Semantic technologies and metadata systematisation for evaluating time series in the context of driving experiments

The design of assistance and automation systems in the automotive domain often has a strong focus on human machine interaction. Therefore test vehicles and simulators play a major role within state-of-the-art development processes. In particular, during assessment/evaluation of developed assistance and automation systems these facilities are used to record extensive data sets (e.g. multivariate time series as well as video and audio data), which describe driver behavior, man-machine interactions, dynamics of real or simulated vehicles, etc. These data has to be explored and analyzed to support further design activities. [1] At the Institute of Transportation Systems of the German Aerospace Center (Deutsches Zentrum fur Luftund Raumfahrt – DLR) e.g. multivariate time series are stored in a relational database. Metadata is used to link additional information to recorded data (e.g. experimental setup, sensor configuration and types of sensors). During the analysis of this data, significant conclusions can arise, that should be associated with measured data or already associated metadata. This newly created information is also expressed as metadata. The main intention is to provide all stakeholders in a development project with a comprehensive view on measured, transformed, and analyzed data as well as to support the flexible integration of further insights.