The Next Frontier in Sustainable Smart City Development

This chapter undertakes a comprehensive exploration of urban, flow, environmental, climate, and geospatial foundation model (FM) classes, analyzing their architectures, functional capabilities, and real-world applications in advancing data-driven decision-making in sustainable smart cities. The findings reveal that urban foundation models (FMs) enable holistic modeling of urban complexity by integrating multimodal data for strategic planning and system-wide optimization. Flow FMs, especially when combined with generative artificial intelligence (GenAI) techniques, accurately capture dynamic urban phenomena under sparse or incomplete data conditions, enhancing generalizability across diverse urban contexts. Environmental and climate FMs provide advanced tools for modeling long-term ecological change, simulating climate scenarios, and informing climate-resilient urban strategies. Geospatial FMs, through large-scale multimodal learning, improve spatial reasoning, urban monitoring, and location-based decision support. The proposed framework, which integrates these categories of FMs, enables an integrated approach to urban intelligence by enhancing urban flow inference, strengthening environmental forecasting and climate resilience, and advancing spatial reasoning, thereby supporting more adaptive, predictive, and sustainability-oriented decision-making in sustainable smart cities. This study advances a computational urban paradigm that positions FMs as critical enablers of technologically advanced sustainable urban development. It provides actionable guidance for urban planners, researchers, policymakers, and technologists seeking to harness FMs to enable more adaptive, sustainable smart city systems.