Construction Jobsites Research Articles

Construction Jobsite Image Classification Using an Edge Computing Framework

Image classification is increasingly being utilized on construction sites to automate project monitoring, driven by advancements in reality-capture technologies and artificial intelligence (AI). Deploying real-time applications remains a challenge due to the limited computing resources available on-site, particularly on remote construction sites that have limited telecommunication support or access due to high signal attenuation within a structure. To address this issue, this research proposes an efficient edge-computing-enabled image classification framework for support of real-time construction AI applications. A lightweight binary image classifier was developed using MobileNet transfer learning, followed by a quantization process to reduce model size while maintaining accuracy. A complete edge computing hardware module, including components like Raspberry Pi, Edge TPU, and battery, was assembled, and a multimodal software module (incorporating visual, textual, and audio data) was integrated into the edge computing environment to enable an intelligent image classification system. Two practical case studies involving material classification and safety detection were deployed to demonstrate the effectiveness of the proposed framework. The results demonstrated the developed prototype successfully synchronized multimodal mechanisms and achieved zero latency in differentiating materials and identifying hazardous nails without any internet connectivity. Construction managers can leverage the developed prototype to facilitate centralized management efforts without compromising accuracy or extra investment in computing resources. This research paves the way for edge “intelligence” to be enabled for future construction job sites and promote real-time human-technology interactions without the need for high-speed internet.

Content annotation in images from outdoor construction jobsites using YOLO V8 and Swin transformer

Digital visual data, such as images and videos, are valuable sources of information for various construction engineering and management purposes. Advances in low-cost image-capturing and storing technologies, along with the emergence of artificial intelligence methods have resulted in a considerable increase in using digital imaging in construction sites. Despite these advances, these rich data sources are not typically used to their full potential because they are processed and documented subjectively, and several valuable contents could be overlooked. Semantic content analysis and annotation of the images could enhance retrieval and application of the relevant instances in large databases. This research proposes an ensemble approach to use deep learning-based object recognition, pixel-level segmentation, and text classification for medium-level (ongoing activities) and high-level (project type) annotation of still images from various outdoor construction scenes. The proposed method can annotate images with and without construction actors, i.e. equipment and workers. The experimental results have shown the potential of this approach in annotating construction activities with an 82% overall recall rate.

Leveraging BIM for Enhanced Camera Allocation Planning at Construction Job Sites: A Voxel-Based Site Coverage and Overlapping Analysis

In the construction industry, the imperative for visual surveillance mechanisms is underscored by the need for safety monitoring, resources, and progress tracking, especially with the adoption of vision intelligence technology. Traditional camera installation plans often move toward coverage and cost objectives without considering substantial coverage overlap, inflating processing and storage requirements, and complicating subsequent analyses. To address these issues, this research proposes a voxel-based site coverage and overlapping analysis for camera allocation planning in parametric BIM environments, called the PBA approach. The first step is to collect information from the BIM model, which is the input for the parametric modeling step. After that, the PBA approach simulates the virtual devices and the construction layout by employing visual language programming and then generates a coverage area. Lastly, the performance simulation and evaluation of various placement scenarios against predefined criteria are conducted, including visual coverage and overlapping optimization for eliminating data redundancy purposes. The proposed approach is evaluated through its application to construction projects. The results from these various implementations indicate a marked decrease in data overlap and an overall enhancement in surveillance efficacy. This research contributes a novel, BIM-centric solution to visual information adoption in the construction industry, offering a scalable approach to optimize camera placement while mitigating overlapping areas.

Influence of Information and Communication Technology Application in Construction Jobsites: A Systematic Review and Bibliometric Analysis

In recent years, the construction industry has witnessed a surge in the adoption and application of emerging technologies. Substantial evidence indicates the impact of these technologies on the lifecycle of construction projects. Among the notable categories of technologies embraced in construction, Information and Communication Technologies (ICTs) stand out. This category encompasses BIM models, reality capture, cloud computing, robotics, and more. However, a literature investigation that specifically focuses on the significant influence of these technologies on construction sites is lacking. To address this gap, this study conducted a systematic literature review, supported by a bibliometric analysis of published articles in the field of construction management. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method, data was sourced from Scopus and Web of Science, covering the period from 2009 to 2023. The findings shed light on nine prominent ICT tools widely adopted on construction sites for project execution, with contributions from 113 countries emphasizing their profound impacts. Despite these positive aspects, this study also identifies notable challenges and provides recommendations for mitigating them. A smart site data integration framework was developed to facilitate the seamless integration of these technologies while ensuring data security. This study provides fundamental insights for future research in this field aimed at informing the construction industry about the importance of embracing cutting-edge technologies for enhanced efficiency on construction jobsites and for timely project delivery.

