Interactive Learning Engineering Concepts in AutomationML

In the era of digitization, industrial engineering process generates a vast amount of data. To organize, store, and exchange such data, dedicated international standards are developed, including the XML-based data format AutomationML (AML). AML is standardized as IEC 62714 and is recommended for managing data flow in continuous engineering. Nevertheless, engineering data is inherently heterogeneous, and the harmonization of various data sources presents a bottleneck in the vision of an integrated engineering toolchain. In this paper, we present AMLLearner - a semi-automated system for learning engineering concepts from AML data. The results of learning are formal ontological definitions of engineering artifacts that naturally serve as knowledge for information integration. Based on the previous works on learning in AML, this paper emphasizes the involvement of end users, i.e. domain experts, for providing suggestions and feedback to the learner. To show the interactivity of AMLLearner, we discuss its characteristics regarding the recent study on interactive machine learning systems.