Digital twin publications in construction (2017–2023): a bibliometrics-based visualization analysis

This study aims to thoroughly explore and visualize the trends and developments of digital twin (DT) literature in the construction field while revealing future research directions for further exploration and exploitation.

The research follows a three-stage methodology. First, the bibliographic data is acquired using the Web of Science database. Second, the bibliometric methods are defined to include co-authorship analysis, citation analysis, keywords co-occurrence, thematic mapping while the software tools include MS Excel, VOSviewer and Biblioshiny. Third, analysis and findings include yearly DT publication output, influential DT publications, leading DT contributors, top DT sources and science mapping of DT literature.

This study identifies top-cited DT publications (35 out of 320) in terms of citations score, local citations score and document average citations per year. Furthermore, the key contributors with respect to authors (58 out of 1147), organizations (55 out of 427) and countries (19 out of 51) are recognized in terms of productivity, influence, activeness and scientific value. Similarly, the major publishing sources (24 out of 58) are identified using the same measures. Regarding science mapping, the DT domain comprises four research frontiers, namely, deep learning and smart city, internet of things and blockchain, DT and building information modeling and machine learning and asset management.

Through a mixed-review strategy, this study introduces a comprehensive analysis of DT literature while avoiding the subjectivity/cognitive bias of traditional review approaches. Moreover, it illuminates the promising and rising DT themes for new/seasoned researchers, institutions, editorial boards and funding agencies.