An integrated architecture for the development and assessment of ADAS

Advanced Driver Assistant Systems act, by definition in natural, often poorly structured, environments and are supposed to closely interact with human operators. Both, natural environments as well as human behaviour have no inherent metric and can not be modelled/measured in the classical way physically plausibly behaving systems are described. This makes the development of a new methodology and especially the development of an experimental environment necessary, reflecting all constraints of the task to be observed, incorporating the influence of each single component and the generation of variations over scenes and behaviours. In this contribution we focus on the development of an integrated architecture based on a simulated reality framework incorporating simple behavioural models of autonomous acting humans, physical plausible behaving scene objects, a dynamic scene generator and an advanced recording and analysing system, operating on-line on all data streams. The architecture allows to develop, assess and benchmark embedded components, as well as whole ADAS.