Table of Contents
ABSTRACT
Urban expansion and associated land use/land cover (LULC) changes have major implications for environmental sustainability within smart city frameworks. This paper presents an artificial intelligence (AI) based remote sensing (RS) approach to monitor and map complex LULC distributions towards informed sustainable land management. Specifically, we develop a deep neural network model that integrates high-resolution satellite data with sustainability-focused LULC classification schemes. Through a case study in the Bangalore Urban region, we demonstrate approximately 90% classification accuracy in mapping fine-scale LULC types related to climate resilience, food security, biodiversity, and urban green space. Results also quantify spatiotemporal LULC changes from 2000-2023, revealing rapid expansion of built-up areas along peri-urban zones and declines in wetlands/forests. In addition to offering a scalable technique for smart city-focused LULC monitoring, the proposed AI-RS framework also produces data to support policy decisions on ecological conservation zones, sustainable urban growth borders, and climate-adaptive landscapes. The discussion of wider generalizability across various metropolitan locations emphasizes the adaptability and significance of integrated AI-RS tools for environmental planning in the face of global urbanization trends that are happening at a rapid pace. For long-term planning and policies regarding sustainable urban expansion, increased monitoring of LULC redistribution is crucial as Bangalore moves forward with its official “Smart City” program. Nagendra et al. (2013) defines the 5230 km2 Bengaluru municipal area to include an extensive and varied landscape made up of huge peri-urban areas, intense cities, and scattered woods. The number of possibilities offered an optimal study setting for delicate spatial scale AI-RS model development and evaluation.
Keywords: Artificial intelligence (AI), frameworks, remote sensing, smart city, infrastructural development
