An AI-driven planning framework for ice and snow towns: integrating optical remote sensing and machine learning

Urban planning in ice and snow towns is uniquely challenging due to extreme climatic conditions, complex terrains, and dynamic snow cover patterns. Traditional methods often struggle to adapt to these rapidly changing environments, highlighting the need for smart, data-driven approaches. This study proposes an artificial intelligence (AI)-driven framework that integrates optical remote sensing technologies with machine learning models to address these challenges. High-resolution satellite and UAV imagery are utilized to extract critical features such as snow cover and terrain stability. A U-Net-based model segments the study area into low-, medium-, and high-risk zones, forming a reliable foundation for hazard assessment. A reinforcement learning module then optimizes urban layouts, minimizing risk exposure while maintaining accessibility and functionality. The framework achieves a mean Intersection over Union (mIoU) of 87.4% and reduces risk exposure by 42% in theoretical validation. By advancing risk-based planning and layout optimization, this research contributes to the development of smarter, safer, and more resilient urban environments in extreme climates, offering a scalable solution for future smart city initiatives.