A longitudinal evaluation of a four-level practical engineering education framework based on OBE-CDIO principles

Abstract

The transition of engineering education toward competency-oriented training is critical for modern intelligent manufacturing, yet empirical models for resource-constrained local universities remain underexplored. Using a national-level automotive engineering experimental teaching demonstration center in China as a case, this study designed and evaluated a progressive four-level practical teaching system. The system integrates Outcome-Based Education (OBE) and the Conceive-Design-Implement-Operate (CDIO) approach and is supported by a Teaching-Learning-Competition-Innovation (TLCI) linkage mechanism. To evaluate its operation, this study conducted a design-based longitudinal investigation using graduate administrative tracking, competency self-assessment, and employer feedback. Multi-source descriptive evidence showed that, during implementation, the graduate employment rate remained around 95% for four consecutive years, and job-major alignment was approximately 80.0%. In addition, the average starting salary of the 2023 cohort increased by 42% compared with the 2020 cohort, students reported relatively high perceived engineering problem-solving ability, and national competition performance also showed positive trends. These findings suggest that combining industry practice with a competition-driven closed-loop mechanism may provide practical reference for improving the quality of engineering talent cultivation and offer an implementable pathway for educational reform in resource-constrained universities.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-56297-9.

Introduction

The rapid development of technologies such as Industry 4.0, renewable energy, and intelligent manufacturing has created new challenges for engineering education¹. Contemporary engineers not only need solid disciplinary knowledge but also require interdisciplinary integration skills to solve real-world engineering problems with blurred boundaries and complex contexts². Accordingly, engineering education in many countries has gradually shifted from a knowledge-transmission paradigm to a competency-oriented framework³. Practical teaching plays a key role in this transition by helping bridge the gap between theoretical learning and professional practice⁴.

Previous studies have extensively examined pedagogical innovation at the curriculum level. For example, Dziuban et al. described blended learning, namely the combination of online and face-to-face instruction, as the “new normal” in higher education⁵. Related studies have shown that, when information and communication technologies are effectively integrated, blended teaching can improve students’ learning performance and satisfaction⁶. Systematic reviews of instructional organization and support strategies in engineering education during the COVID-19 pandemic also indicate that engineering educators maintained a certain degree of teaching quality and learning continuity by adjusting teaching strategies, innovating course design, and optimizing assessment methods^7,8. These studies suggest that the purposeful integration of digital tools with traditional teaching has become an important component of modern engineering talent cultivation.

Beyond micro-level pedagogical innovation, Outcome-Based Education (OBE) and the Conceive-Design-Implement-Operate (CDIO) model have become important macro-level frameworks for engineering education reform and quality assurance⁹. Both have been widely adopted in European and American universities, promoting systematic reforms centered on competency mapping, curriculum structure, and learning experience¹⁰. Prior research also emphasizes that successful engineering education reform depends not only on course design itself but also on sustained faculty development, operational evaluation mechanisms, and evidence-based feedback improvement processes¹¹.

At the implementation level, Project-Based Learning (PBL) and experiential learning models have been shown to facilitate knowledge transfer, interdisciplinary collaboration, and complex problem solving¹². Incorporating sustainability competencies and authentic engineering challenges into training objectives can also enhance students’ sense of social responsibility and systems thinking¹³. Meanwhile, virtual simulation-supported practical activities can improve students’ preparedness and engagement under certain conditions¹⁴. From the perspective of learning motivation, support strategies based on frameworks such as self-determination theory help explain and enhance student engagement and sustained participation in project-based learning environments¹⁵.

Although prior studies have provided an important foundation for OBE-CDIO, practical teaching, project-based learning, and digital instruction, two gaps remain in the existing literature. First, systematic and longitudinal evidence remains limited. Many studies focus on individual courses, local teaching components, or short-term interventions, with limited explanation of how to construct a coherent, multi-level practical teaching system across the entire undergraduate training cycle. They also rarely link the operation of practical teaching platforms with graduates’ career development outcomes. Second, localized implementation research remains insufficient. Existing empirical studies are more often drawn from resource-rich universities or mature engineering education systems, while discussion of how resource-constrained local universities can continuously implement and iterate OBE-CDIO principles remains limited¹⁶. Rather than repeating prior course-level or single-intervention studies, this study builds on this literature by combining long-term institutional development, practical platform operation, and graduate outcome tracking to further analyze the operating mechanism and conditional transferability of an OBE-CDIO practical teaching system in resource-constrained local universities.

In response to industrial transformation, the Ministry of Education of China launched the “New Engineering” reform agenda, with key measures including curriculum optimization and deepened industry-education integration¹⁷. Within this institutional framework, national experimental teaching demonstration centers have become important carriers of engineering education reform. This study focuses on the National Automotive Engineering Experimental Teaching Demonstration Center at a university in Southwest China. Over more than a decade, the Center has combined OBE-CDIO principles with the development of a Teaching-Learning-Competition-Innovation (TLCI) mechanism and has constructed a progressive four-level practical teaching system. This system aims to improve the quality of the cultivation of automotive engineering talent by leveraging disciplinary strengths and standardized practical teaching resources.

This study aims to develop and implement a four-level practical teaching system based on OBE-CDIO principles and to conduct a descriptive evaluation using multi-source evidence, with particular attention to its operating mechanism in the context of resource-constrained local universities, its relationship with student competency development, and its conditional transferability:

1. How can the OBE and CDIO frameworks be integrated to design a coherent four-level practical teaching system suitable for automotive engineering?

2. To what extent do multi-cohort longitudinal data and external feedback indicate that the TLCI-based four-level practical teaching system is associated with improving trends in students’ perceived core competencies and employment outcomes?

3. Under what institutional conditions does this teaching model have adaptability and transferability, particularly for resource-constrained local universities?

