Digital twin technology in engineering education: a systematic review

Background Digital twin (DT) technology is widely adopted in engineering practice but remains inconsistently defined and unevenly implemented in engineering education. Prior studies frequently conflate digital models, digital shadows, and bidirectionally synchronized digital twins, limiting conceptual clarity and cross-study comparability. As DTs expand into cyber-physical learning environments, clearer educational framing and evidence of learning impact are needed. Purpose This review characterizes how DT technologies are conceptualized, implemented, and evaluated in engineering education and identifies gaps in technical fidelity, pedagogical alignment, and assessment rigor. Scope/Method: A systematic literature review was conducted following PRISMA framework. Scholarly publications published between 2017 and 2025 were identified using structured database searches and screened using predefined inclusion and exclusion criteria. The final corpus of 101 studies was analyzed for disciplinary context, DT taxonomy level, synchronization architecture, pedagogical function, and learning assessment approach. Results DT-related publications in engineering education increased sharply after 2021, with most implementations situated in manufacturing, mechanical, and civil engineering. Most studies employed partial or unidirectional DT architectures rather than fully synchronized twins. Instructional uses focused primarily on visualization, monitoring, and system exploration. Few studies employed validated instruments to measure learning outcomes, and most designs remained case-based or exploratory. Recurrent barriers included infrastructure constraints, limited faculty expertise, interoperability challenges, and ethical and governance concerns. Conclusions Current educational DT implementations exhibit rapid growth but substantial methodological heterogeneity. A unified DT classification framework that aligns technical fidelity with instructional intent is needed to support rigorous design, evaluation, and scaling of DT-enabled learning environments in engineering education.