Unraveling intra-urban freight parking patterns: A data-driven geospatial study of shared logistics sector in Hong Kong

Abstract

Insufficient urban parking in densely populated cities has led to challenges like traffic congestion and unauthorized parking. While existing literature extensively covers parking behaviors among private vehicles and taxis for commuters, limited research has been proposed centering on urban shared freight activities. It can be problematic to neglect shared freight activities because these logistics activities often require more flexible parking space for goods loading/unloading near urban destinations. This study therefore introduces a geospatial data-driven approach to investigate urban parking behaviors from shared freight activities using GPS trajectories collected in Hong Kong. The analysis involves three main folds: using rank-size distribution and log-odds ratio to comparatively examine spatial heterogeneity of parking, evaluating illegal on-street parking events via illegal parking index, and quantifying relationships between parking behaviors and urban functions using random forest (RF) and SHapley Additive exPlanations (SHAP). The findings reveal significant concentrations of parking events on weekdays/weekends and with different parking durations. Particularly, high illegal on-street parking concentration is reported in Tsuen Wan and Kwai Chung, where the port container terminals and industrial activities are located. We further discover that industrial and dining densities both indicate significantly positive impacts on the relationship to the increases of overall parking frequency and illegal parking index. This study can be of great interest to current parking management and enforcement strategies in cities and relevant practical implications are discussed.