To answer these questions, this study adopts a design-based case study approach combined with a mixed-methods evaluation strategy^18–20. First, it clarifies the overall goals and core structure of the four-level practical teaching system. Second, it details the institutional development strategies and implementation pathways. Third, it integrates student self-assessments, project performance records, employer feedback, and national employment statistics for multidimensional descriptive evaluation. Finally, it compares this model with mainstream international engineering education frameworks and discusses the practical reference value of the OBE-CDIO-TLCI integrated model for engineering education reform in resource-constrained universities.

Proposed framework and implementation

Objectives and the TLCI mechanism

China established National Experimental Teaching Demonstration Centers to improve practical teaching standards and pilot experimental curriculum reforms. The main function of these centers is to meet national and regional industrial needs. By adopting student-centered teaching methods, they cultivate students’ core engineering skills, autonomous learning ability, and interdisciplinary competencies. To address the mismatch between automotive engineering graduates and industry demands, the Center constructed a systematic talent cultivation framework (Fig. 1). This framework integrates strategies such as curriculum optimization, pedagogical innovation, participation in industry competitions, and deepened industry-university-research collaboration²¹. Figure 1 is a conceptual framework independently constructed and drawn by the authors based on the Center’s development practice and related research. It is used to illustrate the logical relationships among practical teaching objectives.

Fig. 1.

Objectives of Practical Teaching.

The core mechanism is defined as TLCI, which aims to cultivate versatile engineering professionals through the synergy of four dimensions. These four dimensions were developed by integrating the Center’s long-term operational experience with recent research on experiential learning, engineering competition teams, challenge-based learning, systemic PBL, team-based projects, and innovation-oriented programs^22–27. Together, these studies provide the theoretical and empirical rationale for the selection and organization of the TLCI mechanism and the subsequent four-level practical teaching system within a localized OBE-CDIO framework.

1. Teaching: This dimension focuses on innovation in experimental methods and technologies. The design of this dimension is informed by the systematic review by Tembrevilla et al., which emphasizes that experiential learning in engineering education involves the integration of practical activities, theoretical understanding, contextualized tasks, and reflective processes. In the present framework, teaching therefore links conceptual instruction, experimental operation, and practical reflection to strengthen students’ engineering foundations and support the transition from knowledge acquisition to applied problem solving²².

2. Learning: This dimension emphasizes the construction of a multidimensional learning platform. It adopts a comprehensive, practice-oriented training model that combines virtual and real-world teaching tools to promote students’ all-around development.

3. Competition: This dimension uses automotive and transportation competitions to encourage active student participation. Research on engineering competition teams indicates that such participation can provide authentic opportunities for students to practice engineering roles, coordinate collaborative tasks, and develop leadership identity. Accordingly, competitions are embedded in this framework as structured project-based learning settings in which students apply professional knowledge, receive feedback, assume team responsibilities, and develop engineering practice skills²³.

4. Innovation: This dimension serves as the final objective of the system. Challenge-based learning research highlights the importance of real-world challenges, multidisciplinary collaboration, external stakeholder involvement, and iterative evaluation for competency development. In the TLCI mechanism, innovation is therefore understood as the cumulative outcome of teaching, learning, competition, feedback, and project iteration, thereby establishing a “virtuous cycle” in which competition drives teaching, learning, and innovation²⁴.

The four-level practical teaching system

At the National Automotive Engineering Experimental Teaching Demonstration Center, student competency development runs throughout the entire four-year undergraduate curriculum and follows a structured, phased pathway. This staged arrangement is consistent with systemic PBL research on the progression of students’ self-assessed learning outcomes, which suggests that competency development should be understood as a longitudinal and program-level process rather than as the result of isolated course activities²⁵. To further improve training quality, the Center established a progressive four-level practical teaching system, as shown in Fig. 2. Figure 2 is a systematic representation, independently constructed and drawn by the authors, based on the Center’s four-year practical teaching operational experience. Its hierarchical logic is inspired by OBE-CDIO and practical teaching research, but the specific levels, implementation pathways, and activity arrangements are derived from the long-term practice at the study site.

Fig. 2.

Practical Teaching System.

The system comprises four core levels: basic, professional, comprehensive, and innovative. Its progressive logic is consistent with both the chronological sequence of the undergraduate curriculum and the developmental trajectory of students’ professional competencies and vocational literacy²⁵.

1. Basic Level: The Basic Level aims to consolidate core foundational capabilities. It includes general education practice, introductory professional courses, metalworking practice, and career development awareness. These activities are positioned in the early stage of the undergraduate curriculum and lay the foundation for subsequent professional learning and skill acquisition.

2. Professional Level: The Professional Level focuses on the systematic cultivation of professional skills and vocational literacy. Through standardized experimental teaching procedures, it helps students master core engineering knowledge and practical techniques in automotive and transportation engineering. Key contents include professional course experiments, practical technical training, and specific engineering training modules.

3. Comprehensive Level: The Comprehensive Level is essential for developing the ability to solve complex engineering problems. It is designed to guide students in identifying problems in authentic engineering contexts through experimental inquiry, thereby strengthening systems thinking, integrated problem-solving ability, and awareness of innovation. Major activities include comprehensive course experiments, professional internships, and enterprise-based on-the-job training. Research on engineering team-based projects further suggests that professional skills such as communication, project coordination, and role adaptation are more likely to be developed when students engage in structured collaborative project work rather than purely individual technical tasks²⁶.

4. Innovative Level: As the highest level of the practical teaching system, the Innovative Level prioritizes the improvement of engineering problem-solving ability and higher-order engineering thinking. Studies on innovation competitions and programs suggest that their educational value depends not only on participation itself but also on mentoring, reflection, project iteration, and the alignment between competition tasks and learning objectives. Therefore, this level integrates innovation training projects, academic competitions, and project-based research as connected learning activities rather than as isolated extracurricular experiences²⁷.

To ensure the effective implementation of activities at each level, the Center provides strong support in staffing, funding, and equipment investment. It has also established standardized management systems and built a systematic, practical teaching-support architecture, as shown in Fig. 3.

Fig. 3.

Practical Teaching Support System.

