Joint Assessment of Automotive Systems Safety and Security Using Bayesian Networks

Modern developments in automotive systems allow them to perform new complex functions, leading to a situation where standardized functional safety and security lifecycles are sometimes not able to meet their goals both in terms of risk reduction and agility. To address the issue, the authors suggest using a model to jointly represent the safety and security of an automotive system. The model is based on a Bayesian Network (BN). The activities performed in the safety and security lifecycles impact both the structure and conditional probabilities of the BN. The network allows assessment of safety and security properties by calculating the likelihood for the relevant claims both during the development and after it. We demonstrate the implementation of our approach with the example of an advanced driver assistance system (ADAS).