Modeling automotive safety mechanisms: A Markovian approach

Cars embed a steadily increasing number of electric and electronic systems. One of the means at hand to enhance the safety of these systems is to reinforce them with so-called safety mechanisms. The ISO 26262 standard discusses at length how to estimate the contribution of these mechanisms to functional safety. These calculations rely however on fault tree models or ad-hoc formulas that are hard to check for completeness and validity. In this article, we propose generic Markov models for electric and electronic systems protected by first and second order safety mechanisms. These models are of a great help to clarify the behavior of these systems as well as to determine the domain of validity of simpler models such the above mentioned fault trees or ad-hoc formulas. Experimental results make it possible to have a better understanding of which parameters really matter in terms of safety.