VERDI research project

The advancing digitalization and automation of many areas of activity through increasingly complex systems not only raises questions about their appropriate use and social acceptance, but must also address their conformity with European fundamental rights and values as well as their trustworthiness. In the automotive sector, these processes are reflected in the development of driver assistance systems, which are taking over ever more extensive driving functions and challenging the usual understanding of legal responsibility, safety and privacy. Building on the previous work of the Styrian research partners (the University of Graz and the Forschungs- und Kompetenzzentrum - Das Virtuelle Fahrzeug Forschungsgesellschaft mbH in two EU projects (TRUESSEC and SCOTT), the aim of VERDI is to research the trustworthiness of highly automated vehicles (SAE Lv. 3) from an interdisciplinary perspective (law, ethics, psychology and sociology).
Trust is a central concept for analyzing the appropriate use of automated systems: Inappropriate trust ("over-trust") can lead to misuse or abuse, too little trust ("under-trust") to rejection, i.e. non-use of a system. The hoped-for benefits of highly automated vehicles can only manifest themselves if users (as well as all other road users involved) approach automation with "calibrated trust" and thus know the limits of the system and in which situations trust is appropriate.
As part of VERDI, a catalog of criteria based on European fundamental rights for the trustworthiness of highly automated vehicles will be developed along the core areas of transparency, privacy, anti-discrimination, autonomy, respect, protection and functionality adapted from the previous projects. The results obtained will contribute to a future standardization process. Furthermore, a reliable survey methodology will be presented that captures trust on multiple levels in order to reliably evaluate future systems.