Safety and security co-engineering for highly automated vehicles

Highly automated driving will have a great impact on people’s social life, changing the way we perceive mobility and its actual meaning and how vehicle occupants act while traveling to their desired destinations. Future highly automated vehicles (HAVs) will have to be updated periodically to continuously improve them and to keep up with the enormous development speed of the entire automated driving (AD) ecosystem. The updating process as well as the high connectivity of HAVs lead to a high risk of cybersecurity attacks through all kinds of internal and external electrical interfaces. Through such attacks, information could be stolen or, even worse, the control over vehicles could be assumed. Hence, security directly influences safety of vehicles. Attacks must be mitigated during all stages of the vehicle’s life cycle, including development, operation, maintenance, and disposal, to reduce security risks and, consequently, safety risks. Currently, there is no well-defined and officially accepted approach to combine safety and cybersecurity activities. Both the standards for functional safety and cybersecurity have to be met and taken into account accordingly during the (development) processes. In this paper, well-known safety and security methods in the automotive sector are summarized. Safety and cybersecurity co-analysis and co-design methods are outlined for the automotive sector with a focus on HAVs. Furthermore, these safety, cybersecurity, and co-engineering methods are evaluated in practice using a real vehicle and the first results are shown. The examined vehicle is the mobile test platform SPIDER. This platform enables the testing of components and vehicle functions in real-world situations and under harsh environmental conditions, which is a prerequisite to ensure safety.