Application and comparison of STPA and functional safety analysis in ACC system

Abstract

Autonomous vehicles are universally considered to be safety-critical systems, and failures in autonomous vehicles may have catastrophic consequences. Safety analysis and hazard identification of this system have become increasingly significant. However, difficulties are associated with analyzing appropriate safety requirements, including functional safety and the safety of the intended functionality (SOTIF). Functional safety analysis (HAZOP/HARA/FTA) and systems theoretic process analysis (STPA) is an efficient technique to identify potential failure hazards or safety requirements in autonomous driving system. In this paper, a technique integrating which STPA and functional safety analysis (HAZOP/HARA/FTA) is proposed to perform failure analysis and hazard identification in adaptive cruise control (ACC) system, which better identify potential failure and hazards, derive safety goals, ASIL level, safety constraints and safety requirements. Compared with functional safety analysis method, STPA can obtain the requirements of avoiding component failure, and it also focuses on component interactions and personnel decision error, which covers a wider range and provides more cognitive safety requirements for the design and analysis of intelligent driving system.