A metamodel to annotate knowledge based engineering codes as enterprise knowledge resources

The encoding of Knowledge Based Engineering (KBE) software applications is becoming a prominent tool for the automation of knowledge intensive tasks carried out using Computer Aided Design (CAD) technology. However, limitations exist on the ability to manage the engineering knowledge models embedded in these executable KBE applications. This research proposes a metamodel to annotate encoded KBE applications. Resulting from the annotation, XKMs become explicit knowledge resources whose content can be better accessed and managed. The attachment of metadata to data sets in enterprise repositories is a necessary step to identify and index them so they can be queried, browsed and changed. The sophistication of metadata models for these data “items” ranges from the simple indexing using numbers to more sophisticated representations describing their context information (i.e. author, creation date, etc.), their internal structure and their content. Current engineering data repositories like Product Data Management and Product Lifecycle Management systems offer predefined metamodels to annotate a range of engineering data items including CAD files or special types of documents. At the moment, there is no metadata model specifically designed to annotate KBE codes. In this situation, an undifferentiated metadata model needs to be used for XKMs. However, in this case the only information retained by the system about them would be context metadata. Once an instance of the metadata is attached to an XKM, it can be used as its identifier within an enterprise data repository. The proposed metamodel contains abstract entities to annotate XKMs. The resulting descriptive model for an XKM pays attention to its internal structure and its operation at different levels of granularity. The particular design of the proposed metamodel positions it at a level of abstraction between non executable domain knowledge models and executable KBE applications. This design choice is made to support the use of the metadata not only as an informative model but also as an executable one. The achievement of this target is becoming possible through the emergence of semantic modelling standards that allow the description of data models independently from the language of implementation. Using this approach, the generation of code and metadata is made automatically using mapping rules resulting from the semantic agreement between models and specific syntax rules. The immediate application of the developed metamodel is to annotate XKMs within PLM systems. The approach shall contribute not only to systematically store instances of XKMs but also to manage the lifecycle of the engineering knowledge encoded within them. The proposed representation provides a more comprehensive approach for non KBE language experts to understand the code. On this basis, the change on the metamodels can be automatically traced back to the code and vice-versa. During the research, evidence has been gathered from the community of KBE technology users and vendors on the need to support this research effort. In the long term, the research contributes to the use of PLM systems as a platform for engineering knowledge management.