Optimization of Microphone Placement for Audio-based Modeling of Construction Jobsites

Heavy equipment is a crucial resource in the construction industry. Recent studies have shown that analyzing the sound patterns generated by construction machinery can be an effective way to monitor their performance and detect potential operational issues. However, construction jobsites are complex environments that require consideration of multiple factors when creating an audio-based model. To perform efficient audio-based modeling of jobsites, it is essential to optimize the number and placement of microphones to capture the sound emitted by all the operating machines and achieve optimal sound quality. To address this challenge, we developed two optimization methods: (a) an integer programming model that guarantees finding the optimal placement of microphones, and (b) an evolutionary programming model, a heuristic approach more suited to larger problem instances. We evaluated the effectiveness of these models in five different case studies from construction jobsites. Our results showed that the developed models require a reasonable number of microphones to achieve the desired sound quality, demonstrating their satisfactory performance. Notably, both approaches exhibited similar performance in terms of the required number of microphones needed to cover all the machinery.

Productivity Measurement through IMU-Based Detailed Activity Recognition Using Machine Learning: A Case Study of Masonry Work.

Although measuring worker productivity is crucial, the measurement of the productivity of each worker is challenging due to their dispersion across various construction jobsites. This paper presents a framework for measuring productivity based on an inertial measurement unit (IMU) and activity classification. Two deep learning algorithms and three sensor combinations were utilized to identify and analyze the feasibility of the framework in masonry work. Using the proposed method, worker activity classification could be performed with a maximum accuracy of 96.70% using the convolutional neural network model with multiple sensors, and a minimum accuracy of 72.11% using the long short-term memory (LSTM) model with a single sensor. Productivity could be measured with an accuracy of up to 96.47%. The main contributions of this study are the proposal of a method for classifying detailed activities and an exploration of the effect of the number of IMU sensors used in measuring worker productivity.

Construction Work-Stage-Based Rule Compliance Monitoring Framework Using Computer Vision (CV) Technology

Noncompliance with safety rules is a major cause of unsatisfactory performance in construction safety worldwide. Although some research efforts have focused on using computer vision (CV) methods for safety rule inspection, these methods are still in their early stages and cannot be effectively applied on construction job sites. Therefore, it is necessary to present a feasible prototype and conduct a detailed analysis of safety rules to ensure compliance at the job site. This study aims to extend the validation of safety rule analysis through four case scenarios. The proposed structured classification of safety rules includes categorizing them based on project phases and work stages. The construction phase-related rules are divided into four groups: (1) before work, (2) with intervals, (3) during work, and (4) after work. To validate the proposed framework, this research developed prototypes for each group’s scenarios using deep learning algorithms, a storage database to record compliance with safety rules, and an Android application for edge computing, which is required in the “before work” and “after work” groups. The findings of this study could contribute to the development of a compact CV-based safety monitoring system to enhance the current safety management process in the construction industry.

Wearable EEG-based construction hazard identification in virtual and real environments: A comparative study

Hazard identification on construction jobsites is one of the most critical components of construction safety management. To overcome the limitations in the current manual practice, researchers explored the correlation between hazard perception and changes in workers’ electroencephalogram (EEG) signals collected from wearable sensors to identify and differentiate different types of hazards. However, most of such studies are based on experiments in a virtual reality (VR) environment. The unproven applicability of the VR-led results to the real jobsites and the potential discrepancies in results between the physical and virtual environments are the main barriers toward EEG-based ubiquitous hazard identification on the jobsites. To bridge the research gaps, this paper tests the feasibility of identifying construction hazards in physical and VR environments, and compares their results to identify discrepancies. Two experimental environments (physical and VR, respectively), each containing three hazards, were designed with the same configuration. Hazard identification task was performed in both experiments, data (e.g., EEG signals) was collected and preprocessed, and corresponding features were extracted. Hotspot analysis was conducted to infer the existence of hazards by assessing spatial associations. The results showed that both physical and VR settings were effective for hazard detection, while a better performance was observed in the VR setting. In addition, the potential of developing hazard zones as well as particular hazard locations and a gradual increase in subjects’ perceived hazard level were revealed. The results highlight the opportunities of revealing potential construction hazards on the construction jobsite using wearable EEG.

