Autonomous Driving and Cybersecurity by Design

Abstract

So far, real-time requirements for the overall autonomous driving (AD) have been addressed only in a few cases. Cybersecurity and real-time capability are usually addressed separately. However, with regard to a justifiable mobility quality, these requirements are in direct interaction with each other. Therefore, as suggested here, it makes sense to consider the provision of a suitable IT infrastructure with cybersecurity, QoS (Quality of Service) and simultaneous real-time IoT capabilities. The early integration of security and real-time by design, as well as the architecture concepts mentioned, are measures that limit development costs, make the solution modular, scalable and thus sustainable. We introduce the adaptive-real-time-manager (ARM), an innovative concept for continuous assessment and optimization of the real-time capability of autonomous driving systems. The paper also proposes a cloud-broker-concept and simulation as essential building blocks to accelerate the integration of the ARM into an autonomous driving system (ADS). Furthermore, we discuss aspects of multisensory data acquisition and processing, addressing the integration of various data sources and their qualities. Finally, we highlight the importance of driveability for autonomous vehicles, emphasizing its role in comfort, safety, and user acceptance.