Credibility of software-in-the-loop environments for integrated vehicle function validation

Highly Automated Driving (HAD) technology is an enabler for innovation in the automotive industry through the integration of complex vehicle functions for comfort, safety and economy. However, the safety implications of these functions greatly increase the validation requirements as a pre-requisite for their release into the intended environment. Software-in-the-Loop (SiL) environments are emerging as an alternative to traditional testing approaches such as Hardware-in-the-Loop (HiL) systems. This is due to the strong evolution of vehicle simulation technologies as well as the virtualization of electronic control units, vehicle networks, vehicle simulation technologies coupled with the availability of increased computing power. In this paper we provide insights into a novel approach for arguing the credibility of SiL environments for validating the complex functional and non-functional requirements of HAD. These approaches include methodologies to extract test scenarios from real field-data and automatically evaluate their relevance, selection of test cases that validate the credibility of the given SiL environment and a unique metric to indicate the accuracy of the SiL environment in comparison to the reference system. With the help of a HAD use case we then demonstrate the advantages of this novel approach compared to methods currently employed in the automotive industry.