Abstract The rapid digital transformation in the construction industry has stimulated extensive research into utilizing intelligent construction technologies in enhancing working conditions, safety performance, and productivity among construction workers. However, to our best knowledge, a comprehensive literature review that synthesizes the current research status and hotspots in this field is lacking, resulting in a limited understanding of the research achievements and gaps on this topic. Therefore, this research conducts a bibliometric review of research into construction worker in intelligent construction. CiteSpace software was employed to visualize and analyze relevant articles retrieved from the Web of Science and Scopus databases from 2015 to 2024. A total of 167 journal articles were selected and analyzed. The results indicate that the number of publications focusing on construction worker research in intelligent construction has been increasing over the last decade, with the majority published within the latest five years. While early research mainly addressed macrolevel topics such as safety performance, productivity management, and accident prevention, recent hotspots have shifted toward microlevel technology-driven solutions, including hazard recognition, fatigue detection, hazardous object identification, and safety training. Keyword clustering analysis identified five major research themes, including building information modeling, fatigue detection, system, proactive safety management, and electroencephalography. Based on the analysis results, current research hotspots and future research directions were discussed from three integrated perspectives, including intelligent construction perspective for construction workers, construction worker-centric perspective within intelligent construction, and construction worker–intelligent technology interaction perspective. The findings provide a roadmap for researchers to identify emerging trends and prioritize future studies. Moreover, they offer valuable insights for industry practitioners and policymakers to strategically implement intelligent technologies, fostering a safer and more humancentric construction environment.
Abstract The complex nature of human–artificial intelligence (AI) interaction (HAI) introduces new challenges for worker safety. While ensuring human well-being in HAI has become a shared objective, the lack of quantifiable approaches limits our ability to capture its nuanced effects on workers. To address this gap, the present study develops an agent-based modeling framework grounded in stimulus-organism-response (SOR) theory. The model simulates workers’ interactions with AI systems and peers within a bounded two-dimensional space, tracking the resulting cognitive, emotional, and behavioral responses. It comprises three components: worker-agent design, SOR process modeling, and an interaction rule base. Experiments were conducted under three scenarios: (1) repeated simulations with balanced parameter settings, (2) batch simulations with varying parameter settings, and (3) empirical validation at a real smart construction site. Results demonstrate that workers’ safety behavior progressively deteriorates as HAI deepens, with extra-role safety behavior exhibiting the most pronounced decline. This study extends the application of SOR theory to the construction industry and deepens the understanding of HAI in the context of construction safety. Practically, it underscores the cumulative cognitive and emotional effects of HAI on workers, providing both a managerial basis for defining human–AI boundaries and a quantifiable tool for long-term safety decision-making.
Abstract The unsafe behavior of construction workers is one of the major causes of safety accidents. The ongoing digital transformation at construction sites in China is not only enhancing operational efficiency but also shaping the spread of unsafe behaviors among construction workers. Therefore, it is important to study the transmission mechanism of workers’ unsafe behavior under the background of information technology to reduce the occurrence of safety accidents. This study integrates information technology into research on unsafe behaviors among construction workers. It employs the decision-making trial and evaluation laboratory method to identify key influencing factors and constructs a susceptible exposed infected recovered (SEIR) model to decipher the underlying mechanisms through which these factors propagate such behaviors. Based on the findings, targeted interventions are proposed. The results demonstrate that, when the intensity coefficient of digital construction technologies increases from 0.1 to 0.3, the peak infection rate and steady-state infection proportion of unsafe behaviors among construction workers decrease by 9.7% and 14%, respectively. As the infection coefficients of network communication technology and intelligent sensing technology rise from 0.2 to 0.6, the infection peak declines by 9%, while the peak steady-state immune population increases by 70.6%, indicating a restraining effect on the spread of unsafe behaviors. In contrast, stronger conformity psychology, closer coworker intimacy, and higher authority of key figures contribute to an increase in the number of “infected” workers, thereby promoting the propagation of unsafe behaviors. The findings of this study provide a theoretical basis for developing control strategies to mitigate the spread of unsafe behaviors. However, as the research is simulation-based, further validation with real-world data is necessary in future work.
Abstract Hearing protection devices are essential for protecting the health of construction workers in job sites with loud noise; however, they may hinder workers’ ability to detect safety-critical alarms. Although previous research has demonstrated the potential of generative artificial intelligence (AI) models to enhance safety-critical sounds in construction work zones, the feasibility of real-time implementation of such models on wearable AI devices remains unclear. This study addresses that gap by developing and testing a novel AI-augmented hearing protector (AI-AHP) that deploys a real-time safety sound enhancement model on a low-cost Raspberry Pi 5 platform. The model was trained using various machine learning approaches on a data set comprising 4 types of safety-critical warning signals and 38 types of background noise. Controlled functionality testing, field deployment at active construction sites, and a subjective listener survey were conducted to evaluate the effectiveness of the prototype. The study identified DeepFilterNet as the most efficient model for edge computing, achieving fast and substantial improvements in signal audibility, with a maximum Perceptual Evaluation of Speech Quality (PESQ) improvement of 179%. Real-time operation was confirmed with a real-time factor of less than 1 and an end-to-end latency of 0.69 s on the Raspberry Pi 5. Subjective evaluation results showed significant increases in listener ratings, indicating that the enhanced signals were consistently perceived as clearer and more intelligible. These findings demonstrate the feasibility of deploying AI-driven hearing protection devices that improve worker situational awareness, reduce accident risk, and maintain usability in noisy construction environments, with potential applicability to other industrial settings.
Abstract On-site quality assessment of precast concrete (PC) components is essential to ensure structural integrity, but precamber measurement relies on labor-intensive manual procedures due to time constraints and site congestion. This study proposes an automated precamber estimation framework for PC components using two-dimensional (2D) images. The proposed method integrates a reference-object-free geometric estimation, which utilizes predefined design specifications as ground truth, with vision transformer (ViT)-embedded architecture trained on hybrid synthetic data to robustly capture high-level features. Field validation demonstrated that the framework achieves an average measurement error of 1.91 mm, confirming a reliable performance comparable to manual methods. Notably, the automated process reduced inspection time by 79.82% and labor costs by 91.70% compared to traditional manual procedures. These results demonstrate the practical feasibility of the proposed framework for supporting rapid and efficient on-site quality assessment. Furthermore, the framework holds significant potential to extend into a comprehensive dimensional inspection system for managing various geometric deformations beyond precamber.
