Search

The importance of workers' social identity in shaping their safety behaviors has become increasingly acknowledged. Among various social identities (e.g., workgroup, trade, etc.), improving project identity stands out as particularly impactful; it not only promotes safer behaviors among workers but also establishes a shared identity that unites all project members, thereby enhancing their sense of belonging. However, existing management strategies, primarily designed for traditional, long-term organizations, often require substantial managerial efforts to build and maintain high-quality management-employee interactions or organizational reputation. Such strategies may be less effective and economically impractical in the temporary nature of construction projects, suggesting a notable gap in management strategies that effectively foster workers' project identity to enhance safety behaviors. To fill this knowledge gap, the authors aim to develop and evaluate affordable and easy-to-implement managerial interventions that foster workers' project identification and safety behaviors. Drawing on social identity theory, the authors designed project identity-promoting messages and symbols that could be easily embedded into everyday items, such as posters and T-shirts. To evaluate their effectiveness, longitudinal field experiments were conducted, gathering 124 self-reported surveys on project identification and safety behavior from 31 workers before and after the interventions at two separate construction sites over three months. The survey data were analyzed using repeated measures (RM) ANOVA analysis to examine changes before and after the intervention. The results indicate significant improvements in both affective and behavioral dimensions of project identification, as well as in safety participation, demonstrating the potential of social identity–based interventions in improving workers' safety behaviors and project identification. This study contributes to the body of knowledge on construction organizations and construction safety by developing and evaluating practical managerial interventions based on social identity theory, which enhance construction workers' project identification and safety behavior. [ABSTRACT FROM AUTHOR]

The high occurrence of work-related musculoskeletal disorders (WMSDs) in construction remains a pressing concern, causing numerous nonfatal injuries. Preventing WMSDs necessitates the implementation of an ergonomic process, encompassing the identification of ergonomic problems and corresponding solutions. Finding ergonomic problems and solutions within active construction sites requires significant efforts from personnel possessing ergonomics expertise. However, ergonomic experts and training programs are often lacking in construction. To address this issue, the authors applied deep learning (DL)–based explainable image captioning to identify ergonomic problems and their corresponding solutions from images that are prevalent in construction sites. To this end, the authors proposed a vision-language model (VLM) capable of identifying ergonomic problems and their solutions, aided by data augmentation. The bilingual evaluation understudy (BLEU) score was used to measure the similarity between ergonomic problems and solutions identified by the proposed VLM and those specified in an ergonomic guideline. Testing with 222 real-site images, the proposed VLM achieved the highest BLEU-4 score, 0.796, compared with the traditional convolutional neural network-long short-term memory and a state-of-the-art VLM, the bootstrapping language-image pretraining. In addition, the authors developed an explainability module, visualizing which specific areas of the images the proposed VLM focuses on when identifying ergonomic problems and the important words for identifying ergonomic solutions. The highest BLEU score and the visual explanations demonstrate the potential and credibility of the proposed VLM in identifying ergonomic problems and their solutions. The proposed VLM and explainability module greatly contribute to implementing the ergonomic process in construction, identifying ergonomic problems and their solutions only with site images. Practical Applications: To prevent WMSDs, the National Institute of Occupational Safety and Health (NIOSH) recommends implementing an ergonomic process, which encompasses ergonomic problem identification, ergonomic risk assessment, and ergonomic solution identification. The current practice on sites relies on the intermittent implementation of manual ergonomic processes, and thus often falls short in protecting workers against WMSDs due to rapidly changing site conditions and the lack of on-site ergonomic expertise. Addressing this, many automated tools have been developed for ergonomic risk assessment, but none for ergonomic problem and solution identification. Therefore, with these assessment tools, we aim to streamline the recommended ergonomic process in an automated manner. To this end, we propose a deep learning–based explainable image captioning model for automated ergonomic problem and solution identification. Utilizing an ordinary camera (e.g., smartphones and site surveillance cameras), safety managers can easily identify ergonomic problems, assess risk levels, and identify corresponding solutions. Additionally, our model provides justification for its identification by visualizing the reason behind the identified ergonomic problems and solutions. With such an easily accessible and trustworthy model, the on-site ergonomic process can be streamlined, potentially reducing workers' WMSDs. [ABSTRACT FROM AUTHOR]

