We are able to consider the various risk factors when considering automotive safety: for example, malfunction, imperfect environmental recognition, and security violation. We have standards to treat those hazard types: ISO 26262 (functional safety), ISO/PAS 21448 (SoTIF) and SAE J3061 (security), respectively. However, those risk factors don't cover all. We can distinguish the factors relating to the collaboration of multiple ADASs or the difficulty of the decision of motion planner[1][2].In this paper, we provide the approach to identify the risk factors through analyzing the interface classes, that are communication between the environment-system, user-system, and system-system. We'll show the result of this analysis, and we believe this is useful for us to identify the hazard of the system from SAE level 2 to level 5.
[1] https://doi.org/10.20485/jsaeijae.10.1_14
[2] Ito, Masao. "Method of Evaluating the Influence Factor of Safety in the Automated Driving System: The Chasm Between SAE Level 2 and Level 3." European Conference on Software Process Improvement. 2018.
