Integration of Relay-Based LPWAN Network and Artificial Intelligence for Wildfires Detection and Prevention

This paper proposes a low cost and scalable concept for early detection of natural and man caused disaster risks. The emphasis is given to early detection and prevention of wildfires, but the system can be successfully used for other risks as well. One of its main characteristics is that integrates a private long-range LPWAN mesh network of intelligent IoT sensors with machine learning models and AI algorithms. The sensors continuously monitor meteorological and biochemical conditions for wildfire ignition, while AI assesses fire risk, activates response mechanisms where possible, and sends alerts to responsible positions. Because of its low cost and simplicity, the system/concept is interesting for developing countries and entities, who are not able to invest substantial resources in the purchase, installation and maintenance of expensive equipment. Here, we present a state of the art in the field in term of the main functional components and requirements, the concept of the proposed system, a pilot example, as well as adequate conclusions and recommendations.