2025 ASEE Annual Conference & Exposition

Engineering education is largely grounded in constructivist principles, where learning outcomes are often enhanced by individualized assessments. Research supports the idea that tailored interventions, designed to meet the specific needs of learners, can foster a more personalized learning experience. A key aspect of this is targeted feedback, which plays a vital role in student development. This study presents a strategy that enables instructors in chemical engineering courses to create bespoke problem sets and solutions tailored for their students. Ethical AI use and intellectual property contributions are discussed extensively in the text. The issues considered were (1) bias in AI-generated problem statements; (2) academic integrity and plagiarism; (3) data privacy and student information; (4) openness and explanation; (5) intellectual property and copyright; and most importantly, (6) the general framework for ethical use of AI in engineering education.

This approach leverages Python programming, using a modular problem generation and function-based strategy to adapt textbook problem sets and similar resources. AI is employed to vary problem statements. Policies and guidelines on copyright and intellectual property should be followed during the AI phase. It is important to modify the additional information that gives context to problems. Instructors can assign unique values for each student. Instructors also need to set appropriate value ranges and ensure that the program accurately generates custom problems and solutions.

Using Python, ten computational lab activities were generated for approximately 130 students in chemical engineering courses, including chemical engineering calculations, momentum transfer, and separation processes. Flowcharts of the functions used, along with sample activities, are provided. The assessments of these lab activities are discussed, along with the challenges and opportunities for expanding this method to other chemical engineering courses.

This Python-based method for generating personalized problem sets has proven promising in promoting individualized learning experiences for chemical engineering students. The approach shows considerable potential for application in a broader range of chemical engineering courses. Further studies are recommended to see how bespoke problem sets can improve student outcomes in other disciplines as well.

Keywords: personalized learning; bespoke problem sets; python programming; chemical engineering education; computational lab activities; targeted feedback; modular problem generation; individualized assessments; engineering education technology; AI in Education