Simulating Tomorrow's city: An AI Toolkit for 15-minute city

The increasing complexity of contemporary urban systems challenges traditional planning tools, particularly in the context of the 15-minute city, which requires fine-grained temporal and spatial analysis of accessibility, mobility, and daily urban practices. This poster proposes an integrated artificial intelligence–driven simulation framework designed to support the planning, testing, and optimization of 15-minute cities through a step-by-step, data-informed workflow. The proposed approach combines geospatial diagnostics (GeoAI), AI-assisted scenario generation, long-term land-use and socio-economic forecasting, real-time multi-agent simulation, and system-wide optimization models. By integrating established urban simulation platforms (such as UrbanSim, MATSim, and CityLearn) with AI-based decision-support tools, the framework enables planners to explore strategic scenarios, anticipate long-term impacts, and simulate short-term behavioral dynamics at street and neighborhood scales. Beyond its technical contributions, the poster critically addresses key methodological and ethical challenges, including algorithmic bias, data privacy, and the persistent gap between simulated outcomes and real-world urban behavior. The framework is positioned not as a predictive oracle, but as an augmented planning instrument that enhances human expertise while preserving transparency and accountability in decision-making. This work contributes to emerging research on AI-assisted urbanism, chrono-urban planning, and digital twins for cities, offering a replicable and verifiable methodology for researchers, urban planners, and policy-makers seeking to operationalize the 15-minute city concept in diverse urban contexts.