Both physical and human resources require sustained funding and strong institutional support. To cultivate integrated teaching teams, the Center has adopted measures such as establishing specialized practical teaching platforms, upgrading experimental equipment, and implementing a joint guidance model that involves industry mentors and full-time faculty members. The Center deepens industry-university-research collaboration by hiring industry experts, encouraging faculty participation in industry practice, and building a comprehensive faculty development system. These measures provide continuous support for the quality of talent cultivation^28,29. This is consistent with a recent systematic review emphasizing that formal, in-depth industry partnerships and dual-mentor mechanisms are essential to ensuring the sustainability of engineering education and promoting the professional growth of faculty members³⁰. In terms of student career development support, this system does not treat the curriculum scheme itself as the sole driving factor. Instead, it provides students with professional knowledge, project experience, professional norms, and continuous feedback through the collaborative roles of campus faculty, industry experts, and practical platforms. Specifically, through sustained alignment among course objectives, project tasks, industry mentor feedback, and professional norm training, the four-level practical teaching system translates specialized training objectives into career competencies that students can perceive, practice, and develop sustainably.

Core construction strategies

The Center adopts a holistic educational philosophy that integrates knowledge transmission, competency building, professional literacy, and collaborative development. Guided by this philosophy, the Center continuously iterates its experimental teaching system to improve students’ engineering practice skills and innovative thinking, optimizing the training model to match the multifaceted competency structure required by industrial digitalization and intelligent transformation³¹.

Theory-practice integration

The Center has implemented systematic reforms across three dimensions: content, methods, and evaluation. At the implementation level, the Center funds approximately five teaching reform projects each year to balance the transmission of foundational knowledge with the cultivation of higher-order competencies. In terms of interdisciplinary integration, the Center promotes the intersection of automotive engineering and computer science. By incorporating frontier industry practices into the curriculum, this approach responds to the need in interdisciplinary engineering education to “integrate different knowledge domains to solve real problems.” In addition, through cooperation with domestic and international universities, nontechnical elements such as professional ethics and social responsibility have been incorporated into the curriculum, supporting the systematic development and deepening of engineering ethics education³².

To improve students’ ability to solve complex engineering problems, the Center optimized the curriculum structure. While maintaining the balance of total credit hours, it increased the proportion of comprehensive and innovative experiments. As a result, experimental courses now account for more than 27.1% of the total curriculum. In instructional design, the Center adopts a multidimensional structured model:

1. At the micro level, it focuses on in-depth analysis of key concepts and inquiry-based questions;

2. At the meso level, it establishes internal logical coherence within units and horizontal connections across disciplines;

3. At the macro level, it integrates disciplinary knowledge domains with holistic competency cultivation. By combining situated learning and flipped classrooms, this hierarchical design aims to stimulate students’ intrinsic motivation and engagement, thereby better aligning with the competency requirements of industrial transformation³³.

In experimental courses, the Center adopts a “virtual-physical integration” teaching model. This approach not only enhances instructional interaction, standardizes operational procedures, and deepens theoretical understanding but also prompts students to reflect on the development trends and transformation pathways of the automotive industry in the Internet era. For example, the automotive structure course combines physical disassembly experiments with virtual disassembly exercises and requires students to complete component measurement and drawing tasks. This hybrid model leverages the benefits of repeated practice and process visualization to improve students’ operational skills and, to some extent, learning outcomes.

Virtual–physical integration

The Center’s experimental system follows a teaching strategy of “physical practice as the main component and virtual simulation as a supplement.” Using virtual experiments, simulation, and immersive technologies, the Center applies virtual methods specifically to complex scenarios involving high risk, high cost, or unobservable dynamic processes. This approach ensures a “safe, visible, and repeatable” practical learning environment³⁴. In addition, comparative research on the complementarity of physical, simulated, and remote laboratories in terms of learning outcomes and resource efficiency provides an important theoretical basis for this approach³⁵. Related platforms are mainly used for process training, observation of task completion, and recording of competency performance. They provide process evidence for monitoring student competencies rather than replacing the comprehensive evaluation of faculty and industry mentors.

The Center has built a multi-level ecosystem of simulation tools. At the commercial software level, industry-standard tools are included, such as CarSim (vehicle dynamics), PreScan (autonomous driving testing), VISSIM (traffic flow simulation), Simulink (control modeling), and PC-Crash (accident reconstruction), to ensure that the teaching content aligns with industry standards. In terms of proprietary development, the Center has independently developed two types of key tools to address the limitations of commercial software in specific teaching contexts:

1. Traffic Accident Instructional Simulation System: This system integrates scenario reconstruction, interactive training, and skill assessment. Unlike general commercial software, it allows students to customize road environments and collision parameters. This function establishes a closed-loop teaching process from parameter setting to result validation, thereby enhancing instructional interaction.

2. Multi-Platform Data Conversion Interface: To address data compatibility problems among heterogeneous platforms such as VTD (Virtual Test Drive) and PreScan, the Center developed a dedicated format conversion tool. This tool facilitates cross-platform reuse of simulation data and supports complex experimental designs requiring collaboration among multiple tools.

3. By integrating open-source resources, the Center has deployed 13 self-developed simulation platforms covering fields such as traffic engineering and vehicle control. These proprietary tools work in coordination with more than 60 commercial software packages and support over 100 experimental projects each year. In teaching allocation, virtual simulation consistently accounts for approximately 25% of total experimental credits. Existing meta-analyses indicate that virtual laboratories have a positive overall effect on learning performance in engineering education, providing empirical support for allocating resources to this approach.

4. In addition, this “virtual-physical integration” model significantly reduces consumption of physical materials, reflects sustainability in the teaching process, and provides students with a safe space for trial and error in high-risk scenarios. Similarly, a recent meta-analysis shows that integrating virtual laboratories into engineering education can significantly improve students’ motivation and engagement. This approach provides a low-risk and highly interactive environment for exploring complex engineering concepts and can help mitigate resource and safety constraints in traditional laboratories³⁶.

