AI-Driven Fault Detection Using LoRaWAN in Solar-Powered Microgrids for Agricultural Irrigation Systems

This study presents a novel AI-driven fault detection framework tailored for solar-power microgrids used in precision agriculture. The system combines LoRaWAN-enabled wireless sensors with a deep learning model based on Long Short-Term Memory (LSTM) networks to monitor electrical and environmental data in real-time. Unlike traditional threshold-based systems, the proposed method offers adaptive, low-latency anomaly detection with over 95 % classification accuracy. Fault labeling was carried out through manual verification and system-level event logging. Experimental evaluation in a live agricultural testbed demonstrates high accuracy and detection times under 5 seconds. Furthermore, we discuss the feasibility of deploying the model on low-power edge computing devices and outline potential cybersecurity vulnerabilities in LoRa-based infrastructures. This integrated approach supports predictive maintenance, reduces system downtime, and enhances energy efficiency, offering a scalable solution for modern smart farming applications.