C3 AI methods for the sustainable, synergetic use of the investigation results of GRK 2250 for the design and optimization of impact protection systems

Doctoral Researcher
Felix Conrad

Principal Investigator
Steffen Ihlenfeldt

in cooperation with 
Viktor Mechtcherine
Manfred Curbach
Michael Kaliske

Project poster
Download PDF

In the GRK 2250 numerous interdisciplinary results from experiments and numerical simulations are generated. Due to the large methodological spectrum, the data are very heterogeneous and available in different formats, information structures and locations. In order to use the resulting data synergistically and sustainably within the framework of the research training group and beyond, a comprehensive, tailor-made concept for scientific data management and data exploitation must be developed.

The development of such a concept and its implementation in the GRK 2250 is the central goal of the PhD project C3. A holistic methodology is to be developed, which will generate overarching scientific added value from the resulting data.

Concept overview of C3

The method analyzes the different research questions with their optimization goals, methods, associated design and logging of experiments and available data. The systematic overview is used to develop further, cross-project research questions in exchange with the GRK 2250/2 PhD projects. These questions should be answered with the available data, if possible without additional experiments. Furthermore, the aspect of design and optimization of applications under changed target parameters shall be taken up by modelling the necessary methodical procedure and assigning suitable application-specific test series. On the basis of the data sets, AI algorithms to be developed should allow scope for parametric adaptations and ensure strong synergy and high efficiency in the utilization of archived and exchanged data. The data management and evaluation concept is to include the multitude of different examination methods as well as material and structural parameters from the different observation levels in the context of relevant application scenarios and to match and finally link them from methodical view and data view.