PROS2E, a Probabilistic Representation of Safety and Security Events for Diagnosis

Abstract

The recent and increasing interconnection of industrial systems with information technologies creates a new paradigm in which new challenges arise. Being able to provide an explanation when accidents and attacks strike the system is one of them. This article presents an event model designed to provide useful and relevant information for diagnosis of safety and security events. We provide a means for system experts to model elementary events that are automatically recombined in complex and complete scenarios. We specifically focus on the ability to represent countermeasures and sequences of events, recurrent problems in the literature. We also introduce a means to precisely measure the wear of components, an important feature for the model to obtain accurate values. We showcase the use of our model for diagnosis purposes on a cyber-physical system testbed.