Can Artificial Intelligence and Data redefine engineers?--Transdisciplinary revolution in Engineering Education

Research Paper Recommended citation: Zhang, Y., Wu, J., & Zhu, L. (2025). Can Artificial Intelligence and Data redefine engineers--Transdisciplinary revolution in Engineering Education. In Kangaslampi, R., Langie, G., Järvinen, H.-M., & Nagy, B. (Eds.), SEFI 53rd Annual Conference. European Society for Engineering Education (SEFI), Tampere, Finland. DOI: 10.5281/zenodo.17631838. This Conference Paper is brought to you for open access by the 53rd Annual Conference of the European Society for Engineering Education (SEFI) at Tampere University in Tampere, Finland. This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License. Can Artificial Intelligence and Data Redefine Engineers？-- Transdisciplinary Revolution in Engineering Education Y. Zhang a,1, J. Wu b, L. Zhu c, aZheJiang university, Hangzhou, China, https://orcid.org/0009-0003-5797-6426 bZhejiang University of Finance and Economics, Hangzhou, China,https://orcid.org/0009-0006-6069-4237 cZheJiang university, Hangzhou, China,https://orcid.org/0009-0004-2895-1462 Conference Key Areas: Digital tools and AI in engineering education Keywords: Data Science; Artificial Intelligence;Engineering Education Transformation ;Future-Oriented Talent;Transdisciplinary ABSTRACT The integration of artificial intelligence and big data analytics is reshaping engineering education, revealing differentiated institutional responses to technological convergence. Through transdisciplinary case studies of Zhejing University (ZJU) and the University of California, Berkeley (UCB), this research examines paradigm evolution through three structural dimensions: knowledge reconstruction, organizational innovation, and ecosystem support. Two implementation paradigms emerge: ZJU's AI-centric model adopts a top-down transformation by integrating strategic capabilities, whereas UCB’s data-driven approach evolves bottom-up, leveraging existing infrastructure to mediate institutional change. The study conclusively demonstrates that transdisciplinary education, driven by AI's capability for vertical knowledge integration and data's role in horizontal competency expansion, redefines engineers as adaptive sociotechnical synthesizers -- directly answering the central inquiry of technological reconfiguration in engineering education. 1Corresponding Author Initials Last name:Y.Zhang e-mail address:[email protected] 1 INTRODUCTION The rapid advancement of AI and big data is reshaping engineering paradigms, challenging traditional educational frameworks. The longstanding disciplinary demarcations, such as those between mechanics, materials science, and computing, are increasingly obsolete in the face of modern engineering’s demand for transdisciplinary convergence. This shift necessitates the integration of intelligent systems, digital twins, and multi-physics modeling, fostering a holistic approach that bridges traditional silos and equips engineers to navigate complex challenges. Data-driven methodologies are driving reforms in engineering education, focusing on two key areas. First, technological progress demands higher digital competencies among engineers, necessitating the integration of AI and data science into curricula to modernize traditional frameworks (Qian et al., 2025). Second, global initiatives are emerging to develop digitally competent engineering talent. The EU’sDigital Europe Programme and Digital Education Action Plan foster tripartite partnerships (industryacademia-government) to develop competency frameworks (Shi et al., 2021), while the US cultivates data science talent through academic-industry collaborations (Wu et al., 2021). Similarly, emerging economies like China and India are rapidly establishing digital talent systems to support their expanding digital economies. The development of emerging technologies requires the reorganization of knowledge systems and drives innovation within transdisciplinary frameworks, particularly through the enabling capabilities of data science and artificial intelligence. Transdisciplinary approaches have emerged as a pivotal framework and strategic mechanism for engineering education reform, integrating knowledge from diverse disciplines to address real-world challenges (Steiner & Posch, 2006). In curriculum design, transdisciplinary education facilitates knowledge synthesis through themebased learning and project-driven pedagogical models (Firdaus et al., 2018). Empowered by artificial intelligence, higher education has undergone comprehensive digital transformation across instruction, assessment, and management, advancing personalized learning via adaptive systems (Yekollu et al., 2024), intelligent teaching aids(Liang, 2024), and immersive experiences (Wang & Huang, 2025).Contemporary transdisciplinary approaches emphasize comprehensive collaborative curriculum models (Allinson et al.,2022; Bile et al.,2021), utilizing differentiated resources to advance student development across transdisciplinary domains and learning stages. Organizational innovation manifests in two forms: vertically integrated courses that dismantle departmental barriers to restructure knowledge systems; and platforms (e.g., transdisciplinary institutes, open innovation labs) that facilitate multi-agent collaboration, cross-disciplinary knowledge integration, and the cultivation of core competencies (Mejía, 2023). Within ecosystem support, the emphasis on social responsibility, public good, and global sustainability (Bernstein, 2015) promotes an ethics-driven framework for equitable education (Kaufman et al., 2003). This study examines innovative educational approaches at two institutions: Zhejiang University (ZJU) in China, which leverages AI to restructure curricula through disciplinary transformation centered on AI integration; and the University of California, Berkeley (UCB), which integrates data science to coordinate transdisciplinary resources, enabling data-driven pedagogical practices across disciplines. These models not only provide an actionable blueprint for educating the next generation of engineers to address evolving societal demands, but also offer an adaptable framework for global innovation in engineering education. Through a comparative analysis of these talent cultivation models, this study explores how leading universities design transdisciplinary education programs to foster the progressive development of engineers' competencies and drive innovation in engineering education paradigms. This analysis contributes to a strategic understanding of digital talent cultivation in higher education and investigates systemic innovations in engineering education within the era of digital intelligence.. 