Improve Test Quality by Applying a Clustering-based Test Planning Procedure for Customer Experience Vehicle Functions

The automotive industry is undergoing technological change that is not only shaped by electro mobility, but also by the rising importance of customer experience. It puts the customers first and manages their journeys with the help of software. These software-driven customer experience vehicle functions can promote competitive technological advantages for the automotive company. In the ramp-up phase of new car models, testing of electric and electronic customer experience vehicle functions is conducted on prototypes. The functional availability of these functions depends on various requirements. Thus, test planning quickly becomes difficult to conduct manually. In order to optimize the test planning procedure for these functions, we developed a data-driven procedure that automates test planning on test objects. In this paper we introduce a procedure that comprises interaction of hierarchical clustering, weighted parameters and ranking. We evaluated a proof of concept at an Original Equipment Manufacturer and conclude that the procedure leads to higher quality in testing. This use case is a successful application of unsupervised machine learning in the automotive industry.