The main focus of the research group entitled "Data-based material design" is on structural analyses with large-scale facilities, the development of structural models using computer simulation and modelling methods, and the analysis of chemical bonds in oxidic structures and their influence on the chemical and physical properties of ceramic materials. By applying machine learning and deep learning methods, new procedures for the digitalization of materials research are developed and the design of digital twins for fuel cells, gas separation membranes, solid state batteries and thermal protection layers is supported.