A CDIO-Based Reform of Advanced Separation Engineering: AI-Supported Curriculum Innovation and Practice

Advanced Separation Engineering requires graduate students to connect separation principles with process design and engineering decision-making. Using action research, this study developed and refined a reform based on the Conceive–Design–Implement–Operate (CDIO) framework across four annual offerings from 2022 to 2025. The reform integrated authentic engineering projects, AI-supported learning resources, digital process simulation, collaborative inquiry, and continuous assessment. Course evaluations from the 2021 pre-reform historical cohort (n = 46) were compared with those from the pooled reform cohorts (n = 254), supplemented by open-ended responses and interviews. The reform cohorts reported higher ratings on all five common course-evaluation items, with differences remaining statistically significant after Holm adjustment and Hedges’ g ranging from 1.29 to 2.11. Overall satisfaction averaged 3.41 in the historical cohort and 4.58 in the reform cohorts. On supplementary single-item self-report questions, 91.7% considered the CDIO reform helpful for developing comprehensive engineering competence, and 76.4% reported that AI-supported resources and digital simulation helped them understand complex separation processes. Student feedback identified software learning demands, group coordination, verification of AI outputs, and inconsistent industry-expert participation as continuing challenges. This four-year action research study provides a practice-informed reference for integrating CDIO projects and technological support in graduate engineering courses, based on students’ reported experiences.