In order to enable autonomous driving at level 3 or higher, it is essential to manufacture the necessary sensor systems, such as new ultrasonic sensors with improved accuracy, in high quality, large quantities and at low cost. Due to the high requirements on these novel sensor designs, an innovative process control along the entire manufacturing line is aimed at. The goal is to maximize the yield at unchanged cycle time and, at the same time, to keep the manufacturing costs constant.
To achieve this goal, a traceability and machine learning infrastructure shall be developed as part of the ADeUSPro research project. For this purpose, relevant process, quality and meta data must first be identified and collected, and an inline measurement of these data must be realized in real time. Based on this, models are to be developed that monitor product quality and process reliability in an application-oriented manner and support their optimization. The focus here is in particular on the linking of batch and single processes, the associated proper traceability of individual components and the investigation of complex interactions. The infrastructure should be designed in such a way that it can also be transferred to other production lines.