Integrating Digital Twin and Swarm-Based UAV Systems for Autonomous Wildfire Prevention and Detection

The increasing number of wildfires throughout the world has resulted in a global loss of forest areas, which are important for sustaining life. To detect and prevent these disasters, a system comprised of Unmanned Aerial Vehicles, their Digital Twins and a Boids-based autonomous patrolling algorithm is proposed. By replicating an Unmanned Aerial Vehicle and its camera specifications, or by use of a real drone, it is possible to simulate wildfire detection and prevention missions in a virtual map. These vehicles can be controlled manually or set to move autonomously as a swarm by following an algorithm based on the Boids program, where they operate with a bird flock-like movement and avoid colliding with other neighbouring vehicles. Additionally, the use of real-time detection of objects of interest through Computer Vision is also an option. A first solution built upon the proposed system was implemented as a web application for access through a considerable number of devices and validation for the system achieved promising results.