Cognitive Awareness of Intelligent Vehicles

The safety of vehicle driving is within the focus of automotive development. In this work, an approach realizing cognitive awareness is introduced for automotive applications. The advantage of the approach is that the underlying Situation-Operator-Modeling (SOM) concept, can be applied to autonomous driving assuming that changes in the real world are to be considered and understood as a sequence of effects modeled by scenes and actions. For the representational level as part of the cognitive awareness approach, the SOM is used to model and structure the complex scenes of the driver-vehicle-environment interactions in general. First, the scene is modeled as a situation consisting of a group of characteristics, and the set of actions which are performed by the driver are specified and modeled as operators. In this work, passing maneuvers on highways are chosen as example. The implementation of the concept of automated supervision to a real vehicle was firstly realized on a real vehicle, from which driving data are sampled and used to build a model of the Driver-Vehicle-Interaction (DVI). The implementation of this model is realized by a Java-Application, which enables the UDP-based communication between sensor system and a user interface. With the integrated prefilters and operator library, the application provides online processing with raw sensor measurements from the database as input. In conclusion the proposed approach realizes technical cognitive awareness. This task- and problem-oriented awareness can be used for advanced technical solutions, like driver supervision or autonomous driving.