58 - Fraunhofer magazine 4.19 “We hope that our technology will prevent atrocities – or, at the very least, stop the terror escalating.” Dr. Markus Müller, Fraunhofer IOSB contaminated with radioactive material. “Our system detects whether someone in a stream of passersby is carrying a radioactive source and then warns security personnel,” explains Dr. Felix Govaers, deputy departmental head for sensor data fusion at Fraunhofer FKIE. The system is made up of a number of components. Gamma spectrometers, installed in the floor or the walls of a railroad station, are able to detect radioactive radiation and even determine the type of isotope. They cannot, however, localize its source. For this purpose, the network of gamma spectrometers is connected to commercial Kinect cameras, which not only provide images of passersby but also calculate their distance from the sensors. Data fusion software is then used to merge all this data, thereby pinpointing the radiation source and assigning it to a specific person. When installed at strategic locations – e.g., entrances to stations and airports, ticket barriers, departure gates, or in public buildings – technology of this kind could supply secu- rity personnel or surveillance systems with information on a potential dirty bomb attack. Crucially, here too, any decision to intervene remains under human control. Fraunhofer FKIE is currently in discussion with the Düsseldorf company innoRIID, which would like to market the system. Defense against dangers from the skies Let’s imagine a major event, such as the New Year’s Eve celebra- tions at Berlin’s Brandenburg Gate. A drone suddenly appears in the skies above. Is it harmless? Or is it an airborne weapon controlled by terrorists and carrying explosives – or, more likely, an incendiary device – into the midst of the huge crowd? This scenario shows just how urgently we require systems capable of identifying and defending against drone attacks. In AMBOS, a project to develop defense systems against unmanned aerial vehicles (UAV) for authorities and organizations with security roles, 12 partners from industry, research and higher education have now devised such a system. Funded by Germany’s Federal Ministry of Education and Research (BMBF), the project was coordinated by Fraunhofer FKIE. research group at Fraunhofer FKIE. “This not only merges all the data from the various sensors but also monitors the quality of data from each individual sensor.” For emergency services charged with ensuring security at major events, the system helps them assess each situation and react appropriately. Were a drone to approach New Year’s Eve revelers at Berlin’s Brandenburg Gate, security would be able to track its position, altitude, speed and direction of flight on a digital map. Moreover, the system also analyzes this data so as to determine the level of actual threat and suggest possible countermeasures. If appropriate, such measures can be implemented at the push of a button by the head of operations. The initial option is to broadcast an interference signal, which blocks any signal received by the drone from its remote controller. Unless the drone is an autonomous UAV, it will then most likely do one of three things, depending on its configuration: hover motionlessly in the air, land on the ground or return to its takeoff position. The second option to counter a drone is to transmit a pulse of high-energy electromagnetic radiation. This will tangle up the drone’s flight controller, thereby causing, amongst other things, the rotors to stop turning – with the result that the drone will fall from the sky like a stone. The third option is to use a net gun to physically bring the drone down. The project was brought to a successful conclusion at the end of June 2019 with an evaluation program held in the German town of Mosbach. Drones were flown in a variety of scenarios toward an area under surveillance. The system was able to recognize the danger and, under the control of the operations leader, undertook appropriate countermeasures. “We’re in discussion with a number of police departments and German air traffic control about the core system, which features sensor data fusion, positional graphics and decision support, and which can be combined with various sensors and defense measures,” says Stuch. It is to be hoped that such technology will soon be helping to limit horror scenarios involving drones. Twitter scans for greater security The system comprises a radar sensor, cameras, a radio reconnaissance sensor to analyze the radio signal used by the remote controller, and acoustic sensors to detect the sound of an approaching drone. Each sensor plays a complementary role: whereas the radar has a range of several kilometers, it requires a line of sight to the drone – as do the cameras; by contrast, the radio reconnaissance and acoustic sensors do not require a line of sight, but the acoustic sensor has a range of max. a couple of hundred meters and is sensitive to ambient noises. “The system’s great strength is its multimodal character and especially its sensor data fusion software,” explains Hans Peter Stuch, project coordinator and head of the AMBOS Another critical arena in the fight against terrorism is social media, which is being increasingly used by extremists for communication purposes and for announcing planned attacks. In the future, security forces will be able to turn to another technology from Fraunhofer FKIE in order to search tweets and other social media postings for keywords and thereby narrow down terrorist suspects. This technology combines an analysis of recorded phone calls with a search engine for social media postings. Once a judge has autho- rized the wiretap, a suspect’s telephone calls are analyzed in a total of three stages: Firstly, calls are scanned to filter out all the intervals in which no speech is to be heard. Secondly,