The problem is that "common approaches“ like test driving (e.g. Google Self-Driving Car Project) and pure simulations are not suffice for this task due to the high safety requirements for automated vehicles as well as the increasing system complexity. For test driving, multiple billion test kilometers are required in order to statistically prove safety. This approach is not suitable from temporal and economic aspects. Pure simulations by contrast can only conditionally be applied due to their limited validity. Furthermore, the artificially created scenarios can only imitate real situations to some extent. Therefore, there are several methods to combine the „common approaches” in order to increase their potential to validate automated driving. Virtual Assessment of Automation in Field Operation (VAAFO) is such an approach, which was developed at FZD. With VAAFO, the automation can be evaluated and typical critical scenarios can be determined.