Advancing high-stakes workforce training and learning with emerging digital twin technologies

Purpose This research aims to map the global landscape of digital twins (DTs) in high-stakes training. It identifies a potential gap where technical development appears to outpace instructional integration and outlines directions for future academic inquiry. Design/methodology/approach The researchers conducted a bibliometric analysis. The data set includes 611 peer-reviewed documents from the Scopus database. The study uses VOSviewer for visual mapping. Techniques include co-authorship, citation analysis, bibliographic coupling and keyword co-occurrence mapping. These methods visualize the intellectual topography and thematic clusters of the field. Findings The analysis identifies five dominant clusters: algorithmic core, socio-technical frontier, precision validation, industrial backbone and human interface. Human-centric nodes remain comparatively peripheral to dense technical clusters. These peripheral nodes include concepts such as perception, training and competency. The digital architecture of DTs appears to be advancing rapidly, while standardized pedagogical frameworks for workforce productivity remain fragmented in the literature. Research limitations/implications This study relies solely on the Scopus database and English-language publications. Future research should further examine instructional fidelity and explainable AI to better understand the relationship between technical simulation and human cognitive load. Practical implications Human resource development practitioners may consider competency-based training models that integrate real-time DT data. The findings also suggest potential for DTs to evolve as adaptive instructional agents that could support the identification and development of individual skill gaps. Originality/value This study maps the evolving DT research landscape, highlighting key themes and the emerging need for greater integration of technical and pedagogical dimensions.