Wearable biosensors, in conjunction with machine learning, have been employed to develop less invasive monitoring techniques for assessing stress among construction workers during fieldwork. However, existing techniques face limitations in terms of scalable field application due to their subject and context dependency; it is difficult to apply them to new people in new contexts without additional labeled data collection. Therefore, this study developed a stress detection technique that incorporates domain adaptation, simultaneously learning a classifier and a subject- and context-independent features, in this way advancing generalizability. The proposed technique consistently demonstrated superior accuracy compared with benchmarks in classifying stress levels within a testing data set whose subjects and contexts were different from those of training data sets. Thus, the technique can advance generalizability across subjects and contexts. This finding can help us to reliably detect stress for new people in new contexts without additional labeled data collection, thereby contributing to scalable field application of wearable-based stress monitoring at construction sites. [ABSTRACT FROM AUTHOR]

Contact-driven accidents involving actuated excavators have led to a significant number of fatalities in the construction industry. The revolving mechanical arm of excavators poses a major risk of contact-driven accidents for workers in its proximity due to its articulated pose. Detecting the 3D pose of excavator arms is thus essential to prevent contact-driven accidents near excavators. Previous works have attempted to estimate 3D excavator poses using sensor-based or computer vision-based methods. However, existing methods require extensive preparation work, such as attaching physical sensors, calibrating stereo cameras, or collecting 3D training data. As a result, existing methods cannot be easily integrated into the current construction workflow and are seldom applied in real-world situations. The authors propose a projection-based 3D pose optimization method that utilizes excavator kinematic constraints to infer 3D excavator poses from monocular image sequences with no dependency on 3D training data. The proposed method first extracts the 2D excavator pose from images using a keypoint region-based convolution neural network. Then, the 2D pose is reconstructed into 3D by enforcing the rigid excavator kinematic constraints (e.g., arm length) and minimizing the 2D reprojection error of the excavator pose. Tests using a 1:14 miniature excavator model showed a 3D position error of 7.3 cm (or 1.03 m when scaled up to real-world dimensions) for keypoints on the excavator pose, demonstrating the capabilities of the proposed method in estimating 3D excavator poses from monocular images. The proximity measuring capacity of the estimated 3D pose was also evaluated, achieving a mean absolute distance error of 4.7 cm (or 0.66 m scaled). The proposed method offers a 3D excavator pose estimation method using only a monocular camera and without relying on 3D training data. The estimated 3D excavator pose enables safety managers to monitor potential contact-driven accidents near excavators and alert workers of unsafe situations and promotes safer working environments for construction workers near excavators. The authors present a monocular vision-based method for 3D excavator pose estimation, which can serve as the groundwork for monitoring contact-driven accidents near excavators. The method allows safety managers to monitor the 3D pose of an excavator using one single camera (e.g., smartphone, site surveillance camera, or action camera). Unlike previous methods, the proposed method does not require construction professionals to conduct challenging preparation work such as setting up multiple stereo cameras or collecting custom 3D excavator training datasets. For instance, for an excavator 21 m away, the proposed method can estimate the 3D excavator's keypoint positions with an expected 3D position error of 1.03 m. Paired with other onsite information (e.g., utility line or worker location) collected from existing drawings or using other vision-based methods, the proposed method can measure the proximity between the excavator arm and surrounding objects with an expected error of 0.66 m, enabling proactive safety interventions (e.g., proximity alerts). Such safety interventions can make it safer for workers to conduct construction work near excavators. Though originally devised for excavators, the proposed method can be further adapted to monitor the 3D pose of other construction equipment such as backhoes or crawler cranes, enabling a wider range of safety applications for monitoring contact-driven accidents. [ABSTRACT FROM AUTHOR]