Science–education–research synergy

Transforming faculty research resources into teaching content is a core educational strategy of the Center. Following the principle of research-teaching integration, the Center has established a continuous pathway from undergraduate practice to academic research since 2009. In its implementation, the Center has established an institutional mechanism that supports students’ direct participation in faculty-led research projects. For example, the Deep Traffic Accident Investigation project recruits approximately 20 undergraduate and master’s students for research each year. Similarly, in frontier areas such as the hydrogen fuel cell bus battery platform, project teams assign approximately 20 students each year to related work. In addition, research teams in intelligent vehicles, automotive powertrains, clean energy, heavy transportation, and special-purpose vehicles also actively recruit students. Statistics show that approximately 50% of students at the Center join at least one of 16 existing research teams each year for in-depth practical learning. During the research process, mentors guide students in identifying and solving complex engineering problems and help them master frontier professional knowledge and advanced technical skills. It should be noted that the preceding references to “approximately 20 students” are used to describe the annual participation scale of specific research projects. Because participation in such projects varies with project cycles, research tasks, and team arrangements, this study treats these figures only as descriptive information on the intensity of project operations and does not calculate percentages or conduct further statistical inference based on them.

Beyond the internal drivers of academic research, the Center prioritizes feedback loops from external industrial demand. To ensure that the practical teaching system remains aligned with industry development, the Center collects external feedback through multiple channels, including annual graduate tracking surveys, alumni associations, and alumni forums. This “external stakeholder participation-feedback-improvement” mechanism is consistent with systematic review evidence on community and industry engagement in engineering education³⁷. From 2021 to 2024, the Center systematically consulted numerous partner enterprises and more than 100 alumni on issues such as course design and talent cultivation objectives. In response to industry demand for stronger customer communication and technical service capabilities, the Center optimized relevant course content, innovated teaching methods, and placed greater emphasis on simulation training based on authentic service scenarios. These measures effectively improved students’ practical skills and workplace adaptability.

Student innovation and interdisciplinary practice

The Center adheres to the philosophy of “competition-driven learning” and has established a diversified competition system that includes Baja SAE, autonomous driving challenges, and transportation technology competitions. Using this framework, the Center has formed 13 interdisciplinary teams spanning 21 majors, collectively involving approximately 470 students. This collaborative model is consistent with the practical logic in engineering education that emphasizes “collaborative solving of complex problems”³⁸. Here, “approximately 470 students” refers to the cumulative participation across multiple competition teams across different majors and primarily reflects the coverage of the interdisciplinary practice platform. Because this figure represents cumulative cross-team participation rather than a proportion based on a single, stable annual student population, this study does not calculate a percentage.

Under faculty guidance, these teams engage in project-oriented practice and complete the full process from conceptual design, manufacturing, and assembly to on-site presentation and defense over a cycle of approximately ten months. This intensive, closed-loop engineering training helps improve students’ technical integration ability and systems thinking and is also consistent with industry demand for interdisciplinary engineering talent.

The Center further established a “competition-innovation integration” mechanism, encouraging students to transform engineering challenges encountered in competitions into entrepreneurship projects, such as those for the “Internet+” and “Challenge Cup” competitions. Through continuous iteration and improvement, students have successfully bridged the gap between pure engineering practice and innovation incubation. Recently, a student team from the Center won a national bronze award in the National “Internet+” College Student Innovation and Entrepreneurship Competition. This achievement provides illustrative evidence of the model’s practical value and suggests that students can perform competitively in high-level innovation practice.

To adapt to the service-oriented transformation of the automotive industry, the Center expanded the scope of its innovative practice from technological research and development to after-sales service. Based on industry feedback, the Center introduced modules on service process design and user needs analysis into the curriculum. Through school-enterprise collaborative projects simulating real business scenarios, students can strengthen customer communication and technical service capabilities while mastering technical skills. This dual-track “product and service” innovation practice broadens students’ career adaptability and promotes steady improvement in the quality of employment in graduate service positions.

Resource sharing and outreach

The Center not only conducts diverse on-campus experiments but also actively fulfills social responsibility by expanding science outreach and regional cooperation. Since 2020, the Center has successfully hosted more than 400 public welfare activities and received more than 6,000 visitors from primary and secondary schools, universities, government agencies, and enterprises. These activities take diverse forms, including research facility visits, science outreach activities, and laboratory open days, such as those held during Science and Technology Week, the Chengdu Creative Design Exhibition, and youth maker activities (Table 1). As a result, the Center has become an important regional platform for disseminating scientific knowledge and showcasing achievements in engineering education. This community-oriented participation is consistent with systematic evidence on the benefits of community engagement and service-oriented practice in engineering education.

Table 1.

Science Outreach and Collaboration.

Year	Activity Type	Number of Visitors	Participating Units	Highlights
2021	Science research and enterprise visits	1200+	Primary and secondary school teachers and students, alumni, and industry enterprises	Hosted research open days and enterprise technology exchange sessions
2022	University lab open house and science promotion activities	2200+	Universities, the public, and related research institutions	Demonstrated virtual simulation technology, engaged students in interactive participation
2023	Lab visits and transportation safety-themed events	2600+	Local governments, transportation industry associations, university faculty, and students	Virtual simulation teaching demonstrations showcased research achievements.

To expand its social impact, the Center designs customized visit itineraries for different groups, including government officials, industry practitioners, and students. To ensure operational safety and sustainability, the Center has built a diversified funding system. It uses special government funds, industry-university-research cooperation projects, and social donations to ensure the continuity of teaching and research activities. Considering regional differences in resources, the model follows the principle of “adapting measures to local conditions” during dissemination. It advocates the use of information technologies, such as virtual laboratories, to reduce dependence on capital-intensive hardware. In addition, the Center actively promotes resource-sharing mechanisms with universities, insurance companies, and research institutes to leverage complementary strengths and improve the quality of regional experimental teaching.

