Defining a Framework for the Development and Management of Dependability Cases

Dependability cases are a concept that has primarily emerged from safety cases. A dependability case constitutes a clear, defensible, and traceable argument that a system is acceptably dependable to operate in a given operational context. This includes any requirements that are crucial to the stakeholders’ envisioned system operation. A dependability case can be thought of as a driver communicating assurance about the various facets of a system’s operation. The importance of assuring the achievement of dependability attributes is readily apparent in many examples of systems and in specific System of Systems – a class of systems that exhibits a combination of characteristics such as, complexity, autonomy and geographic dispersion. Establishing a dependability case is not a monolithic process, but involves a number of processes and concepts that contribute in overcoming a number of challenges. These include elicitation and apportionment of dependability requirements, resolution of conflicts between dependability attributes and evolution of case. In this paper we present a dependability case framework. The framework includes three processes that have been proposed to overcome the stated challenges, and it can constitute the basis for further additions on the concept. The concept of dependability cases is supported by two pylons. Description of concepts used and their associations (static representation) and definition of the processes that contribute to the construction of the dependability case. The static representation of the framework is defined in a (domain specific) metamodel. Furthermore the processes that participate in the evolution of the dependability case are described along with their prerequisites and contributions to the dependability case.