AI-Driven Optimization for Urban and Vertical Agriculture Planning: A Multi-Model Approach

Urban agriculture has emerged as a critical strategy for enhancing food security, mitigating urban heat islands, and promoting community well-being in densely populated areas. However, the complexity of urban environments poses significant challenges for effective planning and implementation. This paper presents an AI-driven framework to optimize urban and vertical agriculture planning by leveraging advanced machine learning models, including Artificial Neural Networks (ANN), Spiking Neural Networks (SNN), and Large Language Models (LLM). The proposed framework integrates diverse dataset that includes as socioeconomic data, geographic information, and urban zoning regulations to provide actionable insights for decision-makers. The Spiking Neural Network model demon-strated superior predictive accuracy in identifying optimal sites for urban agriculture by effectively handling complex patterns and temporal dynamics in the data. Additionally, the integration of an LLM-powered chatbot into a user-friendly web application enhances interactivity and supports real-time decision-making, guiding users through the prediction process with context-specific recommendations. Experimental results validate the robustness and scalability of the framework across various urban settings, demonstrating its potential to transform urban agriculture practices by providing precise, data-driven recommendations. The findings of this study highlight the transformative potential of AI in urban planning and agriculture, offering a novel approach to fostering sustainable urban development and food security. Future research will focus on expanding the dataset, refining model performance, and enhancing the application's capabilities to support more complex user queries.