Towards increased reliability by objectification of Hazard Analysis and Risk Assessment (HARA) of automated automotive systems

Abstract

Hazard Analysis and Risk Assessment (HARA) in various domains like automotive, aviation, and process industry suffers from the issues of validity and reliability. While there has been an increasing appreciation of this subject, there have been limited approaches to overcome these issues. In the automotive domain, HARA is influenced by the ISO 26262 international standard which details functional safety of road vehicles. While ISO 26262 was a major step towards analysing hazards and risks, like other domains, it is also plagued by the issues of reliability. In this paper, the authors discuss the automotive HARA process. While exposing the reliability challenges of the HARA process detailed by the standard, the authors present an approach to overcome the reliability issues. The approach is obtained by creating a rule-set for automotive HARA to determine the Automotive Safety Integrity Level (ASIL) by parametrizing the individual components of an automotive HARA, i.e., severity, exposure and controllability. The initial rule-set was put to test by conducting a workshop involving international functional safety experts as participants in an experiment where rules were provided for severity and controllability ratings. Based on the qualitative results of the experiments, the rule-set was re-calibrated. The proposed HARA approach by the creation of a rule-set demonstrated reduction in variation. However, the caveat lies in the fact that the rule-set needs to be exhaustive or sufficiently explained in order to avoid any degree of subjective interpretation which is a source of variation and unreliability.