For the last 15 years, safety has been a key topic for Fraunhofer Institute for Experimental Software Engineering IESE. During this time, the institute has become a leading research institution for model-based safety engineering and runtime safety assurance. Particularly with respect to the latter, Fraunhofer IESE is conducting visionary research in numerous projects, spearheading this field of research. Fraunhofer IESE has participated, often with a leading role regarding the topic of safety, in many large research projects such as SPES, SPES-XT, CREST, ARAMIS I and II, CESAR, MBAT, CRYSTAL, EMC², SECREDAS, DEIS, and V&V Methoden. Among the results are a three-digit number of publications, numerous talks, as well as contributions to working groups, standards, and roadmaps. At the same time, Fraunhofer IESE has been engaged in numerous bilateral industry collaborations ranging from large-scale safety engineering via tool development to the investigation of innovative safety concepts for automated driving.
Runtime safety assurance enabled by Digital Dependability Twins
Visions of future transportation systems are enabled by a trinity of technological characteristics – automation, interconnection, and AI. These traits are difficult to handle based on established safety engineering approaches and standards, so new approaches need to be devised. Runtime safety assurance is increasingly being accepted by key stakeholders as the corresponding silver bullet. The means for such assurance can range from easy checks to sophisticated and intelligent reasoning. It is expected that in parallel to increasing degrees of adoption of the aforementioned key traits, systems will need to be empowered to assume increasingly comprehensive runtime assurance responsibilities [ref]. In icri-save, IESE strives to advance previous as well as related work on runtime safety models [ref] to establish a sound safety backbone for runtime assurance in collaborative automated vehicles. We envision a model-based full-lifecycle approach where Digital Dependability Twins are systematically engineered in accordance with state-of-the-art safety engineering practices powered by dedicated tool support (ref). Development time integration scenarios along the value chain shall be supported as well as the transformation of the Digital Dependability Twins into an adequate runtime representation. The anticipated main contributions in icri-save are: 1.) tailoring of the concepts to automated vehicle use cases and advancement of the “safety intelligence” of digital dependability twins; 2.) investigation of corresponding dynamic risk assessment capabilities; and 3.) interfacing and integration of other icri-save approaches with Digital Dependability Twins.