Abstract

The development of innovative driver assistance systems requires the evaluation of the predisposed hypotheses such as acceptance and driving safety. For this purpose, the conduction of experiments with end-users as subjects is necessary. Analysis and evaluation are based on the recording of numerous sensor values and system variables. Video, gaze and physiological data are recorded for the analysis of gaze distraction and emotional reactions of subjects to system behaviour. In this paper, a modular data streaming and processing architecture is suggested and a concept for this architecture is defined for consistent data evaluation, which integrates off-the-shelf products for data analysis and evaluation.