Disclosure Level Analysis of Environmental Indicators of ESG Reports of Listed Developers

近年來永續成為全球議題，企業陸續出版ESG（environment, social and governance）或永續報告書，展現永續績效與作為。然而文獻指出，永續報告書的揭露有待加強。本研究分析建設公司ESG報告書環境指標，檢視國內十家上市建設公司的2020永續報告書的揭露，與國外五家優秀建設公司的比較。研究結果顯示，於32個指標中，十家公司平均揭露9個、揭露率29%，揭露程度17%，約為國外五家公司揭露的一半；揭露邊界多在公司辦公室，較少工地資訊，而國外公司於公司及工地資訊都揭露。揭露率不高的原因有：政府以前未強制規定、揭露未有標準，及公司考量費用。ESG報告編製人員需詳讀GRI準則，了解環境指標要求的資料，定期蒐集與分析，以提高環境指標之揭露率。Sustainability has become a global issue in recent years. More companies publish ESG (environment, social and governance) or sustainability reports to show their sustainable performances and actions. However, literature points out that the disclosure of the report needs improvement. This research analyzed the environmental disclosures of sustainability reports of ten domestic listed developers. The results were compared with the disclosures of excellent sustainability reports of five foreign developers. The research results showed that, among the 32 indicators, the ten companies disclosed nine indicators or 29% on average, and the average disclosure level was 17%. These numbers are about half of the disclosures of the foreign companies. The boundary of disclosure was more limited to the company's office buildings and the information of construction jobsites was less addressed. But the information of both office buildings and jobsites was disclosed by the foreign companies. The reasons for low disclosures can be non-mandatory publication of ESG reports, no standards for disclosures, and the company cost of ESG report preparation. The people who prepare ESG reports need to read the GRI requirements, know the required data, regularly collect and analyze them to raise the disclosure rates of environmental indicators.

Real-Time Personal Protective Equipment Compliance Detection Based on Deep Learning Algorithm

The construction industry is one of the most dangerous industries in the world due to workers being vulnerable to accidents, injuries and even death. Therefore, how to effectively manage the appropriate usage of personal protective equipment (PPE) is an important research issue. In this study, deep learning is applied to the PPE inspection model to verify whether construction workers are equipped in accordance with the regulations, and this is expected to reduce the probability of related occupational disasters caused by the inappropriate use of PPE. The method is based on the YOLOv3, YOLOv4 and YOLOv7 algorithms to detect worker’s helmets and high-visibility vests from images or videos in real time. The model was trained on a new PPE dataset collected and organized by this study; the dataset contains 11,000 images and 88,725 labels. According to the test results, can achieve a 97% mean average precision (mAP) and 25 frames per second (FPS). The research result shows that the detection and counting data in this method have performed well and can be applied to the real-time PPE detection of workers at the construction job site.

The Relationship between Work-Related Stressors and Construction Workers’ Self-Reported Injuries: A Meta-Analytic Review

Mental health concerns are surging worldwide and workers in the construction industry have been found to be particularly vulnerable to these challenges. Stress, depression, addictions, suicides, and other key indicators of poor mental health have been found to be highly prevalent among construction workers. Critically, researchers have also found a link between how stress in the workplace impacts the overall safety performance of an individual. However, the burgeoning nature of the research has stifled the determination of feasible and actionable interventions on jobsites. This paper aims to analyze the relationship between work-related stressors found on construction jobsites and self-reported injury rates of workers. To accomplish this goal, a meta-analysis methodology was used, wherein a comprehensive literature search was conducted to identify key work-related stressors and questionnaires used in the construction industry’s safety domain to assess stress. Using a formal meta-analysis approach that leverages the findings from past studies, a more holistic determination of the relationship between work-related stressors and injury rates among workers was performed. Ninety-eight studies were reviewed, and seven were selected that fulfilled pre-determined validated inclusion criteria for eligibility in the meta-analysis. The results revealed 10 salient work-related stressors among construction workers. Of these ten, seven work-related stressors were identified as significant predictors of injury rates among workers: job control, job demand, skill demand, job certainty, social support, harassment and discrimination, and interpersonal conflicts at work. This study represents a significant first step toward formally identifying work-related stressors to improve working conditions, reduce or eliminate injuries on construction sites, and support future research.

Feasibility Study of Using Nebulizer-Retrofitted UAVs at Construction Projects: The Case Study of Residential Jobsites in Utah

Increased heat stress and air pollution are two major causes of health issues and productivity loss in the construction industry. This will be exacerbated by global warming and an increased possibility of extreme weather conditions in the future. Previous studies revealed that each 1°C rise in temperature reduces construction workers’ productivity by up to 2%. Also, it is found that a 10 μg/m3 particulate matter sized 2.5 μm (PM2.5) fluctuation over 25 days lowers workers’ daily production by 1%. As an effective dust control and cooling strategy, water spraying is used by practitioners to mitigate the effect of air pollution and heat-related problems on construction workers. Considering low costs and high mobility, unmanned aerial vehicles (UAVs) could be considered as a potential alternative for ground-based, stationary water spraying systems. To this end, a case study approach is adopted in this project to investigate the feasibility of using nebulizer-retrofitted UAVs in controlling air pollution and reducing heat stress at construction job sites. The case study consists of two different residential job sites in the state of Utah: one in Salt Lake City and the other one in St. George. The main contribution of this study is to determine the effect of aerial water spraying on air temperature and pollution at job sites. The outcomes of this study show that the average wet bulb globe temperature (WBGT) decreases 1.7° during the flight phase compared to average values for preflight and postflight phases. The aerial water spraying technique yields better and more water-efficient results in decreasing temperature at job sites compared to the existing approaches, such as stationary fans. Moreover, the results of PM variations illustrate that the mean value of particulate change was significantly different between flight and preflight (p=0.005), and flight and postflight (p<0.001) modes. Future studies should include the deployment of multiple drones flying simultaneously at job sites to cover larger areas.