2 METHODOLOGY 2.1 Research Methodology This study employs a dual-case comparative analysis to explore two models of higher education transformation. ZJU and the UCB were selected as case studies due to their status as world-leading engineering research universities with computational specializations. ZJU, ranked 29th globally by QS in 2024 and a longstanding leader in AI research, established the "West Lake Light" AI platform (Qiu et al., 2025). Meanwhile, UC Berkeley, ranked 5th globally by QS, strengthened its leadership through cross-campus data science education via the College of Computing, Data Science and Society (CDSS) (Chayes, 2021). Supported by robust industry-academia networks and regional innovation ecosystems, both institutions exemplify data-driven approaches to engineering education. For data collection and analysis, this study systematically gathered and cross-validated multidimensional qualitative data from institutional documents, conference reports, expert lectures, media publications, digital platforms, interviews and scholarly literature to ensure methodological rigor and enhance the credibility of conclusion through triangulation. 2.2 Research Framework This research proposes an analytical framework grounded in three dimensions of transdisciplinary studies: (1) Knowledge reconstruction —reforming traditional disciplines through blended curricula and cross-domain methods; (2) Organizational innovation —realigning academic units and resources to facilitate cross-faculty collaboration; and (3) Support ecosystem—building cultural inclusivity, institutional support, and infrastructure for talent development, as illustrated in Fig. 1. Building upon this framework, the study applies it to analyze problem-solving strategies for complex societal challenges. It highlights the critical role of integrating multistakeholder perspectives and proposes practical pathways for embedding transdisciplinary approaches within higher education systems.. Fig 1 A three-dimensional framework for analyzing cross-curricular education systems （ Figure created by the authors ） 3 RESULTS 3.1 Strategic Framework for Artificial Intelligence Talent Cultivation at ZJU ZJU has established an AI-driven educational framework through institution-wide coordination, developing comprehensive AI programs across multiple disciplines. This initiative structures curricula at various academic levels and disciplines through three approaches: vertical resource integration, horizontal cross-departmental collaboration, and industry partnerships. The model focuses on developing AI competencies within an innovation-driven educational ecosystem. As illustrated by the authors, the model of ZJU's training program is shown in Figure 2. Fig 2 The AI Talent Cultivation Model at ZJU （ Figure created by the authors ） 3.1.1 Knowledge Reconstruction: Multidimensional AI curriculum ZJU has developed a comprehensive AI curriculum structured across four tiers. The first tier offers discipline-specific core courses, such as the Computer Science 101, which establishes Artificial Intelligence as a foundational subject. The second tier provides micro-credential programs through "AI+X Modules", providing globally recognized credentials to students and enhancing their transdisciplinary skills. The third tier comprises 158 transdisciplinary electives accessible to all students. The fourth tier delivers general education through three AI literacy pathways: Foundation A(STEM/Medicine), Foundation B (Social Sciences), and Foundation C (Humanities/Arts), covering 58 courses and serving over 3,000 students. Faculty coordinators oversee AI instruction, curriculum development, and teaching materials to ensure consistency. This collaborative model integrates educational resources, research initiatives, and industry partnerships to foster an "AI+" ecosystem. 3.1.2 Organizational Innovation: Integrated and Collaborative Organizational Framework ZJU has vertically consolidated resources, establishing the Artificial Intelligence Education and Teaching Research Center in 2024 to coordinate AI initiatives. The center focuses on three key areas: (1) AI curriculum development, (2) technical infrastructure, and (3) pedagogical innovation, with a particular emphasis on integrating generative AI (GenAI). The university’s dual-track academic framework combines "elite specialization" with "transdisciplinary integration." AI undergraduates at the Zhu Kezhen Honors College receive small-class instruction and mentorship, with a curriculum designed to strengthen mathematical foundations and foster innovation through courses that integrate AI fundamentals with cross-disciplinary applications in computer science, mathematics, and neuroscience. Additionally, ZJU has built a collaborative network via partnerships with the Ministry of Education ’s Collaborative Innovation Center for AI, the Yangtze River Delta Research University Alliance, and the Next-Generation AI Education Consortium, forming a national AI education network that standardizes teaching, faculty development, and transdisciplinary evaluation. 