Scenario-based Extended HARA Incorporating Functional Safety & SOTIF for Autonomous Driving : (ESREL2020 PSAM15) ; 01 – 05 November 2020 Venice, Italy

Scenario-based Hazard Analysis and Risk Assessment (HARA) is an effective and realistic way in terms of vehicle automation capability identification (e.g. SAE/NHTSA automation level). For autonomous driving where the system heavily interacts with its environment by means of sensor (fusion) and actuators, hazard consideration has to be expanded from E/E system malfunctioning behavior to also cover Safety Of The Intended Functionality (SOTIF) including cybersecurity. Scenario-based HARA can capture how to analyze hazards considering Functional Safety (FuSa) and SOTIF, but their relation and inter dependencies need to be represented in a systematic way. Moreover, powerful methods like System Theoretic Process Analysis (STPA) are used to identify, define, and analyze hazards but not risk assessments. Therefore, we investigate state-of-the-art on scenario-based HARA and propose an extended HARA that combines FuSa and SOTIF. In particular, we consider (a) functional scenario representation and selection of scenarios (e.g. drawing basic and accident scenarios in HARA) and (b) methods to find the relation and interactions of FuSa and SOTIF at once for each HAD function. Moreover, we execute a scenario-based HARA method with extension by means of demonstrating the process in an applicable case i.e. a transverse guidance assist system and disclose outcome.