An ontology-based approach for transformer generation in a multi-disciplinary engineering environment

Abstract

Modern industries such as energy, automotive, or chemical accommodate a multiplicity of different engineering areas. To satisfy the fast changing requirements of the market, the different engineering areas need to work together within multi-disciplinary engineering projects. This requires the exchange of information between different tools of each area to enable the collaboration with a maximum of a synergy effects. Each tool has a variety of data models which were originally defined with their domain-specific terms and notations. This results in a scenario with the need of semantic integration which deals with the integration of heterogeneous data. A generic approach to face the challenge of semantic integration of engineering data is the Engineering Knowledge Base (EKB). The EKB stores explicit knowledge about engineering plans in different domains to support the access and the management of models and tools across different engineering areas. The main features are data integration based on mappings between local tools on a domain level, transformation between local tools and the domain level, and advanced applications built on these foundations. The creation of a transformation is complex because the engineers need to develop transformation logic as e.g. source code and need to understand the stored knowledge belonging to the tools. This master thesis proposes the TransformerIDE - a solution to support the engineers in their task of creating quality-assured transformations based on the knowledge stored within the EKB. It offers an integrated development environment (IDE) with functions to create, modify, validate, manage and export transformations. The core of the TransformerIDE is the procedural Transformation Language (TL) based on the Transformation Meta Language (TML). This language is tailored to the requirements of the use case and can be easily extended. It consists of simple functions, e.g., trim, concat, with clearly defined input and output. By combining this functions complex transformations can be created. On top of this language the IDE consists of an Editor, Validator, Code Generator, and Transformer Manager. The editor lifts TL from a pure textual structure to a graphical structure with validation, execution, and export functions. The evaluation of the TransformerIDE with the underlying Transformation Meta Language has shown that the automated creation and validation of the Transformer and the Transformations supports the user in reducing the effort and error rate of transformation design and operation processes.