Evaluation of NoSQL graph databases for querying and versioning of engineering data in multi-disciplinary engineering environments

Abstract

In the context of large engineering projects the effective and efficient exchange and versioning of information from different engineering disciplines is essential. Semantic data integration approaches provide the necessary means to overcome the gap between heterogeneous local engineering tool concepts and common project-level concepts which enable the mapping of engineering data coming from different disciplines. However, evaluations have shown that ontology-based stores do not satisfy versioning and query performance criteria in case of large multi-disciplinary engineering projects. In this paper, we present an architecture solution that uses ontologies to describe engineering models and a NoSQL graph database to version individuals. We evaluate the software architectures according to quality attributes, i.e., performance and scalability, in the context of an industrial scenarios, and compare it with related approaches. Main results suggest that the architecture outperforms ontology stores and match solutions relying on relational databases.