A systematic mapping of semi-formal and formal methods in requirements engineering of industrial Cyber-Physical systems

Abstract

The requirements engineering of Industrial Cyber-Physical Systems is extremely challenging due to large system sizes, component heterogeneity, involvement of multi-discipline stakeholders and machines, and continuous evolution. Formal and semi-formal languages, techniques, tools and frameworks can assist by providing repeatable and rigorous structures for eliciting, specifying, analysing, verifying and maintaining requirements. Various approaches have been proposed, but a contemporary and comprehensive study providing a landscape of the state-of-the-art is currently missing. This article reports a systematic mapping study covering 93 primary studies published between 2009 and October 2020. We categorise surveyed studies by current research directions in the use of semi-formal and formal methods for Requirements Engineering phases for Industrial Cyber-Physical Systems. We also identify gaps in current research and develop a novel conceptual model capturing the relationship between available formalisms and Requirements Engineering activities. We find that extensive work has been carried out on the formal analysis and verification of safety and timings requirements. However, the use of semi-formal notations, works on key phases like requirements elicitation and management, and the adoption of industrial standards are largely missing. Moreover, we find no literature providing methods to handle privacy and trust requirements, which have become critical concerns in this area.