H. Sakellaris and P. Bayrasy from Fraunhofer SCAI presented their work on Surrogate Modeling for Thermal Analysis of a High-Speed Electrical Motor using Gaussian Processes. The motivation for their research stemmed from the need for efficient air compressors in fuel cell systems, particularly for applications in high-speed drives in transportation and aviation. The project, named HABICHT, aimed to develop a high-speed drive for fuel cell air compressors, targeting improved performance and innovative thermal management techniques. The team's approach involved breaking down the simulation chain into smaller components, such as electromagnetic design and thermal management of stators and rotors. This method allowed for the use of expert knowledge in specific areas while managing the computational complexity. The innovative cooling concepts included mantle cooling, winding head cooling, and the use of thermally conductive resin. To manage the extensive computational demand, the team employed Gaussian Process-based surrogate models. These models, trained on simulation data, helped predict key thermal parameters like maximum temperatures in various motor components. The surrogate models were particularly beneficial in visualizing temperature profiles and reducing data complexity to manageable core sizes, like maximum temperature values. The surrogate models' reliability was enhanced by incorporating uncertainty analysis, accounting for variations in input parameters like coolant flow rate and resin thermal conductivity. This approach helped in understanding the impact of input uncertainties on the motor's thermal behavior.In conclusion, the Fraunhofer SCAI team successfully demonstrated the use of surrogate modeling in thermal analysis of high-speed electrical motors. This method significantly reduced computational load while providing accurate thermal predictions, crucial for the development of efficient and reliable high-speed drives for fuel cell systems.