AI-Driven Framework For Smart Farming: Enhancing Crop Productivity Through Climate-Aware Decision Support

Abstract

Smart farming uses advanced technologies to optimize agricultural productivity under changing climates. This study presents an AI-driven framework that integrates IoT sensor networks, climate data analytics, and machine learning models to support climate-aware decision-making for crop management. The proposed architecture combines real-time field data (e.g. soil moisture, temperature, and satellite imagery) with weather forecasts and historical climate records to predict crop yield and irrigation needs. We implement and evaluate predictive models on a representative dataset (including synthetic climate–crop data), demonstrating that adding climate variables markedly improves prediction accuracy. For example, a baseline model using only soil factors achieves low accuracy (R² ≈0.21), whereas a climate-enhanced model attains R² ≈0.72. The framework generates actionable recommendations (e.g. adaptive irrigation schedules, fertilizer adjustments) that help farmers mitigate climate risks and boost yields. Experimental results show the effectiveness of climate-informed AI models for precision agriculture. This work provides a comprehensive architecture and case study for climate-resilient smart farming, highlighting the importance of integrating AI with climate analytics for sustainable productivity.