Materials and methods

Study design and ethics statement

The research protocol and all procedures involving human participants followed the principles of the Declaration of Helsinki and were approved by the Ethics Committee of the School of Automotive and Transportation Engineering at the host university (Approval No.: 2024LL(05); approval date: May 20, 2024). The ethical review covered the prospective questionnaire survey and the retrospective analysis of de-identified administrative records. For the survey, participants were provided with a study description, and electronic informed consent was obtained before data collection. For the secondary analysis of administrative data, the committee approved a waiver of informed consent because the data contained no direct personal identifiers and were reported only in aggregate form.

Data collection sources

To describe and evaluate the operation of the four-level practical teaching system, this study employs a descriptive evaluation framework grounded in a multi-source chain of evidence. All evaluation activities are centered on the operation of the Center’s four-level practical teaching system. This evaluation framework has motivational, diagnostic, corrective, and forward-looking functions: it is used to stimulate students’ intrinsic motivation to participate in practical learning, identify students’ competency development needs and course implementation problems, adjust practical activity arrangements based on feedback, and provide a basis for subsequent optimization of the training program. The selection of evaluation tools follows three principles: first, alignment with the objectives of the four-level practical teaching system; second, coverage of both student self-assessment and external evaluation; and third, a verifiable basis for reliability and validity.

On this basis, this study uses multiple forms of evidence, including student questionnaires, longitudinal administrative archives, and employer feedback, for cross-validation. The main sources of evidence include:

1. Longitudinal cohort tracking evaluation: Based on administrative archives, this study tracked graduates from the 2020–2023 cohorts to examine their employment and career development. Data were obtained from historical records of the Academic Affairs Office and Career Services Office, and all personal identifiers were removed. To ensure transparency, Wilson 95% confidence intervals are reported for proportion indicators, and t-distribution 95% confidence intervals are reported for continuous variables. Long-term indicators were calculated only for eligible cohorts, while others were marked as N/A³⁹.

2. Student satisfaction and competency self-assessment survey: In 2024, the “Satisfaction and Effectiveness Evaluation Questionnaire for the OBE-CDIO Four-Level Practical Teaching System” was distributed through an online platform to junior and senior students and graduates from the previous three years. A total of 312 valid questionnaires were obtained. The survey focused on three core competencies: engineering problem-solving ability, interdisciplinary teamwork, and engineering literacy. It included 13 indicators measured using a five-point Likert scale. Detailed indicators are provided in the Supplementary Information.

3. Employer feedback: Through alumni and employer feedback mechanisms, 125 valid employer evaluations were collected in 2024 to examine graduates’ job fit and professional performance. These feedback data served as third-party evidence, complementing and corroborating student self-assessments and longitudinal archives through triangulation⁴⁰.

Statistical analysis

This study used SPSS to analyze the survey data. For the three dimensions of the student competency questionnaire, namely engineering problem solving, interdisciplinary collaboration, and engineering literacy, Cronbach’s α coefficients were 0.842, 0.857, and 0.889, respectively. The Kaiser-Meyer-Olkin (KMO) values were 0.816, 0.826, and 0.882, and Bartlett’s tests of sphericity were significant for all dimensions (p < 0.001), indicating that the data were suitable for factor analysis. Principal component analysis showed that a single principal factor was extracted for each dimension, with cumulative variance explained of 67.84%, 70.06%, and 69.22%, respectively, suggesting acceptable convergent validity. In addition, this study conducted item-deletion sensitivity analysis. The results showed that Cronbach’s α did not increase after deleting any single item in any dimension: the α values after deleting individual items were 0.788–0.810 for engineering problem solving, 0.813–0.827 for interdisciplinary collaboration, and 0.862–0.868 for engineering literacy, all lower than the original α values for the corresponding dimensions. No obvious redundant item was found that could be deleted to improve internal consistency.

To avoid interpreting the data as causal improvement in a post-test design, this study uniformly reports Likert means, standard deviations (SDs), and 95% confidence intervals (CIs), and strictly interprets them as students’ perceived levels of competency improvement⁴¹. In addition, descriptive statistics and frequency analyses were used to characterize the distribution of high-score categories. The item-deletion analysis supports the internal consistency of the current item set, but measurement invariance and cross-sample stability still require further examination in future research.

Methodological statement and limitations

All analyses in this study should be interpreted as descriptive and correlational evidence. Because of the absence of a rigorous control group or quasi-experimental design, observed differences across cohorts or groups may be affected by multiple external factors, including macroeconomic fluctuations, changes in industry demand, the COVID-19 pandemic, student self-selection bias, and changes in regional employment structure. Therefore, this study uses these results only to describe trends, provide contextual reference, and form multi-source corroboration, rather than to infer the net causal effect of the TLCI mechanism. The relevant statistical criteria and sample information are specified in the titles or footnotes of the corresponding tables in the Results section.

The virtual learning space in this study primarily refers to a virtual simulation environment for engineering practice training. Its purpose is to reduce experimental risk, improve the repeatability of training, and expand access to practical resources, rather than to discuss an online psychosocial space in the sense of anonymous interaction or identity presentation.

Results

In recent years, the Center has offered more than 90 undergraduate experimental courses each year, covering over 270 experimental projects and completing approximately 120,000 student-experiment sessions. Notably, 90% of these experiments are classified as comprehensive, design-oriented, or innovative. These figures indicate that the Center has built an integrated practical teaching supply chain spanning courses, projects, and platforms, with a structural feature characterized by a high proportion of comprehensive, design-oriented, and innovative experiments.

Competition participation and academic achievements

Since 2007, the School of Automotive and Transportation Engineering has relied on the national experimental teaching demonstration center to continuously advance the development of a student-centered practical teaching system. With support from national and provincial education projects, the school has deepened school-enterprise cooperation mechanisms and actively built academic competition platforms. The school has established seven specialized student competition teams focused on core research and practice in fields such as internal-combustion-engine vehicles, electric vehicles, and autonomous driving technologies. To date, 538 students have joined these competition teams, including 430 vehicle engineering majors, accounting for 56.7% of all students in that major.