3.1.3 Ecosystem Support: Comprehensive Support for AI literacy Innovation Capacity ZJU has established a four-dimensional competency framework, encompassing systematic knowledge, practical skills, innovative thinking, and ethical awareness, through its Undergraduate AI Literacy Standards. This framework underscores ethical algorithm design and the cultural responsibility of technology in serving societal values. To cultivate AI talent, the university has built a comprehensive technological ecosystem that integrates computational infrastructure, data management systems, and platform solutions. Key components include a no-code AI development environment, a unified teaching-research platform, domain-specific models, and intelligent agent systems co-developed by faculty and students using the DeepSeek V3/R1 framework. Strategic industry partnerships have led to the establishment of the Kunpeng Scientific-Educational Innovation Centre (in collaboration with Huawei) and the WisdomMO platform, enabling personalized learning analytics and cross-domain resource integration. 3.2 Strategic Framework for Data Science Talent Development at UCB UCB has developed an integrated data science education framework through three strategic dimensions: (1) embedding data science principles and discipline-specific skills across academic programs; (2) implementing interdepartmental governance reforms to facilitate resource sharing and collaboration; and (3) fostering an open ecosystem that connects education with real-world societal challenges. As illustrated by the authors, the model of UCB's training program is shown in Figure 3 . Fig 3 The Data Science Talent Cultivation Model at CDSS （ Figure created by the authors ） 3.2.1 Knowledge Reconstruction: Comprehensive Knowledge Integration and Curriculum Co-development UCB has established the School of Computing, Data Science, and Society (CDSS), introducing an undergraduate data science program centered on five foundational courses: Data8 (Foundations of Data Science), Data100 (Principles and Techniques of Data Science), Data140 (Probability Theory for Data Science), Data104 (Data Ethics and Humanities), and Data102 (Data, Reasoning, and Decision Making). This core curriculum integrates lowerand upper-level competencies, emphasizing computational thinking, analytical reasoning, and practical problem-solving in realworld contexts, which comprises three core components: (1) the Domain Emphasis (DE), where senior data science majors undertake DE courses that apply the ‘Data+’ paradigm through discipline-specific applications; (2) the Feature Modular Curriculum, which offers tiered data science modules for non-majors, structured around a ‘Disciplinary Adaptation + Progressive Cultivation ’model; and (3) Connector Courses, designed to bridge core data science tools with domainspecific challenges, thereby creating applied learning pathways for students with basic data literacy. 3.2.2 Organizational Innovation: Collaborative Network and Resource Sharing CDSS fosters cross-institutional collaboration through strategic academic partnerships, enhancing resource integration across four key areas: (1) Joint management of Electrical Engineering and Computer Science (EECS) programs with the College of Engineering improves faculty mobility, curriculum alignment, and collaborative research efforts; (2) A partnership with the Social Sciences Division has established the Social Science Data Lab (D-Lab), providing tailored data infrastructure, methodological training, and technical support for social science researchers; (3) The Computational Precision Health (CPH) initiative with the University of California, San Francisco (UCSF) integrates UC Berkeley’s expertise in computational statistics with UCSF ’s clinical informatics, advancing cross-campus research and co-developed teaching programs; and (4) CDSS supports transdisciplinary research through institutes such as the Bakar Institute of Digital Materials for the Planet and the Institute for Data Science, enabling cross-domain collaboration. Cross-campus faculty appointments ensure alignment between teaching and research priorities for CDSS-affiliated scholars. 3.2.3 Ecosystem Support: Cultural Foundations and Program Development CDSS advances equity, accessibility, and human-centered values through interdisciplinary integration with fields such as public policy, health sciences, education, social work, law, and clinical medicine. This vision is realized through inclusive programs, including the Data Scholars initiative and Berkeley Unboxing Data Science, alongside strategic academic partnerships. These efforts address historical disparities in data science access by providing learning opportunities for underrepresented faculty and students. Central to these efforts, the Data Science Education Programme (DSEP) incorporates innovative teaching through its Connector Course Series and Modular Curriculum, focusing on three key pillars: (1) practical project development in interactive programming environments, (2) training in statistical analysis, and (3) implementation of scalable educational technologies. DSEP’s infrastructure employs tools like the Otter autograder and MOOC platforms, ensuring pedagogical robustness and academic excellence. 