Model based Systems Engineering is only as powerful as the enabling data foundation it is working on. Data foundation could be improved by applying semantics and big data approaches. The proposed Semantic DataLake and machine learning concept is based on a new data handling (Extract, Transform, Load or ETL) paradigm and will boost Industrial Engineering Data Management and Analytics to the next level. Today's common approach of defining sophisticated conceptual datamodels, as a basic concept of data handling, has found his limitations and seems not valid for solving future data management tasks. Still today standard MBSE is lacking a birds-eye-view, which could be the foundation for an overall evaluation and analysis of progress and maturity. Engineering data and meta data are captured and stored automatically. However, mainly not taken into account for Engineering improvements. Scoping to close the gap between existing sophisticated conceptual datamodels/ontologies and semantically weak DataLakes, promising concepts have emerged the last years. Observing various industrial sectors and even social networks, the concept of “capturing everything” is tackled by applying concepts of semantic DataLakes. One example is the Big Data Europe (BDE) initiative, which created a platform, containing a couple of interesting features. Our proposal is to check transferability and apply following main features on a to be developed big data engineering platform: create a high performing DataLake apply a semantic layer on top of “raw format DataLake” Enable search across all platform data, apply scalable semantic analytics stack and machine learning approaches. The final goal is to augment the model based engineering process by model analyses, based on machine learning. We are confident, that the components developed as part of the BDE initiative, transferred to ESA engineering tasks, merit a more in-depth evaluation and knowledge transfer into the MBSE and Space domain.
