Policy-aware multi-source remote sensing for automated urban zoning compliance monitoring

Unregulated construction in rapidly urbanizing region(s) disrupts planned development, strains quality and regulatory infrastructure system, and increases public safety risks. Conventional monitoring approaches often lack scalability and sufficient spatial precision, reducing their effectiveness for timely action. This study introduces an automated geospatial framework that operationalizes zoning regulations by integrating digital surface models, digital terrain models, and openly available building footprint datasets. Structural height compliance is evaluated through a rule-based inference engine applied to OSM building data, while a convolutional neural network applied to very high-resolution satellite imagery identifies boundary violations. The framework produces violation layers and interactive GIS-ready maps that support evidence-based regulatory action. The technical contribution lies in a multi-modal data fusion pipeline that merges raster elevation models, vector zoning boundaries, and spectral imagery for accurate and scalable identification of unauthorized construction. The approach strengthens existing geospatial compliance monitoring practices and offers a reproducible method for supporting urban governance in real-world metropolitan environments.