Because workers' abnormal psychophysiological responses (e.g., high levels of stress and fatigue) are directly or indirectly linked to disorders and accidents at construction sites, monitoring workers' abnormal psychophysiological responses during ongoing work enables preventive interventions, thereby improving their health and safety. As such, wearable biosensors (e.g., wristbands) have been extensively applied with machine-learning models in construction fields as a means of continuous and less-invasive psychophysiological monitoring. However, there is a significant knowledge gap in how to validate machine-learning models that monitor human responses from biosignals. Specifically, despite the importance of generalizability across different people and contexts for psychophysiological monitoring tasks, current validation methods do not ensure different subjects and contexts between training and testing data sets, and thus overestimate the generalization performance of models. To address this issue, the authors propose a new independent subject and context testing method, leave-one-subject-and-context-out cross validation (LOSCOCV), which ensures that training and testing data sets are collected from different subjects and contexts. The proposed LOSCOCV method's generalizability estimation performance was compared with current validation methods through conducting a test wherein machine-learning models were developed to detect construction workers' stress levels from biosignals collected during their ongoing work. The proposed LOSCOCV method showed statistically lower errors in estimating machine-learning models' generalizability than other benchmarks. The results indicate that LOSCOCV is more valid than current validation methods in assessing models' generalizability for tasks that monitor human responses from biosignals. Accurately tracking generalization performance is fundamental to efforts toward advancing the generalizability of models. This study therefore significantly contributes to the field's use of biosensors and machine learning to monitor construction workers' psychophysiological responses—ultimately advancing their health and safety. [ABSTRACT FROM AUTHOR]

• A research model on construction workers' safety participation was developed. • The research model explains the psychological mechanism of safety participation. • Transformational leadership (TL) has a positive effect on safety participation. • Communication climate (CC) positively affects safety participation. • Workers' project identification mediates the effects of TL and CC. Introduction: Safety participation has gained increasing attention as an important dimension of workers' safety behaviors. Although previous studies attempted to identify factors affecting workers' safety participation, only a few studies paid attention to the psychological mechanisms behind it. Therefore, this study aimed to develop and test a research model that explains how management factors are implicated in workers' safety participation. Specifically, this study focused on project-based organizations (e.g., construction projects) because employee psychological mechanisms may have a unique nature in such transient employment. Method : The hypotheses in the research model of the psychological mechanism of construction workers' safety participation are tested using survey data from 261 construction workers. Results : The results indicated that construction workers' safety participation is influenced by project identification after controlling the shared variance of safety compliance. Project identification also mediates the effects of transformational leadership and communication climate on safety participation. Practical Applications : This study offers researchers and practitioners an explanation of how management factors influence construction workers' safety behaviors and clarifies the role of project identification play in explaining the effects of management factors on safety compliance and safety participation. [ABSTRACT FROM AUTHOR]

Recent advancements in wearable health devices equipped with biosensor systems (e.g., heart rate (HR) sensor) have provided an ample opportunity to continuously measure and understand workers' physical demands from construction work. Specifically, a relative measurement of physical demands, which is a percentage of HR reserve (%HRR), is convenient and useful by normalizing individual differences of HR. Since affordable HR monitoring using wearable devices (particularly, a comfortable wristband-type device: wristband hereafter) becomes available, %HRR-based physical demand measurement, which can be continuously calculated without interfering with workers' ongoing work, provides an enormous potential to protect workers' safety and health and to sustain expected productivity. This research investigates the usefulness of affordable %HRR-based physical demand measurement using a wristband from a case study of 19 workers in construction sites. The aim of the analysis is to examine the potential of this continuous measurement in capturing any significant physical demand variations, by investigating in-depth information on factors affecting physical demands (e.g., work tasks, individual and environmental factors). The results show that workers' physical demands are highly variable according to their working patterns (i.e., direct work, and indirect work including tool/equipment/material handling, traveling, and preparatory work), combined influences of work tasks, as well as individual and environmental factors (e.g., age and heat stress). These results demonstrate the need for continuous physical measurement during workers' ongoing work so that any significant high physical demands, which need to be avoided if possible, can be captured. The findings of this paper show that the continuous measurement of physical demands using a wristband provides rich information to understand, manage, and design physically demanding construction work (e.g., flexible work-rest cycle and managing demanding indirect work) by balancing workloads throughout a day and/or reducing unnecessary physical demands beyond direct work. By anticipating potential health and safety problems from excessive physical demands, as well as productivity loss before they occur, this research will have an ameliorative impact across the construction industry. [ABSTRACT FROM AUTHOR]

