Evaluating software architectures using ontologies for storing and versioning of engineering data in heterogeneous systems engineering environments

Large systems engineering projects involve the cooperation of various stakeholders from different engineering disciplines. Individual stakeholders apply various tools and related data storage approaches that (a) might hinder seamless interoperability and (b) include limited capability to support data versioning. Project-level concepts enable the mapping of engineering data coming from different disciplines. However, it remains open how to store data on project level that enable flexible and efficient data access from different disciplines in different environments and enable backtracking to previous versions in case of defects and/or human errors. While semantic data integration provides fundamental solutions for bridging semantic gaps between common project-level concepts and the local tool concepts used by each discipline, semantic storages have been developed to query and reason over gathered data rather than versioning frequent instance changes inherent to such engineering projects in distributed and heterogeneous environments. In this paper we evaluate three software architectures using ontologies in different ways and compare selected quality attributes, i.e., performance and scalability, in the context of an industrial scenarios. Main results suggest that architectures relying on a relational database for versioning individuals still outperforms traditional ontology storages.