4 DISCUSSION AND CONCLUSIONS 4.1 Conclusions Based on the tripartite framework, a comparative analysis of the models is provided in Table 1. We observe that ZJU and UCB have adopted distinct approaches—top- down and bottom-up, respectively. ZJU’s model prioritizes technology (AI) as its core driver, initiating the process by cultivating advanced technical competencies and systematically extending these into specific application domains to ultimately drive systemic transformation. In contrast, UCB’s approach treats foundational general competencies (data skills) as its cornerstone, emphasizing broad data literacy and establishing collaborative frameworks to facilitate transdisciplinary integration, thereby driving innovation to address complex challenges. Table 1Comparative Analysis of Cultivation Models: UCB vs. ZJU The ZJU：AI-Centric Model (TopDown Transformation) The UCB：Data-driven Approach (BottomUp Restructurions) Knowledge Reconstruction Created a hierarchical discipline structure with vertical integration to facilitate multilevel transdisciplinary connections. Introduced three course typologies to enable bidirectional knowledge exchange between data and non-data science disciplines. Organizational Innovation Implemented a hub-and-spoke model with dual mechanisms: elite oncampus training and industryacademia partnerships. Established resource-sharing protocols through multi-stakeholder partnerships to maximize educational resource efficiency. Ecosystem Support Fostered a core academic culture while building a technology-driven educational ecosystem. Developed support frameworks integrating institutional mechanisms and technological infrastructure to ensure equitable access. ZJU has established an AI-Centric talent development framework utilizing a topdown, three-tier model. The technical tier emphasizes core AI competencies, including generative modelling and reinforcement learning. The application tier integrates AI methods into engineering disciplines, such as mechanical engineering and materials science, promoting transdisciplinary innovation through AI+Field collaboration. The system tier transforms traditional engineering systems by shifting operational paradigms from localized optimization to comprehensive system redesign, the model is shown in Figure 4. In contrast, UCB adopts a bottom-up, datadriven approach, similarly structured around three tiers. The foundation tier develops cross-disciplinary skills in data collection, cleaning, and analysis, prioritizing statistical reasoning and data visualization as universal communication Fig 4 ZJU's Artificial Intelligence-based Positive Triangle Talent Cultivation Model Fig 5 UCB's Inverted Triangle Talent Cultivation Model Based on Data Thinking tools. The integration tier fosters collaboration across the data lifecycle, spanning acquisition, modelling, validation and iteration, as exemplified by digital twin experiments and knowledge graph development. The innovation tier employs datadriven decision-making to address complex engineering challenges—such as supply chain optimization and environmental systems modelling, facilitating a transition from traditional methods to intelligent, data-driven systems,the model is shown in Figure 5. 4.2 Implications This study posits that modern engineering education has entered a transdisciplinary innovation phase, necessitating the integration of data science and AI as pivotal transformative drivers. This shift unfolds through two synergistic mechanisms: (1) strengthening foundational capabilities through the development of data infrastructure, and (2) expanding disciplinary boundaries through AI-driven innovation. These mechanisms constitute complementary components of an educational ecosystem, wherein data-driven networks and AI-powered growth mutually reinforce advancement. The paper examines three dimensions to provide conceptual guidance for developing future-ready, data-proficient professionals. The first dimension underscores the development of adaptive knowledge systems and the reconceptualisation of educational approaches. AI-driven educational innovation has shifted from progressive enhancements to a fundamental restructuring of knowledge frameworks. This transformation prioritises the creation of integrated knowledge networks that prepare students for unpredictable challenges. Higher education must evolve from a focus on discipline-specific content delivery to the development of multifaceted problem-solving competencies, demanding a transition from linear knowledge dissemination to interconnected structures. The second dimension focuses on restructuring institutional boundaries to facilitate the transformation of educational ecosystems. Institutions must leverage their strengths to transcend conventional constraints, integrating disciplines while aligning technical training with societal imperatives. Key actions include dismantling disciplinary silos, delineating societal objectives, establishing robust evaluation systems, and creating adaptable academic structures. The third dimension enhances industry–academia collaboration through integrated data systems, aligning educational practices with professional requirements. This model combines theoretical learning with practical training to develop professionals proficient in computational analysis and engineering applications. (1) co-developing curricula in collaboration with industry partners; (2) incorporating real-world R&D projects into coursework; and (3) implementing blended-learning approaches that merge computational thinking with engineering problem-solving, establishing an iterative enhancement cycle that links education, practice, and industry feedback. In conclusion, we argue that the ZJU model is particularly well-suited for universities prioritising specific disruptive technologies as part of their core strategic objectives, given their strong technological specialization and high entry barriers. In contrast, the UCB model is well aligned with institutions that regard data as a foundational resource and decision-making cornerstone, aiming to tackle complex cross-domain challenges while fostering broad innovation capabilities and adaptability. As datadriven transformation continues to reshape the concepts, content, and methodologies of engineering education—progressively shifting it from closed systems to open ecosystems, and from traditional models to integrated and innovative paradigms—we will persist in investigating future-oriented talent cultivation to realize its broader societal impact.