2025 Vol.2
Urban Spatial Intelligence and Planning
Urban Scaling Revisited: Size, Scale, and Shape
Abstract: In this research, we define four different sets of relationships that tie together theories and methods that describe and explain how cities and their spatial locations change as they scale. By scaling we mean changes in the size of urban phenomena such as population that take place as cities grow and more generically, how cities change over time. These relationships cover city size distributions and the rank size rule, urban density functions that relate to how dense populations are with respect to their location around the central area of cities, how gravitational interactions between locations scale with respect to distance, and finally how attributes relating to size in cities such as income scale allometrically as cities change in size. These four relationships can be associated with those who first popularised their form, which in the order we introduce them, are what we call Zipf’s Law, Clark’s Law, Tobler’s Law, and Marshall’s Law. Having described their form, we illustrate their application to the distribution of employment and populations for small areas and for whole cities in the UK, reflecting a form of spatial intelligence for planning informed by the principles of urban scaling.
Spatiotemporal Intelligence and Refined Urban Governance in the New Era: Theories, Methods, and Application Approaches from an Urban Science Perspective
Abstract: Refined urban governance is an important dimension of the modernization of the national governance system and governance capacity, and spatiotemporal intelligence plays a significant role in this endeavor. This paper, from the perspective of urban science and considering the objectives and tasks of the New Era, attempts to discuss the theoretical basis, methodological framework, and application pathways for spatiotemporal intelligence to support urban governance. This paper first briefly reviews the research spectrum of spatiotemporal intelligence and points out that existing urban science research has established a solid foundation for urban governance from the perspectives of geographical patterns, mechanisms, and regulation, while also setting theoretical goals that can be achieved. Many methodological studies have also provided a rich toolkit for urban spatial pattern analysis, interaction and temporal dynamic pattern mining, problem attribution and trend prediction, as well as operational optimization. This paper further demonstrates, through specific cases, the application pathways of the aforementioned spatiotemporal intelligence tools in the complete closed-loop of urban governance, from urban situation awareness, spatial management, response and disposal, prediction and warning, to institutional building.
Research on Demand Measurement and Simulation Optimization of Unmanned Last Mile Delivery in Campus
Abstract: Unmanned last-mile logistics, as an emerging technological approach, has seen preliminary applications in practice. However, systematic and quantitative research on its implementation effectiveness and optimization strategies remains insufficient. This study aims to quantify the actual demand for unmanned last-mile logistics on campus and evaluate the social, economic, and ecological benefits across different automation levels and scenarios, providing scientific guidance for campus logistics planning. Taking the Tsinghua University campus as a typical application scenario, a logistics demand measurement model is developed by integrating the YOLOv8 and ByteTrack algorithms to automatically identify and count delivery riders from surveillance data. On this basis, combined with questionnaire data, a simulation is conducted using the AnyLogic platform to analyze the impact of unmanned last-mile logistics under various scenarios. The results indicate that unmanned last-mile logistics offers significant advantages in improving delivery efficiency and reducing energy consumption. However, it also leads to increased customer waiting times, and the service level still needs to be improved. Strategies such as increasing robot or rider speed, adding more robots, and expanding delivery hubs can enhance delivery efficiency, reduce customer waiting times, and lower energy consumption. Based on the simulation results, it is found that at least 40 robots are required on campus to complete delivery tasks. Furthermore, the layout of 4-5 delivery hubs, each equipped with 15-20 robots, is found to be highly cost-effective. It is also recommended that robot and rider speeds be controlled between 13-15 km/h to balance safety and efficiency.
Paradigm Shift of Street Visual Intelligence in Urban Planning
Abstract: Streets serve as the foundational elements of urban environments. Their visual characteristics directly reflect urban spatial quality and influence residents' well-being. Street view imagery captures street scenes from a human perspective, providing a unique viewpoint for micro-level urban environment analysis and serving as an indispensable resource for urban planning. This study systematically reviews the development of street view image applications in urban planning, especially focusing on their integration with advancing artificial intelligence (AI) techniques. From early manual audits to feature extraction powered by statistical machine learning, and later to automated analysis driven by deep learning, the efficiency and accuracy of street view images utilization have markedly improved. In recent years, the incorporation of self-supervised learning and large language models has markedly enhanced the application potential of street view images, enabling more complex urban analysis compared to earlier approaches. By reviewing the application of street view data across different technological stages, this study illustrates how artificial intelligence has reshaped its analytical paradigms and explores its promising potential for future urban planning.
