Radioactive, chemical or biological substances are undetectable to humans in threatening situations and difficult to detect with remote sensing. Researchers at the Fraunhofer Institute for Communication, Information Processing and Ergonomics FKIE use specially equipped drones and robots to quickly and precisely localize radioactive sources.
Chemical, biological, radiological, nuclear and explosive substances (CBRNE substances for short) can pose a threat to the public and to emergency services. For example, in 2023, a cesium capsule measuring just a few millimeters in size that had fallen from a truck led to a massive search operation in Australia. The recent increasing number of hybrid attacks and various destabilization attempts have exacerbated the threat situation. Two research departments at Fraunhofer FKIE are therefore working intensively to address the question of how we can use drones (unmanned aerial systems, UAS) and robots (unmanned ground vehicles, UGV) to provide the best possible support to people in these threatening situations. The performance of systems like this has been repeatedly assessed over many years at the European Robotics Hackathon (EnRicH) at the Zwentendorf nuclear power plant and at the European Land Robot Trial (ELROB). These events are both jointly organized by researchers from the Cognitive Mobile Systems department in alternating years. They contribute to testing drones and robots for practical suitability under realistic operating conditions and to further developing them based on the results.
Under contract to the Bundeswehr Research Institute for Protective Technologies and CBRN Protection (WIS), researchers in the Sensor Data and Information Fusion department are developing an UAS that can quickly and precisely identify and localize radioactive sources. An technology demonstrator has already been tested in field trials at the WIS site in Munster. This system can precisely track down a radioactive source to within a few meters in only a few minutes. "The cesium capsule in Australia could only be found after days of searching with handheld detectors. We could have found the radioactive capsule much more quickly using our UAS," says Claudia Bender, a research scientist at Fraunhofer FKIE, who designed the technology demonstrator together with her colleague Torsten Fiolka.
The researchers are particularly specialized in complex data processing, sensor data fusion and automation. The detection process is highly automated and consists of an exploration phase and a search phase. During the exploration phase, the UAS flies over the target area and continuously collects data from the surroundings. As soon as a deviation from the background radiation is detected, the system switches to search mode.
In this mode, the flight path of the drone dynamically adapts to the information it has already accumulated from the surroundings as well as current sensor data. This is done using stochastic methods that estimate the probability of different positions of the source. "Once the pilot launches the drone, it initially follows a fixed flight pattern. As soon as sufficient sensor data is available, the system switches to adaptive search mode, using the accumulated information to independently calculate where the source might be," the researcher explains. "It then generates waypoints until it has localized the hazardous substance and reports the precise position of the source." A spatial heat map depicts the radiation levels over the areas that have been scanned. A probability map can also be used to indicate the cell with the highest probability of containing the radioactive material.
The drone is equipped with a gamma detector that measures radiation levels as well as additional sensors for the detection process. These are supported by electro-optical and infrared cameras, an Intel NUC computer for data processing, an inertial measurement unit (IMU) and a LTE comms module for monitoring the data from the ground. The cameras show the live image viewed by the drone. They can detect objects such as people, buildings and vehicles and can display these on a map with georeferencing. The IMU records the position and movement of the drone in 3D.
The technology demonstrator is the product of research under the HUGIYN project (Highly Automated UAS for Detecting and Identifying y-Emitting Nuclides). In the SLEIPNIR follow-up project (Automated Airborne Detection and Identification Platform for Nuclides and Isotopes from Radioactive Sources), the researchers' goals include increasing the airspeed of the UAS and simultaneously localizing multiple as well as moving nuclides.
In situations that are too hazardous for humans, uncrewed ground vehicles provide support along with drones. Robots equipped with CBRNE sensors and autonomous assistance functions then take over the detection process. The intelligent combination of CBRNE detection sensors, navigation strategies and geodata processing with uncrewed ground vehicles is a key research area for Frank E. Schneider, Deputy Head of the Cognitive Mobile Systems department at Fraunhofer FKIE.
“An official radioactivity monitoring system1 was installed in Germany in preparation for events like Chernobyl involving radioactive material. A grid of sensors at intervals of several kilometers are distributed throughout Germany. These continuously report current radioactive emissions levels to the German Federal Office for Radiation Protection (BfS). We compare the radiation intensity detected by the robot in preliminary reconnaissance in a potentially hazardous area with the normal level reported by the radioactivity monitoring system for this area. We use data fusion to combine the results in a radiation map, the heat map, in order to confirm or rule out a potential threat," Schneider explains the deployment scenario.
A further focus of the Cognitive Mobile Systems department is on developing intelligent navigation and assistive functions for facilitating the control of CBRNE robots. These functions ensure that the robot covers the entire hazard zone and can thus reliably localize the radiation source. Additional measurements by the robot then inform the control center regarding the size of the hazardous area, the type of radioactive material found and the positions or coordinates through which to route the safety corridor. Experts and emergency services can use the data obtained in this way to determine further steps.
Triggered by a mouse click on a live video image, a smart click and grasp system enables the robot's gripper arm to automatically pick up objects and place them in a different location that is also designated by a mouse click. In practical tests, the system was able to autonomously grasp spilled material, check it for radiation, transport it away and place it in a special container. The innovative assistive function even enables complex movements such as opening car doors, meaning that it can even access radioactive material in closed spaces.
Schneider's team is also developing new operating concepts that use special sensors to generate a photorealistic 3D model and a virtual reality environment. This enables the operator to view objects from different angles without having to move the robot. “This essentially gives the operator enhanced situational awareness,” the researcher says. Furthermore, sensors attached to the human hand and arm can transfer the operator's arm movements to the robot. The manipulator can thus be controlled intuitively. "We call this new support function “jacket control.” This also enables emergency personnel who are not trained specialists to control the robot intuitively.”
1The BfS ambient gamma dose rate monitoring network includes roughly 1,700 measuring points distributed throughout German
Fraunhofer-Gesellschaft