An experimental study on UML Modeling errors and their causes in the education of model driven PLC programming

Based on previous studies on programming errors and their causes, the presented study examines UML modeling failures in the area of teaching machine and plant automation that were made by 102 mechanical engineering students and analyzes their reasons. Though poor concentration, misunderstandings and overlooking of key aspects within the task have been mentioned by about one third to one fifth of the subjects, they could not explain the objective modeling performance. Instead, the main reason for the shortcomings in the subjects' UML models seems to be difficulties to “translate” their mental model of the system into an UML model. In particular, structure modeling was quite difficult to handle. Lack of time also was a problem, but resulted primarily from starting with the rather complicated structure modeling part instead of the behavior modeling part.