A Distributed Hierarchical Spatio-Temporal Edge-Enhanced Graph Neural Network for City-Scale Dynamic Logistics Routing
Fundamental Scientific Reports in Multidisciplinary Areas · 10.06.2026 · 2 (02) · 320-332
DOI: 10.66238/fsrma89 ↗Kurz gefasst
Fachartikel in der Zeitschrift Fundamental Scientific Reports in Multidisciplinary Areas, online erschienen am 10.06.2026. Verfasst von Zihan Han und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
City-scale logistics routing has become increasingly challenging as metropolitan road networks expand to tens of millions of edges and traffic conditions evolve rapidly under high-volume mobility demands. Conventional centralized routing algorithms and monolithic graph neural network (GNN) models face significant limitations in scalability, latency, and real-time adaptability, restricting their effectiveness in large-scale urban logistics systems. To address these challenges, this paper proposes a Distributed Hierarchical Spatio-Temporal Edge-Enhanced Graph Neural Network (HSTE-GNN) for dynamic routing over ultra-large road networks. The proposed framework partitions the city-scale road graph into regional subgraphs that are processed in parallel across distributed computing nodes, enabling efficient learning of localized traffic dynamics. Within each partition, an edge-enhanced spatio-temporal module jointly captures node states, dynamic edge attributes, and short-term temporal dependencies. To ensure global routing coherence, a hierarchical coordination layer aggregates cross-region representations via an asynchronous parameter-server mechanism, allowing timely global route updates under high-frequency traffic changes. This distributed and hierarchical design balances local responsiveness with global consistency, significantly improving scalability and inference efficiency. Experiments conducted on real-world large-scale traffic datasets from Beijing and New York validate the effectiveness of the proposed approach. Compared with strong spatio-temporal baselines such as ST-GRAPH, the distributed HSTE-GNN achieves 34.9% lower routing delay, 14.7% lower MAPE, and 11.8% lower RMSE, while improving global route consistency by 7.3%. These results demonstrate that the proposed framework provides a scalable, adaptive, and efficient solution for next-generation intelligent transportation systems, particularly for large-scale logistics platforms requiring rapid and reliable routing decisions
Zitieren
APA
Han, Z., lm3193@columbia.edu, L. M., & Zhang, J. (2026). A Distributed Hierarchical Spatio-Temporal Edge-Enhanced Graph Neural Network for City-Scale Dynamic Logistics Routing. Fundamental Scientific Reports in Multidisciplinary Areas, 2(02), 320-332. https://doi.org/10.66238/fsrma89
Vancouver
Han Z, lm3193@columbia.edu LM, Zhang J. A Distributed Hierarchical Spatio-Temporal Edge-Enhanced Graph Neural Network for City-Scale Dynamic Logistics Routing. Fundamental Scientific Reports in Multidisciplinary Areas. 2026;2(02):320-332. doi:10.66238/fsrma89
BibTeX
@article{Han2026,
author = {Han, Zihan and lm3193@columbia.edu, Lingran Meng and Zhang, Jingwei},
title = {A Distributed Hierarchical Spatio-Temporal Edge-Enhanced Graph Neural Network for City-Scale Dynamic Logistics Routing},
journal = {Fundamental Scientific Reports in Multidisciplinary Areas},
year = {2026},
volume = {2},
number = {02},
pages = {320--332},
doi = {10.66238/fsrma89}
}
EUVIMED