Early detection of forest fires: When drones read smoke signals
Pilot project by FAU and Fraunhofer IISB testing long-distance drones to monitor forests
Forest fires are no longer only restricted to southern European regions, they are now increasingly also a problem in Germany. The interdisciplinary project Evolonic from FAU and the Fraunhofer Institute for Integrated Systems and Device Technology (IISB) is focusing on developing effective solutions for the early detection of forest fires. Together with the fire services in Erlangen and Nuremberg, they are conducting trials on using long-distance drones for the autonomous surveillance of wooded areas. On August 9, the Bavarian Minister of the Interior Joachim Herrmann and the mayor of Erlangen Dr. Florian Janik visited Erlangen to find out all about the project.
A long-distance drone from Evolonic equipped with sensors for detecting smoke, which it is hoped will be able to detect forest fires at an early stage in future, has been flying over the Mark forest to the north-east of Möhrendorf since July. The trial, which is set to run until the end of August, is aimed at testing the system’s long-term stability and gather experience on how the components behave under various weather conditions. “We are investigating false alarms, communication with the control center at the Nuremberg fire department and the performance of the livestream. Based on our findings, we can fine tune the system and make it more effective at detecting forest fires at an early stage,” explains Tobias Raczok from Evolonic.
The flight zone for the trial flights in Möhrendorf was chosen carefully in consultation with the fire department and forest management in order to rule out the possibility of any danger to air traffic or bystanders. Technical measures have been taken to ensure safety through a redundant propulsion system and communication with the drone. The drone is also fitted with systems that allow it to detect any air traffic and automatically swerve out the way.
The pilot project in Möhrendorf is only the first of two trial runs planned for this year. In September, students will test the system’s accuracy for detecting forest fires during a real fire trial in Saxony Anhalt. In the long term, the plan is for the system to fly over defined forest areas autonomously and frequently, using computer-assisted image recognition to detect smoke at an early stage. When it detects smoke, the drone triggers an alarm and forwards all the necessary information and images to staff at the control center and fire service commanders. This should allow the fire service to take action more rapidly and pinpoint the area to focus on, at the same time as increasing the safety of the fire services by giving them a better overview of the situation.
Minister of the Interior Joachim Herrmann enthused about the project: “This mixture of traditional fire service skills and cutting-edge modern technical support will enable us to raise our high protection standards in Bavaria even further. That is extremely important. Forest fires are becoming more dynamic and more intense. It is all the more important that we discover fires quickly to allow our fire services to tackle the blaze from the outset.”
Erlangen’s mayor Dr. Florian Janik adds: “Weeks of dry weather and forest fires throughout Germany this summer have opened our eyes once again to the importance of tackling forest fires in a truly efficient way. Once again, the University and the Fraunhofer Institut are keeping their finger on the pulse and striving to find application-based solutions to problems facing us today.”
Prof. Dr. Martin März, Director of Fraunhofer IISB and holder of the Chair of Power Electronics, Power Engineering and Energy Technologies at FAU, has been in charge of the interdisciplinary student project Evolonic since it was launched in 2008: “Over all the years, students in the team have focused on topical issues of relevance to the future. The impressive dedication and enthusiasm students bring to the project, their ability to manage their team and their projects completely independently and to solve highly complex interdisciplinary tasks inspire my greatest respect. I am particularly pleased that students in the Evolonic team are not only focused on the sporting aspect of their project, although they have already won first place in two international competitions. They also want to work towards providing a benefit for society, as the impressive project we are introducing today for detecting forest fires quite clearly shows.”
The fire department in Erlangen has been behind this innovative project from the outset: “The fire department in Erlangen has been preparing intensively for vegetation and forest fires, providing training and purchasing the relevant equipment. It is extremely important to detect and pinpoint forest or vegetation fires as early and as accurately as possible in order for us to be able to start tackling the blaze as soon as we possibly can,” explains city fire inspector Friedhelm Weidinger.
Detection based on findings from FAU’s Innovation Lab
The foundations for the detection of forest fires were laid in summer semester 2022 during collaboration between Evolonic, the fire department of the city of Erlangen and the Innovation Lab for Wearable and Ubiquitous Computing, a teaching unit at the Chair of Machine Learning and Data Analytics at FAU. The aim of the innovation lab is for students to get involved in solving practical issues from system and software engineering in conjunction with wearables and IoT devices in a creative and innovative way.
During the course, the students developed a concept for using drones and a sensor system based on image recognition to detect forest fires at an early stage. Following on from on these results, the students of the interdisciplinary project Evolonic at FAU and the Fraunhofer IISB developed the system until it was ready for application and integrated it into their drone. During the Digital Tech Fellows Program run by the FAU Digital Tech Academy (DTA), the students also laid the groundwork for launching a start-up.
Phone: + 49 151/22332499