Decentralized, Secure and Cognitive Architecture for Automotive CyberPhysical System of Systems

—In this paper, a holistic, decentralized and cognitive design and operation approach that supports CyberPhysical System of Systems (CPSoS) autonomic (without human intervention) behavior, making such systems aware of their physical and cyber environment and reacting to it accordingly so that they constantly match their intended purpose. We propose using a model based design approach to describe a CPSoS in a holistic and abstract way and to allocate computational power/resources to the CPS end devices of the System by determining and generating autonomously what cyber-physical processes will be handled by a device’s each heterogenous component (processor cores, GPUs, FPGA fabric) and software components (software stacks). The proposed solution uses this methodology to strengthen reliability, fault tolerance and security at system level but also to support CPS designs that work in a decentralized way, collaboratively, in an equilibrium, by sharing tasks and data with minimal central intervention. Also, the proposed system supports the interaction of the CPSoS with their human users/operators through extended reality modules (AR glasses, haptics interfaces) to increase human situational awareness but also to include human behavior in the CPSoS design and operation phase. The proposal key points are highlighted in this paper and their usage in an automotive use case that involves connected cars is presented.