Using AI and autonomous drones

to identify and analyze wildfires

 

 

In an attempt to provide early mitigation techniques for wildfires that are both affordable and effective, a swarm of self-coordinating drones for firefighting has been developed.

The drones can automatically detect and investigate fires, relaying all the information to the fire squad. This is made possible by AI technology that combines thermal and optical vision, and was developed in Sheffield at the School of Electrical and Electronic Engineering.
Utilizing swarm technology created by the University of Bristol and overseen by fire and rescue personnel, the drones may then autonomously self-coordinate as initial responders to quickly apply fire retardant onto the fire, keep an eye on the situation, and return to base.

“Autonomous wildfire detection and localization, especially in remote and large geographic areas, with hot weather is a challenging task that has not been fully achieved yet with the current technology. “We have been creating AI and computer vision techniques at the University of Sheffield that can function in a variety of weather scenarios and effectively extract visual information from incoming video data.

“The primary focuses of our work are effective human-UAV collaboration, assessing the reliability of the established solutions, and ensuring the resilience and scalability of the proposed methodologies.”While autonomous single drones have been utilized in some firefighting situations, the technology of swarms offers a greater degree of reliability and efficiency.

“The technology has the potential to revolutionize wildfire fighting operations by providing firefighters with advanced tools for remotely fighting spreading fires,” said Dr. Lyudmila Mihaylova, a professor of signal processing and control at the University of Sheffield’s School of Electrical and Electronic Engineering.

In an attempt to create early mitigation measures for wildfires that are both affordable and successful, Lancashire Fire and Rescue, which in 2018 had to fight a blaze over 18 square kilometers of moorland near Bolton for 41 days, has tested a swarm of self-coordinating drones for firefighting purposes.

In the course of the study, Lancashire Fire & Rescue worked with several highly regarded robotics and artificial intelligence professionals from the Universities of Sheffield and Bristol, as well as Windracers (https://windracers.com/), a British firm that creates autonomous freight planes. Together, they have developed cutting-edge technology that can recognize and extinguish fires on its own before they grow into uncontrollable wildfires.

“The sooner we respond to fires, the less damage they cause to firefighters, communities, infrastructure, and the environment,” stated Justin Johnston, Chief Fire Officer of Lancashire Fire and Rescue Service.

Each Windracer ULTRA self-flying freight aircraft can carry up to 100 kg of fire retardant. These aircraft have also been used to transport packages to the Orkney Islands and gather scientific data in Antarctica. During the summer, they can fly in an autonomous search pattern to keep an eye on dangerous areas; a swarm of drones could be able to cover an area the size of Greece.

In the UK, wildfires have increased in size, frequency, and severity. This tendency is influenced by variables like shifting land use, rising temperatures, drought conditions, and climate change. Over 44,000 wildfires were reported in 2022, a 72% increase from the year before.

supplied by Sheffield University