This project develops an advanced intelligent system to protect livestock from predators using artificial intelligence and deep machine learning. The system combines thermal imaging and conventional cameras with sensors to detect and predict predator movement in real time, regardless of lighting conditions. The key innovation lies in the autonomous activation of preventive measures (acoustic/light) and the use of Edge Computing for fast data processing directly in the field. The device is designed to operate in remote areas thanks to photovoltaic power and solves the problem of limited data through synthetic AI training. The expected output is a prototype and a utility model that will enhance the digital transformation of agriculture.