Validation and Assessment of AutomationML Models

By Martin Glatz .
Advised by Manuel Wimmer

It has been recognized that during the construction of industrial pants, which involves various engineering disciplines, the used software tools are often heterogeneous without the possibility to transfer data to the next project phase. To address this problem Daimler initiated in cooperation with the universities of Magdeburg and Karlsruhe the development of the data-exchange language AutomationML (AML). AML is a standardized neutral, free to use, XML based data exchange format which uses Computer Aided Engineering eXchange (CAEX) as its top-level format. The usage of CAEX is further refined within AML. Due to the restrictions and extensions in the AML specification a valid standard CAEX structure is not necessarily a valid AML CAEX structure. Therefore, validity checks available for standard CAEX are not sufficient for AML CAEX structures. Beside validity to the formal definition of AML another vital aspect for model quality is the definition and assessment of modelling guidelines and software-metrics, which is already widely used in classical software development. Therefore the overall aim of this thesis is to create a validation and assessment framework for AML CAEX models. The framework on the one hand provides a formal way to validate AML CAEX models not only on the CAEX but also on the refined AML CAEX definitions. On the other hand the framework creates a possibility to assess a model based on modelling guidelines and model-metrics (e.g., naming conventions, maximal model depth, etc).