A new approach towards increasing Cyber-Physical Systems dependability

Cyber-Physical Systems (CPSs) represent a new generation of digital systems, where cyber entities and physical devices cooperate towards a set of common goals. The continuously evolving requirements concerning CPSs have led to the necessity of creating a dependability assurance framework. The present paper introduces a methodology for creating a dependability run-time model of CPSs in a dynamic and evolving context. The new model combines ontology development techniques with a fault forecasting analysis process in order to define a dependability domain ontology, which plays a central role in the CPS architecture and drives various aspects of these systems, especially the ones related to system diagnosis. An ontology-driven CPS for environmental monitoring shows how the proposed approach is able to assure system dependability. The environmental monitoring case study CPS in this paper has an open architecture with a high degree of flexibility and scalability, it is composed of wireless sensors nodes for data acquisition and transfer, and it includes the ontology and specific software packages for the CPS proper operation.