Spatial patterns and community structures of urban freight: Network insights into diesel and electric truck mobility

Understanding the mobility patterns of urban freight trucks is crucial for optimizing freight systems and urban planning. With the increasing adoption of electric trucks in urban areas, the spatial dynamics of urban freight are undergoing substantial changes. However, the lack of research on the mobility patterns of electric trucks has resulted in an incomplete understanding of the evolving freight network structure. This study constructs urban freight networks using large-scale trajectory data from both diesel and electric trucks and proposes an integrated community detection framework based on graph-encoder architecture. By simultaneously optimizing network topology and spatial attributes, this method captures more realistic network community partitioning and spatial interaction structures within these networks. Key findings show stark structural differences: diesel truck networks form hierarchical “core-periphery” structures dominated by peripheral hubs, where expanding community size increases low-degree nodes and fragments connectivity. Electric truck networks, in contrast, form non-hierarchical, highly interconnected spatial clusters that directly integrate industrial, commercial, and residential zones, maintaining robust connectivity across large communities and reducing delivery path lengths. These findings provide critical insights for electrification-driven urban freight transformation, enabling policymakers to leverage spatial differences between diesel and electric truck networks to optimize freight flows in support of the co-design of environmentally sustainable urban freight systems and socially inclusive urban development.