INNERVATE

Germany plays a leading role in the automotive industry, and vehicles have to undergo tests such as the moose test, in which dynamics and driving behaviour are tested, to ensure road safety. The INNERVATE project aims to speed up the approval process through the use of AI technologies. Data collection is optimised and the certifying person is supported by an interactive ‘over-the-shoulder’ learning process to make the process more efficient.

The certification process consists of three main phases. In the first phase, each driving manoeuvre is performed to a high standard according to predefined criteria and collected and evaluated using specialised electronic equipment. This begins with the installation and commissioning of the sensors and data acquisition in the vehicle, followed by the execution and evaluation of individual driving manoeuvres on the test track. Finally, the collected measurement data is analysed comprehensively.

In the second phase, which HBRS concentrates on, the quality of the manoeuvre control is guaranteed. Here we develop a solution for the automatic evaluation of manoeuvre reproduction (statistical approximation of the cost function) and to support the test drivers with information about incompletely recorded parts of the manoeuvres. The aim is to analyse measurement series, identify errors and provide suggestions to achieve data completeness. We will also ensure that even fragmented data is meaningfully included in the final evaluation.