Abstract:

Generative AI is increasingly used to assist instructors in the preparation of teaching material. In engineering education, some topics require visual representations and repeated practice in order to consolidate understanding. Petri nets are a common modelling tool in automation and control, particularly for describing discrete event systems.

In the undergraduate course Systems and Control Engineering, generative AI was explored as a support tool for producing exercises related to Petri nets. The language model was able to generate simple Petri net diagrams using ASCII representations. These diagrams were incorporated into multiple-choice questions designed to evaluate students’ understanding of the laboratory practices. The questions asked students to interpret the behaviour of the net, determine which transitions were enabled, or analyse the distribution of tokens across places.

The generated material was used in short questionnaires applied at the end of laboratory sessions. The purpose of these questionnaires was to evaluate whether students had understood the concepts addressed during the practical activities. In addition to Petri net exercises, the model also generated questions covering other topics discussed during the laboratory sessions. All generated material was reviewed by the instructor before being used in the course.

Students later completed a short survey based on Likert-scale questions. The survey explored the clarity of the exercises and the usefulness of the material for understanding laboratory content.

Keywords:

Generative AI, multiple-choice questions, Petri nets, engineering education, discrete event systems, laboratory assessment.