Construction projects are usually labour intensive, and human resource (HR) issues contribute significantly to a project's final costs. From this perspective, a tool that can help construction managers reduce their HR costs can potentially generate improvement in the project cost. In this paper we propose a simulation-based approach that sheds light on the dynamics of workforce skill evolution as the project progresses, thereby assisting construction managers in adjusting their HR policies. The proposed approach uses a system dynamics (SD) simulation model that dynamically tracks the effects of alternative HR policies. After the development and validation of the SD model, the SD model is extended to capture operational details and their interaction with workforce skill evolution, adopting a hybrid SD and discrete event simulation (DES). The hybrid model has been applied to an experimental case of structural steel fabrication projects, in which we demonstrate that there is a considerable room for cost-saving in HR. The hybrid modeling approach introduced in this paper can be employed by construction managers for possible improvements in HR management, as well as researchers for an in-depth understanding of the dynamics in workforce skill evolution. [ABSTRACT FROM AUTHOR]

This paper introduces BlendCon, a fully automated framework capable of simultaneously synthesizing and labeling construction imagery data. This framework simulates a construction site by orchestrating 3D mobile objects against a 3D background and produces multimodal labels for target entities. The effectiveness of the synthetic data in training object detection models was thoroughly validated. For the construction worker detection task, a YOLOv7 model trained with synthetic data nearly matched the performance of a model trained with real data: it achieved 71% AP@0.5–0.95 compared to 75% for the real data-trained model. Moreover, the model trained with synthetic data surpassed its real data counterpart in scenarios requiring stricter IoU thresholds, particularly above 85%. Acquiring a sufficient quantity and diverse range of imagery data has been a primary challenge in construction studies that focus on automation and digitization through deep neural networks. BlendCon can significantly contribute to addressing this data scarcity challenge. • BlendCon: automated framework to synthetize and label construction imagery data. • It generates a wide spectrum of construction images with varying imaging condition. • Its labels include 2D/3D bboxes, 2D/3D poses, segmentation mask, and depth map. • It successfully produced 985,660 images within 2 days, proving its usability. • 71% AP@0.5–0.95 on worker detection task only with synthetic data (75% for real). [ABSTRACT FROM AUTHOR]

Construction workers are at high risk of work-related musculoskeletal disorders (WMSDs) due to physically demanding manual-handling tasks in awkward postures. Although existing observational methods to identify ergonomic risks are inexpensive and easy to use, they are seldom used in construction sites because they are time-consuming, subject to observer bias, and require well-trained analysts. To address these drawbacks, this paper proposes a vision-based method to automatically classify workers' postures for ergonomic assessment. Specifically, it proposes a vision-based method that eliminates the need to collect extensive training-image datasets by employing classification algorithms to learn diverse postures from virtual images, and then identifies those postures in real-world images. The experimental tests showed about 89% classification accuracy in automatically classifying diverse postures on images, confirming the usefulness of virtual training images for posture classification. The proposed method has potential for automated ergonomic risk analysis, and could help to prevent WMSDs during diverse occupational tasks. • A computer vision-based method for automated postural ergonomic risk assessment is proposed. • The method identifies risky postures on images by learning image patterns according to different body postures. • The method shows robust posture classification accuracy comparable with the accuracy attained by human observers. [ABSTRACT FROM AUTHOR]

Abstract Struck-by accidents have resulted in a significant number of fatal and nonfatal injuries in the construction industry. As a proactive safety measure against struck-by hazards, the authors present an Unmanned Aerial Vehicle (UAV)-assisted visual monitoring method that can automatically measure proximities among construction entities. To attain this end, this research conducts two research thrusts: (i) object localization using a deep neural network, YOLO-V3; and (ii) development of an image rectification method that allows for the measurement of actual distance from a 2D image collected from a UAV. Tests on real-site aerial videos show the promising accuracy of the proposed method; the mean absolute distance errors for estimated proximity were less than 0.9 m and the mean absolute percentage errors were around 4%. The proposed method enables the advanced detection of struck-by hazards around workers, which in turn can make timely intervention possible. This proactive intervention can ultimately promote a safer working environment for construction workers. Highlights • A computer vision method for UAV-assisted remote proximity monitoring is presented. • A CNN-based localization, YOLO-V3, is applied for robust object localization. • An image rectification method is developed for efficient distance measurement. • A test on a real-site video illustrates the promising accuracy: around 4% of MAPE. • The method can provide the advanced detection of struck-by hazards around workers. [ABSTRACT FROM AUTHOR]

