Integrating AI and IoT for Precision Agriculture: A Hybrid Deep Learning Approach to Sustainable Farm Management

Abstract

The research centers on the synergy between Artificial Intelligence (AI) and IoT for precision agriculture, aiming to overcome some of the largest challenges facing agricultural production, including the scarcity of water resources, fertilization and excessive crop application, and a need for more climate-resilient farming. For the two cropping seasons of the field, Randomized Complete Block Design (RCBD) was used, and 15 smart farms and 300 plots were used. Utilizing a mix of 450 IoT-based sensors collecting data from soil moisture sensors, soil temperature sensors, humidity sensors, and drone-based imaging sections, a combined hybrid deep learning model was created (Long Short-Term Memory networks + Random Forest algorithms) to predict crop growth and detect disease. The model has been able to achieve prediction accuracy, precision, recall, and F1 score of 96.4%, 95.8%, 95.2%, and 95.5% with a root mean square error of 0.18. This precision agriculture system that was implemented resulted in an increased yield of 31.2%, water saving of 27.8%, reduced pesticide usage by 21.7%, and reduced fertilizer usage by 24.5%. The model explained 78.2% of the variance in the farm sustainability outcomes, and AI-based decision support explained the largest proportion (standardized β = 0.46, p < 0.001) for the sustainable management of the farm. Using principal component analysis, three more components were identified that accounted for 84.6% of the variance—AI for productivity, IoT for environment monitoring, and sustainable use of resources. The novelty of this work lies in the novelty of the hybrid deep learning architecture of crop growth prediction optimized with disease detection, which enables real-time farm management with data-driven solutions. The findings highlight the potential for AI and IoT in agriculture to transform the sector by enhancing its sustainability, resource efficiency, and environmental footprint.