Automated Solar Potential and EV Charging Capacity Assessment for Smart Cities: An AI-Driven Geospatial Tool Applied to Commercial Hubs in Baku

As smart cities expand rapidly, there is an urgent need for accurate, scalable information on solar energy potential and the readiness of the electric vehicle charging network. Commonly used feasibility studies often depend on manual surveys and static parameters, which lead to a high potential for error and make the process lengthy and cumbersome for urban energy planning. The research presents "Solar Roof Academic Analyzer," a web-based decision-making tool developed for automating quantitative feasibility parameter extraction in the multicriteria decision-making (MCDM) procedure. This system was built with a web framework using React, OpenStreetMap, and Google Gemini 3.0 AI to allow real-time automated building footprint extraction from OpenStreetMap polygons and the collection of local weather information. Using a practical example of the 28 Mall commercial center in Baku, Azerbaijan, we demonstrated the possibility of determining the roof area, total solar power capacity, and estimated annual energy production in real-time using the developed tool. In the given case, the total area of the roof is 9608 m², where the total usable area is 8167 m². This area allows the installation of 3807 solar panels (450 W) with a total capacity of 1,713.15 kWp and potential annual energy production of 100 MW of 1,897.48 MWh. In addition, the determination of total capacity allowed the estimation of the EV charging potential of 42 charging stations and provided decision makers with the information necessary to plan and optimize urban infrastructure.