An approach for managing semantic heterogeneity in Systems of Systems Engineering

Semantic heterogeneity is a significant challenge to integration in Systems of Systems Engineering (SoSE) due the large variety of languages, domains and tools which are used in their construction. In this paper we envision a strategy for managing this heterogeneity by decomposing domain specific languages into their “building block” theories which can be independently analysed, and used as a basis for linking with similar notations. This provides a systematic approach to building a tool-chain which integrates the different theories, methods and tools used in SoSE. Our approach has been piloted on the development of theories enabling machine-supported analysis of SysML models of SoSs. We conclude that the approach has further potential and identify lines of future research, notably in techniques for handling mixed discrete and continuous behaviour, timebands, mobility and model integration in SoSE.

I. Introduction

Systems of Systems Engineering (SoSE) is a collection of techniques that support the development and maintenance of a potentially complex aggregate of independently owned and managed systems that are relied upon to provide an emergent service. The nature of the systems that form this aggregate vary in terms of the domain of application, level of independence, ownership, manageability, etc. This variance potentially leads to a plethora of seemingly incompatible models, methods and techniques, and the effective engineering of SoSs depends on their coordinated use.