SAFER2
The Sensor and AI Fusion for Enhanced PeRformance and Reliability (SAFER2) project addresses the fusion of dedicated sensors with ML models for consistent monitoring, evaluation and control of complex systems that appear in several areas of DLR. The aim is to combine DLR's existing cross-focus expertise from the fields of sensor technology and machine learning to be brought together to address essential issues at the intersection. The project spans five demonstrators each focused on challenges of specific DLR research fields: turbo-machinery, AI mobility, wind turbine blade design, aircraft icing detection, and non-intrusive structural damage detection.
As the DLR Institute of Data Science, we are involved in one of the demonstrators, the Transonic Grid Wind Tunnel. Machine learning methods will be used to identify sensor failures in (near) real time and to recognise, evaluate and compensate for erroneous and/or grossly unexpected measurement data during experiments in the wind tunnel. In addition to gaining knowledge, the aim is to achieve greater test reliability, particularly in the area of turbo-machinery, and test stand reliability and, as a result, higher test stand availability.