Research on Emotional Space Identification and Regeneration Strategies for Urban Neighborhoods Based on Large Language Models: A Case Study of the North Sichuan Road Neighborhood in Shanghai
Abstract: Addressing the challenge of quantifying human emotional elements in urban renewal, this study constructs a research framework of "data collection – emotion recognition – urban renewal strategy" to explore human-centric renewal pathways driven by artificial intelligence technologies. By integrating large language models (LLMs) and knowledge graph technology, the research systematically synthesizes massive heterogeneous data from social media, resident hotlines, news supplement, and in-depth interviews. Taking Shanghai's North Sichuan Road neighborhood as a case study, it identifies eight categories of emotional spaces and conducts emotional scoring for 3 697 spatial data points. Key findings include: (1) There is a strong coupling relationship between emotional hotspots and physical urban fabric; (2) Nostalgia and joy collectively form the neighborhood's emotional foundation; (3) Negative emotional spaces exhibit functional mismatch-driven nodal clustering patterns. The study demonstrates that the synergistic application of multi-source data and AI technologies expands observational dimensions for emotional space analysis, providing a feasible approach to integrate technical tools with humanistic values in precision-oriented urban renewal.
Resilience of Shanghai's Urban Spatiotemporal Structure under Varying Flood Scenarios
Abstract: Urban space and function are highly coupled, forming a complex spatiotemporal system that not only accommodates and responds to various urban challenges but also serves as a critical target for enhancing urban resilience and informing planning interventions. Understanding the resilience dynamics of real-world urban systems and their intricate socio-economic effects is fundamental to the development of resilient cities. This study integrates multi-source data, including mobile signalling, road networks, and housing prices, to construct a simulation framework for urban commuting spatiotemporal structures. It further assesses their dynamic resilience characteristics and spatial-social disparities under varying levels of flood scenarios. By identifying resilience bottlenecks, the study proposes targeted planning interventions and applies the framework to Shanghai to unveil regional resilience variations and diverse resilience typologies. The findings indicate that the resilience of urban commuting spatiotemporal structures exhibits multidimensional complexity, with disparities among social strata exceeding those related to gender and manifesting as uneven socio-spatial exclusion effects driven by risk distribution. Moreover, the mitigation of commuting efficiency bottlenecks demonstrates significant temporal sensitivity and spatial heterogeneity, enabling precise intervention targeting. The results validate the effectiveness of the proposed approach, elucidate the flood response mechanisms of Shanghai's commuting spatiotemporal resilience, and offer insights for resilience optimisation and refined urban planning.
Climate Inequality to Urban Flooding Risks Driven by Multidimensional Factors
Abstract: Urban flooding has emerged as a significant challenge for densely populated cities, particularly in the context of increasing extreme rainfall events. This study focuses on the Guangdong-Hong Kong-Macao Greater Bay Area, developing a “hazard-exposure-vulnerability” framework to assess flood risks and resource equity using a variety of data sources. By employing a Bayesian-optimized LightGBM-SHAP method, the study uncovers the mechanisms underlying key disaster-inducing factors. Additionally, the distribution of flood defense resources is evaluated for spatial inequities using the Dagum Gini coefficient and Lorenz curve. The findings indicate that flood risk decreases from coastal to inland areas, with high-risk zones concentrated in core cities and densely riverine regions. There is a pronounced spatial imbalance in the allocation of safety resources, with the fragmentation of planning and infrastructure isolation in emerging areas exacerbating risk exposure. Key disaster drivers shift from topographical and physical factors to hydrological and ecological factors as risk levels rise. Impervious surface rate and vegetation coverage are identified as critical variables influencing overall flood risk. This research integrates risk mechanism analysis with equity-based assessments, offering scientific support and policy recommendations for climate-adaptive urban planning.