Automotive Electronic Systems

• Applicant: Professor Dr.-Ing. Markus Maurer
• Subject Area: Computer Architecture, Embedded and Massively Parallel Systems
• Term: from 2013 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 206480214

Nowadays, automated driving is widely discussed by not only the research community, but also the automotive industry. With this growing interest in vehicle automation, extensive research is currently conducted in the field of functional safety for automated driving functions. So far, however, most research groups consider the driver to be the ultimate fallback solution in case of a malfunctioning system. When designing fully or partially automated vehicles, the driver has to be considered to reside "out of the loop", meaning that he/she is not responsible for handling the vehicle in critical driving situations. Therefore, the safety requirements for the design of embedded systems for automated vehicles are considerably higher than those for state-of-the-art driver assistance- or partially automated vehicle systems. In cases of malfunctions or changing external conditions, e.g. signal losses, the system itself needs to react in such a way that a safe operation of the vehicle can be guaranteed. Thus dependability and availability of embedded systems are key issues for the construction of automated vehicles. As these two aspects are a main focus of Controlling Concurrent Change (CCC), sub-project C2 sees great potential in applying the CCC mechanisms to vehicle guidance systems. Hence C2 will focus on three goals during the second project phase: The first goal is to show the scalability of the CCC approach with regard to complex embedded systems for automated driving. In this context the applications and algorithms developed in first phase of CCC will be extended resulting in a complex overall system which integrates state-of-the-art environment perception algorithms. At this point, the extension of the CCC project to open networks offers great potential to increase the reliability and integrity of (sensor) data shared in a group of vehicles. The second goal focuses on the methodic research of a holistic approach to self-representation of automated vehicles. Two main challenges will be addressed here: C2 will develop methods to model the current abilities and skills of the system. In addition to this system model, suitable performance metrics will be developed for the quality of service of executed functions as well as for the integrity of sensor data, particularly for sensors such as cameras, multi-layer laser scanners and radar. The third goal is to make use of the aforementioned methods in the CCC framework. With its flexible reconfiguration mechanisms, the CCC framework shows great potential to provide a platform able to integrate redundancy mechanisms, degradation strategies and self-healing for embedded systems. The models and performance metrics developed for the assessment of the system's abilities will provide a basis for the selection of suitable fallback strategies in case of system failures.

• DFG Programme: Research Units
• Subproject of: FOR 1800: Controlling Concurrent Change (CCC)