Project

The overall goal of the BESKID project is the development of two AI systems that enable the performance of fire spread simulations in rail vehicles based on a small amount of experimental data. These scientific approaches aim to significantly simplify the calculation of design fires, thereby enabling their practical application on a larger scale. The use of such detailed simulations allows for the quantification of intrinsic and extrinsic influences, as well as the effects of structural and system-specific features of the vehicle on the fire progression. For example, the impact of ventilation or firefighting systems on the fire progression can be quantified.

Work Packages

• WP 1 - Definition of requirements and scenarios
• WP 2 - Creation of training databases
• WP 3 - AI-based determination of material parameters
• WP 4 - Design of AI-based fire simulations
• WP 5 - Small-scale reference experiments
• WP 6 - Validation experiments at the laboratory scale
• WP 7 - Real fire experiments
• WP 8 - Simulation of a design fire
• WP 9 - Market analysis for spin-off opportunities
• WP 10 - Dissemination

Information and Contact

Funding Agency: BMBF, see also BMBF project outline
Program: Research for Civil Security Announcement: "Artificial Intelligence in Civil Security Research II"
Project Manager: VDI, contact person: Dr. Christian Fenster
Total Funding: 2,2 Mio. Euro
Project Duration: October 2022 to September 2025
Coordinator and Contact Person: Prof. Dr. Lukas Arnold