Towards automated safety analysis for architectures of dynamically forming networks of cyber-physical systems

Dynamically forming networks of cyber-physical systems are becoming increasingly widespread in manufacturing, transportation, automotive, avionics and more domains. The emergence of future internet technology and the ambition for ever closer integration of different systems leads to highly collaborative cyber-physical systems. Such cyber-physical systems form networks to provide additional functions, behavior, and benefits the individual systems cannot provide on their own. As safety is a major concern of systems from these domains, there is a need to provide adequate support for safety analyses of these collaborative cyber-physical systems. This support must explicitly consider the dynamically formed networks of cyber-physical systems. This is a challenging task as the configurations of these cyber-physical system networks (i.e. the architecture of the super system the individual system joins) can differ enormously depending on the actual systems joining a cyber-physical system network. Furthermore, the configuration of the network heavily impacts the adaptations performed by the individual systems and thereby impacting the architecture not only of the system network but of all individual systems involved. As existing safety analysis techniques, however, are not meant for supporting such an array of potential system network configurations the individual system will have to be able to cope with at runtime, we propose automated support for safety analysis for these systems that considers the configuration of the system network. Initial evaluation results from the application to industrial case examples show that the proposed support can aid in the detection of safety defects.