Key points are not available for this paper at this time.
Purpose To address the challenges of limited research methodologies and weak theoretical foundations in the field of intelligent construction for large-scale infrastructure, this study employed CiteSpace combined with Delphi method. By integrating technology diffusion theory, a three-tiered framework was constructed to fill up the research gap in this domain. Design/methodology/approach This study examined literature on intelligent construction of large-scale infrastructure indexed in the Web of Science database from 1998 to 2025. 981 valid documents were obtained via multiple rounds of screening. CiteSpace software was employed for co-citation analysis, keyword clustering and co-occurrence analysis. Clustering results were validated through two rounds of the Delphi method. A three-tier research framework was constructed by integrating technology diffusion theory, and theoretical-practical validation was achieved through a case study of an airport terminal building. Findings This study revealed that research on intelligent construction technologies in large-scale infrastructure developed rapidly from 2012, with publication output over the past five years surpassed the preceding two decades quadrupled. Core literature primarily focused on BIM and digital twin-related researches. Three major future development trends have been identified: life cycle closed-loop data management, deep integration of intelligent construction technology clusters and sustainable intelligent construction. Additionally, this study also involved implications on research, practice, policies and society. Research limitations/implications First, the data source is single, relying solely on the WoS database, which mainly covers Europe, China and North America. This imbalance in geographical distribution may affect the comprehensiveness and objectivity of the research to some extent. Second, there is a potential bias in citation analysis, as most highly cited literature comes from European and American journals, which may downplay the technological innovation contributions of non-English-speaking countries. Finally, the keyword unification process carries the risk of semantic simplification, ignoring the subtle differences in application scenarios of different terms, which may obscure the hierarchical differences of research themes. Practical implications (1) Industrial practice: Enterprises are advised to build “BIM + IoT + AI” integrated platforms for real-time data integration across design-construction-operation; adopt modular intelligent technologies, such as prefabricated components with RFID chips in intercity railways, to enhance efficiency and precision. (2) Policy suggestions: Governments should invest in national building data sharing platforms and edge computing clusters, establish dual standards for human–machine safety and carbon accounting, and provide tax incentives for smart low-carbon projects. Social implications (1) Safety and well-being: Intelligent equipment replaces high-risk tasks, reducing accident rates via AI behavior recognition; modular construction minimizes on-site operations, decreasing dust and noise pollution. (2) Low-carbon sustainability: Green materials combined with intelligent energy management enable precise carbon emission control; digital twin technology integrates dynamic data to optimize carbon footprint prediction for infrastructure like airports. Originality/value This study innovatively integrated CiteSpace bibliometrics and Delphi validation method and proposed a novel perspective of “theoretical support-practical verification” dual-driven knowledge graph of intelligent construction for large-scale infrastructure. Different from existing single technology focused reviews, this study combined the technology diffusion theory and proposed a three-layer framework of “knowledge dissemination-technology development-practical application”. This framework can overcome the fragmented limitations of existing studies and achieved the systematic connection of theory, technology and practical engineering. Furthermore, this study innovatively proposed three future research trends, which exerts a profound impact on the in-depth application of intelligent technologies globally.
Liu et al. (Sat,) studied this question.