Construction workers' attitudes and behaviors are one of the most important factors of a construction project's performance. As the attention paid to the impact of social influences and norms on worker behavior grows, agent-based modeling and simulation (ABMS) emerges as a research tool for studying workers' group behavior. With ABMS, researchers can uncover the underlying process of group behavior emerging from individuals' interactions in an organization. However, validating agent-based simulation with real data is the greatest challenge in using ABMS for organizational behavior research. With this background in mind, the objective of this paper is to propose a methodology for creating an empirically supported agent-based model for studying workers' behavior influenced by social norms. The proposed methodology suggests that empirical data collected by a questionnaire can be used for ABMS in three steps: (1) testing the agent behavior rules used in an agent-based model (i.e., testing the modeling assumptions), (2) demonstrating the model behavior's qualitative agreement with real workers' behavior (i.e., testing the simulation results against real data in a qualitative manner), and (3) creating a specific agent-based model with the model parameters that correspond to a specific empirical case. A specific agent-based model created in this way can then be seen as a scenario generator that corresponds to a specific reality and can be used to answer what if questions. Therefore, the model can be used to develop policies/interventions to improve workers' behavior in a given situation. The proposed methodology is illustrated by a study on construction workers' absenteeism that was conducted by the authors. This paper contributes to the body of knowledge of workforce management in construction; the proposed methodology provides a means of simulating workers' group behavior and developing policies/interventions to improve worker behavior at the group level in construction. [ABSTRACT FROM AUTHOR]

The use of wearable sensing devices (e.g., GPS and physiological sensors) can open a new door toward occupational safety and health management in physically demanding and hazardous construction (e.g., tracking worker's locations in dangerous working area and monitoring of worker's physiological status). These potential benefits cannot be achieved, however, if construction workers do not recognize the value of wearable devices as well as the details of their usage. This study thus investigates determinants for workers' adoption of wearable technology in the occupational work context. Specifically, the scope of this study is to test hypotheses regarding workers' intention to adopt two representative wearable devices for occupational safety and health, a smart vest with an embedded indoor GPS for location tracking, and a wristband-type wearable activity tracker (i.e., wristband) with physiological sensors. The research results indicate that perceived usefulness (PU), social influence (SI), and perceived privacy risk (PR) are associated with workers' intention to adopt (IA) both smart vest and wristband. Also, workers' experiences using wearable devices positively moderates the association between PU and IA of smart vest and negatively moderates the association between SI and IA of smart vest. In the work context, foremen are more likely to be influenced by PU than workers with regard to using a wristband. By considering the different functions, benefits, and challenges of each device, and by taking into account individual and job characteristics, the results of this study provide crucial insight into the process of motivating workers to adopt each device in their work, which can promote the continued and appropriate use of wearable technology in occupational safety and health management. [ABSTRACT FROM AUTHOR]

Although deep neural network (DNN)-powered visual scene understanding is a driving factor in a transition toward construction digitalization and robotic automation, a shortage of construction training images has been a roadblock to achieving DNNs' maximum performance potential. This data shortage becomes more problematic in digitally monitoring field workers who perform a variety of activities in an unstructured outdoor construction environment. To address this issue, the authors present a construction worker-centric image synthetization approach that can automatically synthesize and label limitless artificial human images with diverse poses, activities, and outdoor imaging conditions. Using synthesized construction worker-centric images, the authors conduct training experiments to characterize the effects of synthetic images on DNN-powered worker detection. In addition, the authors explore the hybrid effects of synthetic and real images on DNN performance. Results showed that a synthetic image-trained model potentially performs well in diverse field conditions and can even detect construction workers who are missed by a real image-trained model. It was also shown that a hybrid use of synthetic and real images can reduce the number of necessary real training images by 50% and improve DNN performance by 16% on average, compared to when only one of the two data sources are adopted. Moreover, the data hybridity enabled DNNs to reach its near-maximum performance while scaling down the size of a real training dataset by up to 80%. These findings indicate that synthetic images have promising potential for worker-centric DNN training in that they enable higher performance while reducing the human effort needed for real construction image collection and labeling. This capability can help to address the problem of data shortage in construction and enable the training of more accurate and scalable DNN models. Furthermore, this will stimulate the development and implementation of visual artificial intelligence for robotic automation and digitization. • Presented a new construction worker-centric image synthetization approach • Showed the potential of synthetic construction images in training a worker detection DNN model • Synthetic images could fill in a real dataset's missing contexts • Increased the DNN performance by 16.3% while reducing the number of real images by 50%. • Human effort needed for real construction image collection and labeling can be reduced [ABSTRACT FROM AUTHOR]

