Curriculum reform and practice of a graduate-level course based on the CDIO engineering education model: a case study of Pattern recognition

This paper examines the current state of graduate education in the discipline of Control Science and Engineering, as well as the Control Engineering professional master’s program at our university. It identifies key shortcomings in graduate-level courses, particularly regarding the integration of theory and practice, the cultivation of research capabilities, and academic writing training. To address these issues, this paper proposes and explores concrete reform measures and practices based on the Conceive-Design-Implement-Operate (CDIO) engineering education model, taking the Pattern Recognition course as a case study. The reform was implemented on 235 graduate students, with its effectiveness evaluated by a mixed method of course assessments and classroom observations. Key results show that the proportion of students achieving excellent performance, which is above 90 points, increased from 19.1% in the 2022–2023 cohort (i.e. before the reform) to 48.3% in the 2023–2024 cohort (i.e. after the reform), the average score rose from 86.9 to 88.9, and classroom observations consistently indicated high engagement and participation. In addition, over ten course papers were recommended for presentation at university-level academic conferences. The research findings suggest that the CDIO-based teaching approach has achieved remarkable results in enhancing students’ practical competence, research capabilities, and academic writing skills. This CDIO-integrated teaching paradigm can serve as a reference for teaching reform in other engineering disciplines at the university, as well as in engineering disciplines at other universities that aim to strengthen the connection between theoretical learning and engineering practice.