Robotic aircraft can also get to places where other vehicles are not or only very difficult. Researchers at Johannes Kepler University in Linz, led by Oliver Bimber, have developed a new drone system consisting of a commercially available octocopter and a thermal imager. The advantage of this solution is that the individual shots of the thermal camera can be combined into a single integrated image, which is then analyzed by artificial intelligence. Bimber said the new system is based on a measurement principle called Synthetic Aperture Sensing used in radio telescopes, which maps objects remotely based on their reflections. The person skilled in the art added that the result obtained from the recorded measurement data can be further enhanced by connecting several sensors. When a robotic plane flies over a larger forest, hundreds of photos are captured with a lens of just a few mm, and then charged as if the lenses were several meters in diameter.
While in traditional cases the recognition rate is below 25 percent, in this method it is above 90 percent. The method can be used to search for missing persons as well as to check border sections or to determine the location of fire nests in forests. Unmanned aerial vehicles are ideal for agricultural or wildlife monitoring purposes. There is only one problem at the moment: the method has only been integrated into battery-powered drones, while a traditional robotic aircraft can be in the air for several hours and cover hundreds of miles. It would also be important to amend aviation law so that rescue operations with unmanned aircraft do not require special permits each time. According to Bimber, otherwise the system will be mature in 2-5 years and can be used under real conditions.