Through hands-on training, the competition teams have effectively integrated multidisciplinary knowledge from mechanical engineering, materials science, industrial design, and management. This process appears to have supported students’ teamwork, interdisciplinary collaboration, and engineering innovation capabilities. Over the past four years, students have participated in more than 300 competitions and won more than 120 awards, including more than 70 national awards (Table 2).

Table 2.

Academic Competition Achievements.

Competition Name	Award Levels	Achievements	Year
National Undergraduate Smart Car Competition	15 nationals’ first and second prizes	Students’ design and innovation abilities are highly recognized	2017–2024
China Automotive Engineering Society Baja Competition	10 national second prizes, 20 individual top-three awards	Teams excelled in problem-solving and engineering capabilities	2017–2024
China Formula Student Competition	10 national second and third prizes, “Most Influential Team.”	Teams gained widespread recognition for innovation and creativity	2017–2024

The career achievements of these outstanding alumni provide case-based evidence of the four-level practical teaching system’s talent-cultivation quality. For example, students who participated in the “Intelligent Cockpit” practical training project successfully transitioned into frontier fields such as autonomous driving by drawing on experience accumulated through competitions and practical training.

1. Graduate A won a silver medal in the Automotive Technology category at the 44th WorldSkills Competition.

2. Graduate B currently serves as a product manager at a precision machinery enterprise, leads several key technical projects, and was awarded the title of Sichuan Provincial Model Worker in 2020.

3. Graduate C is currently a technical manager at an automotive research institute and has participated in several National Natural Science Foundation of China projects and major projects of the Ministry of Industry and Information Technology.

Student self-assessed competency improvement

Table 3 shows that students’ self-assessments of the three core competencies ranged from moderate to high, with engineering problem-solving ability receiving the highest score. This trend is highly consistent with the empirical findings of Yang⁴². Prior research suggests that introducing structured PBL aligned with industry needs may help bridge the gap between theoretical teaching and industrial application and may improve students’ ability to solve complex engineering problems while increasing their learning engagement. This subjective perception was further corroborated by longitudinal archives (Table 4) and employer feedback (Table 6).

Table 3.

Core Competency Self-Assessment Results.

Core Competency Dimension	Assessment Items	Mean (1–5)	SD	95% CI	Interpretation
Engineering problem-solving ability	Ability to identify problems and propose feasible solutions (Q1–Q4)	4.17	0.51	[4.11, 4.23]	High perceived level
Interdisciplinary collaboration (role adaptation, coordination)	Team communication, role adaptation, and coordination (Q5–Q8)	3.76	0.53	[3.70, 3.82]	Moderate-to-high level
Engineering literacy (ethics, standards, regulations)	Ethics, sustainability awareness, and safety standards (Q9–Q13)	3.96	0.53	[3.90, 4.02]	Moderate-to-high level

The scale used a five-point Likert rating. The table reports means, standard deviations (SDs), and 95% confidence intervals (CIs). “Result interpretation” refers to descriptive thresholds for contextual understanding only and does not indicate causal change. All table analyses follow the general disclaimer in the main text and provide only descriptive evidence and contextual reference; they do not support causal inference. Specific details for each table are provided in the main text or table captions, or follow standard statistical practice.

Table 4.

Longitudinal Administrative Tracking Results.

Indicator	Definition/Calculation	2020	2021	2022	2023	Overall
Employment rate (%)	Proportion of Employed in Employment Status	94.03 [91.20, 95.99] (n = 362)	93.77 [90.89, 95.78] (n = 361)	96.62 [94.31, 98.02] (n = 372)	97.13 [94.93, 98.39] (n = 372)	95.38 [94.22, 96.32] (N = 1,467)
Starting salary (RMB/month)	Mean ± SD; 95% CI (employed/further study only)	5295 ± 758 [5217, 5374] (n = 362)	5153 ± 941 [5056, 5250] (n = 361)	6148 ± 787 [6068, 6228] (n = 372)	7519 ± 858 [7432, 7607] (n = 372)	6040 ± 1261 [5976, 6105] (N = 1,467)
Starting salary growth (2020→2023)	(Mean 2023 – Mean 2020)/Mean 2020 × 100%	—	—	—	+ 42.00%	—
Job-major alignment (High, %)	Proportion of Job Relevance Level = 3 among the employed	78.45 [73.93, 82.38] (n = 284)	78.39 [73.86, 82.33] (n = 283)	79.03 [74.61, 82.86] (n = 294)	84.14 [80.08, 87.50] (n = 313)	80.03 [77.90, 81.99] (N = 1,174)
High job satisfaction (≥ 4 after ≥ 3 years, %)	Proportion of Job Satisfaction ≥ 4 among eligible follow-up	82.04 [77.76, 85.66] (n = 297)	82.00 [77.71, 85.62] (n = 296)	N/A	N/A	82.02 [79.05, 84.65] (N = 593)
Employment rate (Competition subgroup, %)	Employment rate within the subgroup (Competition = Yes)	—	—	—	—	97.40 [95.68, 98.44] (N = 524)

Table 6.

Employer Feedback Summary.

Competency Dimension	Typical Performance Description	Center Proportion	National Benchmark
Engineering ethics and safety literacy	Enhanced awareness of engineering ethics, safety standards, and sustainability	63%	47%–49%
Career development potential	Higher likelihood of promotion within five years	76%	66%
Job–major alignment	High relevance between job position and academic major	80%	74%
Overall workplace performance satisfaction	Employer satisfaction with graduates’ role adaptation/performance	82%	77%

Longitudinal employment outcomes

Table 4 shows cohort-specific trends in employment rate, starting salary, and job-major alignment for this major. Statistical tests indicated significant cohort differences in employment rate [χ²(3) = 7.89, p = 0.048, Cramér’s V = 0.072] and highly significant differences in starting salary [F(3,1463) = 617.72, p < 0.001, η² = 0.559], with the 2023 cohort showing a large effect size compared with the 2020 cohort. No significant cohort difference was found in job-major alignment. Notably, competition experience was positively associated with employment. Tracking data collected through the National College Graduate Employment Platform and the Center’s alumni feedback mechanism showed that graduates who participated in research and competition projects generally had higher starting salaries than the major average and achieved faster career development within two to three years after graduation.

