Résumé:
This thesis presents the design and implementation of an intelligent soil health monitoring system to support sustainable agriculture. In response to the limitations of traditional soil monitoring methods, the proposed solution integrates Internet of Things (IoT) technology, artificial intelligence (AI), and multi-sensor hardware. The system, named TerraSense, enables real-time collection of agro-environmental data (temperature, humidity, pH, nutrients, etc.), wireless transmission via LoRaWAN, and intelligent analysis using an XGBoost classifier to predict plant health status (healthy, moderately stressed, or highly stressed). Field tests in pots, gardens, and wheat fields confirmed the system’s high accuracy, energy autonomy via solar power, and user-friendly web interface. This work contributes to the advancement of affordable, proactive, and ethical precision agriculture.
