Drone swarms follow the flames
Autonomous drone swarms are used in the FireMan project. Cameras on the drones detect smoke and can monitor the progress of the fire. The collected data is processed with onboard computers, and key data is sent to the command centre, which makes forecasts on the progress of the fire and optimises the suppression measures.
The project makes use the controlled burning used by Metsähallitus for forest restoration and collects data for training drones during them. The drones record the fires before ignition and during ignition and burning, and the data is used to develop algorithms that can reliably predict the progress of smoke and the fire. In summer 2024, the project researchers took part in several controlled burnings, and data from these is used in further drone training.
The FireMan project is headed by Research Professor Eija Honkavaara from the FGI. Honkavaara specialises in drone research.
‘In the tests done in the summer, we successfully tested a new AI method with which a smart camera mounted on a drone detected a small restoration fire from about four kilometres' distance by using real-time object identification’ says Honkavaara about the recent development.
Digital twins help with fire suppression
The research makes use of accurate digital models called ‘digital twins’ to predict the spread of fire. A digital twin of the terrain is created with open geospatial data to which the real-time location data of drones is combined. With a digital twin, the chief fire officer can create a forecast on the direction the fire is likely to spread and assess the effectiveness of different firefighting techniques, for example.
‘We have created a situational awareness center in the cloud of the IT Center for Science (CSC), to which the drones send data. The data is processed in real time to become geospatial data compatible with maps. In this way, data is produced about the progress of the fire. In September, we will test the cooperation in a swarm consisting of several drones in collecting situational awareness data’, says Honkavaara.
More development is still needed for using drones in suppressing forest fires. The drones require a strong network connection for data transmission, which usually is not available in a forest. To solve this challenge, the researchers are studying whether a separate network could be transported to fire areas, for example. To enable the drones to support fire suppression as efficiently as possible, their data processing capabilities are also being developed. The drones process the data on site, and only transmit the key data to the control centre, which reduces the volume of data to be transferred.
Drone airspace management is also developed as part of the project to prevent collision with other air traffic, such as airplanes and helicopters. The drones must also be able to safely fly as a swarm.
More information
Eija Honkavaara, Research Professor, FGI, +358 29 531 4716, firstname.lastname@nls.fi
The FireMan project is funded under the EU’s Recovery and Resilience Facility that promotes the green and digital transitions through research. The funding is coordinated by the Academy of Finland. The project is a collaboration between the FGI, the universities of Jyväskylä and Oulu and VTT Technical Research Centre of Finland.