1. Values are percentages, with Wilson 95% confidence intervals or t-distribution 95% confidence intervals in parentheses.

2. Sample sizes vary by employment status and data availability; N/A indicates that data have not yet been obtained for cohorts with fewer than three years after graduation.

3. Starting salary growth was calculated only for the direct comparison between the 2020 and 2023 cohorts.

4. High job satisfaction was measured using a five-point Likert scale. Competition grouping refers to students who participated in disciplinary competitions.

5. All data are descriptive. Because of the lack of a control group, observed differences should not be interpreted as causal effects; statistical significance is reported only where explicit tests are provided.

To provide external context, this study compared the employment indicators for the vehicle engineering major with those for related engineering majors at the university and with indicators from external experimental centers⁴³. Table 5 shows the performance of the vehicle engineering major relative to these reference groups on key employment indicators. This external comparison serves as a contextual supplement to internal longitudinal tracking.

Table 5.

Employment Quality Comparison.

Indicator	National Experimental Teaching Demonstration Center	Mechanical Engineering	Materials Engineering	Off-campus Experimental Center B
Employment rate (%)	95.4%	92.1%	90.8%	89.5%
Job satisfaction (≥ 4) after ≥ 3 years of work (%)	82%	78%	67%	76%
Job–major alignment (High, %)	80%	76%	68%	77%
Starting salary growth rate (2020→2023)	42%	28%	25%	20%

Employer feedback

Table 6 summarizes employers’ evaluations of graduates’ professional literacy and job fit, with indicators echoing those used in related research⁴⁴. This third-party perspective further enriches the multi-source chain of evidence.

Discussion

Interpretation of findings

This study constructed and implemented a four-level practical teaching system within the OBE-CDIO framework. Multi-source descriptive evidence suggests that this model is associated with improving trends in graduates’ perceived core engineering competencies, competition participation, and employment quality indicators. In the survey results, students rated engineering problem-solving ability highest, which may be related to the project-driven and virtual-physical integration strategies in the curriculum reform. Meanwhile, the positive association between competition experience and employment rate suggests that competition-driven learning within the TLCI mechanism warrants further study. Alumni tracking and employer feedback jointly indicate that students who participated in research and competition projects performed well in engineering practice skills, interdisciplinary communication, and workplace adaptation, but these results should still be interpreted as descriptive associations rather than strict causal effects.

Comparison with international engineering education models

Compared with internationally common paradigms, the Center’s four-level practical teaching system is not a simple replication. Instead, it represents localized adaptation and structural reconstruction of core OBE-CDIO principles, tailored to the context of a regionally oriented, application-focused university.

1. Comparison with the ABET framework: Although ABET emphasizes outcomes assessment and continuous improvement-oriented quality assurance, the Center’s model goes beyond outcomes. It provides a scaffolded process-support system through a four-level progression of basic, professional, comprehensive, and innovative training.

2. Comparison with the German dual system: The German model is known for its intensive enterprise apprenticeship. On this basis, the Center’s model integrates an additional dimension of interdisciplinary synthesis. Studies show that while the dual system significantly enhances vocational competence, it also places higher demands on organizational coordination⁴⁵.

3. Comparison with the Nordic CDIO model: The Nordic model is usually driven by design projects. The Center’s model further strengthens the integration of research and teaching by directly incorporating undergraduates into faculty research teams and national-level competitions.

Institutional recognition and scalability

The implementation outcomes of this system have generated institutional recognition and some external dissemination. The Center has established a practical teaching framework with OBE as the core and deep integration of the four key dimensions of TLCI. Institutional recognition provides external support for the relevance and institutional acceptance of these reforms: the vehicle engineering major was designated as a national specialty, and the Center was officially approved as a national experimental teaching demonstration center (Table 7).

Table 7.

Teaching Achievements and Publications.

Category	Name	Level	Awarding Institution
Teaching Achievements	National Distinguished Specialty: Vehicle Engineering	National	Ministry of Education
	National Experimental Teaching Demonstration Center: Automotive Engineering	National	Ministry of Education
	Excellence in Engineering Education Program: Vehicle Engineering	Provincial	Sichuan Provincial Department of Education
	Provincial First-Class Specialty: Vehicle Engineering	Provincial	Sichuan Provincial Department of Education
	University Teaching Excellence Award: Collaborative “Teach-Learn-Compete-Innovate” Model	University	The university where the experimental center is located
	Provincial First-Class Course: Vehicle Engineering Technological Innovation Practices	Provincial	Sichuan Provincial Department of Education
Publications	Automotive Electronics and Control Technology Fundamentals		Mechanical Industry Press
	Automotive Electrical and Electronic Technology		Mechanical Industry Press
	Automotive Operational Materials		China Communications Press

In terms of educational research outcomes, teaching staff have consistently focused on improvement-oriented teaching research and reflective practice, disseminating their experience through peer-reviewed publications (Table 8). In addition, the Center’s core textbooks have been adopted by multiple universities, suggesting that these teaching resources have been adopted beyond the Center.

Table 8.

Research Achievements and Social Impact.