With increasing concerns regarding occupational safety and health, managing excessive physical workloads of workers is critical to prevent workers' fatigue, injuries, errors, or accidents at physically demanding workplaces such as construction. In this regard, heart rate (HR) is an effective physiological indicator of workers' physical demands. Currently, off-the-shelf wearable activity trackers (e.g., wristband-type) can monitor a worker's HR with its embedded photoplethysmography (PPG) sensor. However, PPG signals can be highly affected by signal noises resulted from user's movements, and thus the exact HR extraction from a wristband-type PPG may not be sufficiently accurate during intensive construction tasks. In this paper, we investigate the accuracy of a PPG sensor embedded in a wristband-type tracker to see if it can be used for construction. Through field data collection from seven construction workers, we conduct a comparative HR analysis between a PPG sensor and an electrocardiography (ECG) sensor in a chest strap used as ground truth. The results show that a PPG-based HR sensor in a wristband-type activity tracker has a potential for practicable HR monitoring of construction workers with 4.79% of mean-average-percentage-error (MAPE) and 0.85 of correlation coefficient for whole datasets (4.44%, 4.52%, and 5.33% of MAPEs and 0.89, 0.70, and 0.61 of correlation coefficients during light works with < 90 bpm of HRs, moderate works with 90–110 bpm of HRs, and heavy works with > 110 bpm of HRs, respectively). Because there is still room for improvement of the accuracy, particularly during heavy works, we also investigate the factors affecting the accuracy of HR monitoring using inequality statistics. From this secondary investigation, we found the major sources of error including noises from motion artifacts. With advanced noise-cancellation techniques, it is expected that that field HR monitoring using wearable activity trackers can be used to evaluate worker's physical demands from diverse construction tasks in a non-intrusive and affordable way. As a result, our work will help manage excessive workloads (e.g., flexing work/rest plans) so that a worker can sustain his/her given tasks during working time in a safer and healthier way. [ABSTRACT FROM AUTHOR]

• A worker-centered process for assessing workers' heat strain is presented. • The developed data-driven models can discern heat strain levels with 92% accuracy. • A series of experiments was administered in a chamber room to simulate heat stress. • It is feasible to detect workers' heat strain in advance based on their biosignals. • The proposed process holds promise to reducing heat-related illnesses in the field. Construction workers are at a high risk of exposure to excessive heat generated by several factors such as intensive physical activities, personal protective clothing, and frequent heat events at construction sites. Previous studies attempted to evaluate the occupational risk of heat stress by concentrating on environmental variables or the self-assessment measures of perceived heat. Despite their potentials, most of these approaches were intrusive, inaccurate, and intermittent. More importantly, they mainly overlooked the disparities in workers' physical and physiological characteristics. To address these limitations, this study proposes a heat-stress risk-assessment process to evaluate workers' bodily responses to heat – heat strain – based on the continuous measurement of their physiological signals. To this end, workers' physiological signals were captured using a wristband-type biosensor. Subsequently, their physiological signals were decontaminated from noises, resampled into an array of informative features, and finally interpreted into distinct states of individuals' heat strain by employing several supervised learning algorithms. To examine the performance of the proposed process, physiological signals were collected from 18 subjects while performing specific construction tasks under three predetermined environmental conditions with a different probability of exposure to heat stress. The analysis results revealed the proposed process could predict the risk of heat strain with more than 92% accuracy, illuminating the potentials of wearable biosensors to continuously assess workers' heat strain. The long-term implications of this study can be capitalized as guidelines to improve systematic evaluation of heat strain and promote workers' occupational safety and well-being through early detection of heat strain at construction sites. [ABSTRACT FROM AUTHOR]