Hazardous scenarios in environments with poor visibility, such as catastrophes in tunnels with dense smoke and intense heat present major challenges to fire and rescue teams. When it becomes too dangerous for humans, mobile robots are increasingly being deployed. A well-known example of the use of robots is in the investigation of the Fukushima atomic power station after the catastrophe there in 2011. Existing robots soon reach their limits in rough environments, however. Cameras and laser scanners are often no longer able to supply reliable evidence e.g. in smoke, dust, fog, rain or snow.
In collaboration with national and international partners, the SmokeBot project of the Department of Real-Time Systems at the Faculty of Electrical Engineering and Computer Science, Leibniz Universität Hannover, is developing a mobile robot that can also be deployed in hostile environments. A combination of innovative sensor and camera technology are integrated in a mobile robot. "The combination of the different sensors is completely new," explains Prof. Dr.-Ing. Bernardo Wagner from the Department of Real-Time Systems.
In June, the robot is to be tested under real-life conditions in the "Brandhaus", a training centre belonging to the fire brigade in Dortmund. The Dortmund fire brigade is one of the partners in the SmokeBot project, together with universities and industrial partners from Sweden, Austria and the UK, as well as the Fraunhofer Institute for High Frequency Physics and Radar Techniques (FHR). SmokeBot is supported as an EU project within the programme Horizon 2020.
"The robot is a prototype," says Prof. Dr.-Ing. Wagner. "Before it can actually be used in accidents in tunnels, in major fires and in bomb disposal, or in poison gas attacks, the software and hardware still have to be optimised for the rough conditions." The first results are promising: the fusion of the different sensors - cameras, laser scanners, depth cameras and radar - is as yet unparalleled. Specially developed rotating radar sensors are being used for the first time in such a system. They are robust in poor visibility and can be deployed when conventional cameras and laser scanners reach their limits. The radar sensors deliver comparatively inaccurate readings, however. The challenge therefore lies in combining the radar signals with usable data from laser scanners and thermal imaging cameras so as to create as accurate a model as possible of the surroundings.
"Electronic noses" are also deployed - gas detectors that issue warnings in time when hazards arise. "A heat shield is also integrated, which unfolds to protect the robot when extreme temperatures occur," explains Bernardo Wagner. In addition, the system data can be combined and compared with the fire brigade's emergency plans and maps.
Note to the editorial office:
For further information please contact Prof. Dr.-Ing. Bernardo Wagner, Faculty of Electrical Engineering and Computer Science, telephone +49 511 762 5516, E-Mail wagner@rts.uni-hannover.de
