Ontology-Based Pattern for System Engineering

System engineering is a multi-domain process that encompasses the design, realization, delivery, and management of complex systems or system of systems. The Model-Based System Engineering (MBSE) approach is commonly accepted by the system engineers community that depends up on the creation of centralized models to produce the expected deliverables. Standard metamodels such as UML, SysML, or NMM/NAF are typically used to describe the relevant concepts for these descriptive models. However, there is a need to also use domain specific languages (aka ontologies) to ease the communication between all the system engineering stakeholders. The author proposed an approach in previous works to reconcile the usage of complex but necessary predefined metamodels with dedicated ontologies. This solution speeds upthe creation of model-based documents. However, the implementation of such approach revealed that the modeling users are expecting a solution in-between the frozen metamodel andthe specific ontology approach; a set of predefined modeling features addressing recurrent engineering concerns completed by project specific concerns. Among the recurrent concerns there are the requirement elicitation, the functional analysis, the system interface definitions.... This paper shows how this balance can be addressed through ontology-based patterns developed as modular modeling features blocks. Since these blocks are applied in the context of model-based system engineering we also named them MBSE Enablers. The paper proposes a solution to a new issue raised by this pattern reuse expectations; a dynamic mapping is required between the building blocks and the existing models. The proposed method is based on the category theory which brings a theoretical foundation to ensure models are correctly managed. The global idea of the extended approach is to speed up again the modeling tool customizations letting the system engineers focusing as far as possible on the systems to be designed.