The lecture starts with the initial situation in assisted and automated driving including a short definition of several levels of automation, a description of technical features within this levels and time of introduction on the market. Afterwards public expectations shall be discussed taking into account several stakeholders. Politics and society expect an increase of road traffic safety on the one hand. On the other hand, customers implicate benefit, e.g. increase of comfort, as well as harm, e.g. loss of control or vulnerability by cyber-attacks. In this context it is necessary to touch upon some legal issues, too. Safety and security requirements for highly automated driving including system verification and validation are currently being developed. In a first step, they can be derived from existing regulations and standards, e.g. UNECE regulation or ISO 26262, and from approval procedures for existing vehicle systems, e.g. passive safety systems. But there are some important limitations in this way. On the one hand we have to deal with technical limitations of sensors and algorithms regarding environment recognition and situation awareness in system safety design (keyword: functional deficiencies). On the other hand, we have to approve a huge number of environmental conditions, traffic situations and driver behaviour. It is not possible to cover this variability only by real testing as it is recently done. So virtual testing will play an important role in frame of the approval of systems for highly automated driving. The lecture gives an overview about recent activities in this field and closes with an outlook to further development of safety and security requirements as well as verification and validation procedures for highly and fully automated driving.
展开▼
