Customer centricity gains increasing importance in vehicle development. With respect to durability testing and approval of prototypes, it causes the requirement for testing procedures which highly comply with the occurring loads during customer operation. In this manner, lightweight construction is enabled and considerable economic advantages are raised by avoiding oversizing as well as undersizing of components. As current state of the art, vehicle manufacturers use special test courses on proving grounds or public roads which must be endured damage-free by the vehicle prototype until a compulsory mileage is reached. Alternatively, the course's loads can be reproduced on a test rig to reduce the required test time. The definition of these test courses is based on long-standing experiences. According to new trends in the automotive sector, as for example electrification and shared mobility, it can be predicted that the customer loads will change in accelerating cycles. Correspondingly, these previous experiences cannot be used offhandedly for such new questions and existing test courses become invalid. To solve the described problem of inflexibility, this article presents a technology for automatic generation of appropriate test courses from customer field data. The developed algorithm transforms customer-oriented load spectra into a test course that reproduces the occurred loads as good as possible. For this purpose, digital map data is combined with measured track load data and consecutive road segments are assembled to a complete test course by heuristic methods. This resulting course requires a minimum driving distance and induces approximately the same loads as the destination load spectra derived from customer field data. The optimisation can be processed simultaneously for several different load channels. Via using automation and heuristics, a good performance arises and enables a fast as well as flexible generation of tailored test courses pursuant to current demands. In contrast to the inflexible test course concept used so far, the described technique represents a good improvement of customer-oriented testing methodology. Therefore, it provides a great novelty value. In the article, the process of an automatic test course design is demonstrated and validated by means of a real example, which shows the practical applicability. The latter is also proved by the fact that the described method has already been used successfully in the research and development of Volkswagen AG.

Mr. Florian Grober, Volkswagen AG, GERMANY; Dr.-Ing. Andreas Janßen, Volkswagen AG, GERMANY; Prof. Dr.-Ing. Ferit Küçükay, TU Braunschweig, GERMANY;