. This paper proposes the velocity-originated spatio-temporal clustering of trajectory point sequences (VSTC-TPS) method, which effectively uses both spatial and temporal dimensions of GPS data while also having a self-adaptive threshold to account for the diversity in GPS data quality of different HDTs. This method was tested using a data set of 712 million GPS records from more than 100,000 HDTs, representing all HDT operations in Shanghai, China, over a single week. The VSTC-TPS method demonstrated superior performance in terms of speed and accuracy when compared with several existing methods, as validated by ground truth data. Additionally, a network community detection method is applied to the vehicle stops identified by VSTC-TPS as potential applications of this method, revealing significant logistical connections and distinct freight communities within Shanghai. The proposed VSTC-TPS method supports more informed decision making among stakeholders in the logistics sector and provides valuable insights that contribute to sustainable urban development and the welfare of city inhabitants. It offers a comprehensive tool for enhancing the efficiency of urban freight systems and reducing their environmental impact, thereby supporting the development of smarter, safer, and more sustainable cities.