A framework for capturing and formalizing modeling knowledge of model-based design with modeling logs and spoken annotation

With the development of computer-aided engineering (CAE) tools and their libraries in recent years, model-based design (MBD) has become widespread for system design across multiple fields. Properly capturing and managing the modeling knowledge behind skilled engineers' modeling processes makes those processes more explicit, facilitating the sharing and reuse of models. Real-time knowledge management is an initiative that aims to support engineers by simultaneously capturing and formalizing this knowledge with modeling. This study proposes an implementation method based on the framework. We automatically capture modeling operations based on the function call log of a 1D modeling tool in a non-intrusive way for engineers. After that, the operation log is expressed in the graphical issue-based information system (gIBIS) format, a discussion model, via the design operation template. gIBIS includes real-time spoken annotations made by engineers to record the operations’ interpretations and rationales explicitly. In a case study, we used function call logs of OpenModelica, an open-source software for 1D modeling. Also, we defined a mapping rule for OpenModelica and formally translated it into a gIBIS representation. The proposed method was applied to the design of jumping mechanisms, capturing the knowledge hidden in the modeling process of a skilled engineer. As a result, the modeling operations and the associated annotations could be expressed structurally, and detailed logs of the modeling process could be explicitly recorded.