Research Area Technology - T3:Quality assurance

Data fusion for quality assurance with artificial intelligence

Advisors: Lanza (wbk), Thompson (MMRI)


In the second generation of the IRTG, project T3 will focus intensively on the further development of the multi-sensor system based on the first generation and data evaluation with the help of artificial intelligence. After initially applying the measurements still on SMC, during the second generation, these will then be expanded to include long-fiber reinforced thermoplastics. In this context, special attention will also be paid to the tape bending process and the resulting complex 3D geometries developed by T2. Defects, such as incorrect fiber orientation or bending errors, can occur. For this reason, the multi-sensor system will be extended with further measurement systems, e.g. for airborne noise detection. In order to enable the evaluation of these measurements, the data fusion approaches will be expanded. Furthermore, it has been shown that data evaluation, for example, thermography, produces different patterns for different defects. These defects are expected to be classifiable and could be quantitatively analyzed by means of artificial intelligence. A close cooperation will exist in this generation with the project T2 to analyze the process of tape bending and investigate the related defects. For a better understanding of the process, close cooperation with T1 is aimed at, as well. Furthermore, there will be a detailed focus on the investigation of the effects of the measured defects on the mechanical properties, which will be examined together with C3. A hybrid data-driven and model-based (based on simulations) quality control loop will be developed for this purpose. This requires strong cooperation with D1 and D2.
 

Figure: Shadowing effects on 3D geometries