Category	Achievement Name	Level	Awarding Institution
Research Achievements	New Energy Vehicle Provincial Innovation Team	Provincial	Sichuan Provincial Department of Science and Technology
	Sichuan Provincial Automotive Safety Industry Collaborative Innovation Center	Provincial	Sichuan Provincial Economic and Information Department
	Sichuan Provincial Postdoctoral Innovation Practice Base	Provincial	Sichuan Provincial Department of Human Resources and Social Security
	Sichuan Engineering Research Center for Intelligent Control and Simulation of New Energy Vehicles	Provincial	Sichuan Provincial Development and Reform Commission
Academic Promotion and Social Impact	Research on China’s Formula Student Competition and Talent Cultivation		Tongji University Press
	Sichuan Provincial Outstanding Expert: Xu Yanhai	Provincial	Sichuan Provincial Department of Education

To enhance the system’s transferability, the Center has gradually developed standardized implementation and optimization tools, including practical teaching feedback standards and course evaluation operational manuals. Through peer exchange and regional cooperation, relevant experience has provided a reference for practical teaching reform at some universities. For resource-constrained local universities, the reference value of this model does not lie in replicating a single hardware platform, but in drawing on its four-level progressive structure, TLCI feedback loop, virtual-physical integration, dual-mentor support, and resource-sharing mechanism, and adapting these elements locally according to different universities’ funding conditions, faculty base, and industrial environment.

Sustainability and social impact

The four-level practical teaching system incorporates the principle of “engineering for the benefit of society” into its design and emphasizes the social value of technical skills. Specific examples include the traffic accident investigation team cooperating with local institutions to convert professional data into traffic safety teaching materials for primary and secondary school students, and the intelligent driving team developing assistive systems for people with visual impairments. These projects illustrate how engineering education can respond to the needs of vulnerable groups and cultivate students’ awareness of engineering ethics. Prior research shows that learning experiences oriented toward social contexts and public welfare can promote students’ sociotechnical integration thinking and sense of responsibility⁴⁶.

To assess graduates’ performance in social responsibility, a targeted follow-up survey of the 2020–2024 cohorts showed that approximately 42.5% of respondents had participated in public welfare projects or corporate social responsibility (CSR) activities within three years after graduation. In response to these findings, the Center updated courses such as Engineering Ethics and Social Responsibility over the past two years and introduced a modular Corporate Social Responsibility Practice Training component. Related research indicates that embedding ethical norms into engineering courses with assessable learning outcomes can support students’ gradual development from normative awareness to responsible engineering judgment in complex situations⁴⁷.

Limitations

This study has several limitations that should be acknowledged:

1. Regional differences in resource availability: Some institutions, especially those in western China, face infrastructure constraints that may hinder full implementation of the system and limit its broader scalability.

2. Differences in students’ interdisciplinary foundations: Students from non-core majors often show gaps when participating in advanced projects, thereby increasing teaching demands.

3. Limitations in data traceability: Historical data were not collected using a unified evaluation tool, making strict before-and-after comparison within the same institution impossible.

These limitations are common in real-world educational reform research and highlight areas for future improvement.

• (5)Limitations related to sample and measurement tools: The sample in this study was mainly drawn from a single center and its associated training system, and employer feedback and student self-assessments may be affected by sample coverage, common method bias, and high similarity among items. Although item-deletion sensitivity analysis showed that Cronbach’s α did not increase after deleting any item in any dimension, suggesting a certain degree of stability in the current item set, the number of scale items, inter-item correlations, construct dimensionality, and measurement invariance may still affect the measurement results. Therefore, conclusions about scale structure and measurement stability should be interpreted cautiously. Future research should expand cross-institutional samples and further examine scale structure, item discrimination, measurement invariance, cross-sample stability, and the boundary conditions of model applicability.

Conclusions

This study developed and implemented a four-level practical teaching system under the OBE-CDIO framework, with the TLCI mechanism as its core driver, and descriptively evaluated its operation using multi-source evidence, including longitudinal administrative records, perceived competency surveys, and employer feedback. The findings show that the system is associated with positive changes in indicators such as students’ perceived engineering competencies, competition participation, and employment quality, but the results should still be interpreted as descriptive and correlational evidence rather than strict causal effects.

This work makes three main contributions:

1. It presents student perceived competency results in a standardized manner while avoiding over-causal interpretation of the results;

2. It clarifies the implementation pathway of the TLCI mechanism in a real engineering education context;

3. It provides a descriptive evaluation framework that can be further validated and locally adapted in subsequent research.

Future research will advance in three main directions: first, developing a cloud-based virtual simulation sharing platform to improve access to practical teaching resources for resource-constrained universities; second, exploring artificial intelligence-driven personalized preparatory courses to support competency bridging for students with different academic foundations; and third, expanding sample sources, conducting cross-institutional and cross-regional external validation, and further streamlining and reviewing the evaluation scale to examine the applicability of this model under different resource conditions. Overall, this study provides an implementable pathway and evaluation framework for practical engineering education reform in resource-constrained local universities.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, F.C. and J.Y.; Methodology, F.C. and P.C.; Software, S.L. and X.M.; Validation, Z.M., Y.Z. and J.C.; Formal analysis, F.C. and S.L.; Investigation, P.C. and X.M.; Resources, J.Y. and J.C.; Data curation, P.C. and Z.M.; Writing—original draft preparation, F.C. and P.C.; Writing—review and editing, J.Y. and J.C.; Visualization, F.C. and S.L.; Supervision, J.Y. and Z.M.; Project administration, J.Y.; Funding acquisition, J.Y. and J.C. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Intelligent Vehicle Industry Collaborative Training Special Research Topic of the Sichuan Association of Higher Education (Grant No. GJQZ-2025-ZD-06), the Education and Teaching Reform Research Project of Xihua University (Grant No. xjjg2025017), and the Foundation of the Chengdu Future Intelligent Manufacturing Industry Development Research Center, Chengdu Vocational & Technical College of Industry (Grant No. 2025WLZZ-YB03).

Data availability

All data generated or analysed during this study are included in this published article and its Supplementary Information files. The underlying datasets are provided as Supplementary Data 1 and Supplementary Data 2. These files contain de-identified and aggregated data/statistical summaries required to reproduce the analyses reported in this paper; no individual-level administrative records or personally identifiable information are shared.

Declarations

Competing interests

The authors declare no competing interests.