Data - and Knowledge-Driven Urban Safety Risk Evolution, Forecasting, and Early Warning

¹ Department of Safety Science and Engineering, School of Engineering and Technology, China University of Geosciences (Beijing), Beijing, 100083, China
² Institute of Urban Safety and Environmental Science, Beijing Academy of Science and Technology, Beijing, 100054, China

^a 3002230050@email.cugb.edu.cn
^* Corresponding author: daibq@163.com

Abstract

This paper reviews recent advances and remaining gaps in urban safety and disaster risk assessment, intelligent transportation and predictive control, warning signals and cross-domain forecasting, and data governance and risk management. Evidence shows that complexity science, machine learning, and multi-source data fusion substantially improve understanding of mobility patterns, forecasting of infrastructure evolution, and accuracy of disaster assessments. Trajectory optimization, multi-level risk assessment, and microclimate simulation advance the safety of autonomous driving and urban air mobility. Probabilistic models, multimodal analysis, and artificial intelligence enhance early risk recognition. Policy analysis, digital twins, and AI-generated content (AIGC) offer new avenues for risk prevention and control. Major challenges remain, including inadequate dynamic adaptability of models, limited cross-domain collaboration, and insufficient integration between data governance and risk management. Future work should strengthen model self-adaptation and updating, optimize heterogeneous data fusion, and build collaborative mechanisms that align data governance with risk management to improve the safety and resilience of complex urban systems.