KIBIDZ – Risk analysis for rescue workers and buildings in case of fire

Real-time risk analysis for fires in buildings is central to saving lives and limiting damage. Buildings must be viewed holistically. The causes of fires that can be derived from the composition of the smoke gases allow conclusions to be drawn about the maximum fire temperature to be expected. Elaborate fluid mechanics simulations are required to determine smoke propagation in buildings and the temperature distribution in structural components. From the simulation data, it is possible to estimate which escape routes must be avoided at all costs due to the existing smoke gas concentrations or acute danger of collapse as a result of material damage. This requires a suitable model of the building that is constantly updated throughout the entire life cycle, the so-called digital twin. This Digital Fire Building Twin also allows the clarification of issues such as the placement of sensors, the connection of these sensors via IoT protocols in advance of the fire event.

Currently, there is a lack not only of these suitable model formalisms for the digital twin and of suitable analysis methods, but also of suitable and networked sensor technology: on the one hand, cameras must reliably detect fire and smoke developments. This should be done directly in the camera, i.e., smart cameras for fire emerge. This must be supplemented by additional sensor technology for analysing smoke gases or temperature development. It is crucial that this sensor data is networked by means of IoT solutions and condensed into an overall building image, the digital building twin, by comparing it with information such as building data and prior knowledge of fire technology. This condensation into a predictive building twin is done using machine learning methods. With the help of this additional information, false alarms can also be reduced, so that the emergency services can be requested in a more targeted manner and expensive shutdowns of equipment, evacuations, etc. can be minimized.

Project duration: 01.11.2020 to 31.12.2024

The KIBIDZ project is funded by the German Federal Ministry of Defense (BMVg) under the DTEC funding program.

Project Management

Prof. Dr.Ing. habil. Wolfgang Weber
Helmut-Schmidt-Universität |
Universität der Bundeswehr Hamburg
Fakultät für Maschinenbau
Professur für Statik und Dynamik

E-Mail: [email protected]


Letzte Änderung: 23. February 2023