Five factors affecting the on-body placement of wearable tactile safety promotion device for construction workers-on-foot

PurposeTactile communication that relies on the human sense of touch replicated using vibration motors is increasingly being explored for seamless communication on construction jobsite. However, the technological efficacy cannot secure the users’ acceptability of the tactile communication devices. This study aims to assess the factors affecting the wearability of such a portable tactile device based on the responses from practicing professionals.Design/methodology/approachThe investigation adapted a three-step phenomenological interviewing approach to seek feedback from construction personnel in Texas, the USA, regarding the viability of wearable tactile communication. The interviewees expressed various opinions about the on-body placement upon exposure to a portable tactile feedback prototype developed for this study, which was used to derive inferences regarding the factors affecting its on-field acceptability.FindingsAll the participants of the round-table study (11 out of 11) considered tactile feedback as a viable mode of communication on construction jobsite. Seven professionals supported the integration of a tactile device with the hard hat, whereas the rest preferred tactile eyeglasses. Weatherability, rechargeability, traceability, safety and social receptivity were identified as the major factors affecting the on-body placement of the wearable tactile communication device.Originality/valueThis paper presents a roadmap to gain construction industry opinion on the factors that can affect the on-body placement of a wearable tactile communication device. The five aforementioned factors impacting tactile communication acceptability were used to evaluate 10 potential on-body placements. The findings have implications for research and development of wearable tactile devices and the subsequent acceptability of such a device on the jobsite.

The development of a safety assessment model for using Unmanned aerial systems (UAS) in construction

In recent years, Unmanned Aerial Systems (UAS) have become very popular in the construction industry due to their versatility, ease of use, cost-effectiveness, and safety advantages. Several research studies investigated UAS use to assist with various tasks through different construction phases. However, the potential negative safety impacts to construction workers associated with UAS use in construction jobsites have not been studied adequately. Therefore, this study aims to fill this gap by developing a practical model to establish, assess, and improve the risk mitigation programs that construction companies have to control safety risks associated with UAS use in construction with a specific focus on workers’ occupational safety and health. The components of the model including the safety factors and mitigation methods were identified, verified, and quantified through a mixed approach relied on a review of the literature and a three-round Delphi process. The proposed model is expected to enable practitioners working in the construction industry to 1) understand and recognize the risks associated with the use of UAS in construction as well as their causal factors, 2) measure and evaluate the effectiveness of their own safety control programs for UAS-assisted projects, and 3) adjust and update their own safety control programs using effective mitigation strategies.

Fatigue Detection Post Physical Activity Using Machine Learning Algorithms

Construction work is purely effortful and the prevention of injuries at construction job sites is essential for encouraging worker’s well being and health which is generally overlooked at the construction sites. World’s construction industry is one amongst those having unsatisfactory work health issues. A large number of laborers and construction workers have to undergo fatigue risk at their job place. This paper describes the current state of the art of the research carried out in case of fatigue assessment after performing some physical activity providing an insight into fatigue, its detection and an overview of the causes of risk fatigue and its countermeasures. A number of subjective and objective fatigue assessment approaches have been used that have further stimulated the inclusion of latest and advanced approaches for fatigue detection. Although individual’s knowledge regarding the fatigue detection approaches has been enhanced, there is as yet minimal research in the field of fatigue detection post physical activity. In this study the stress would be on objective fatigue detection techniques where the acquisition of physiological data of construction workers is required. The study provides a ground for detection of physiological fatigue based on motion capture data in the form of videos.

Quantifying the Effect of Mental Stress on Physical Stress for Construction Tasks

Because of labor-intensive and physically demanding tasks, construction workers are known to be at a higher risk of developing physical fatigue. Recent studies have shown that construction workers are also often exposed to considerable mental stresses. While a number of studies have proposed methods and tools to measure and monitor physical and mental stress disjointedly, there is a need to explore their interaction. The literature indicates that no previous study has endeavored to evaluate the effect of mental stress on physical stress for construction tasks. This investigation is necessary to better comprehend work demands of construction tasks. Accordingly, entailing a randomized crossover design and simulated manual material handling experiment, this study evaluated the effect of cognitive task-led mental stress on physical stress using both subjective (ratings of perceived exertion) and objective measures (heart rate, skin temperature, and skin conductance). The results revealed that cognitive task-led mental stress led to a significant increase in subjective ratings of perceived exertion accompanied by an increase in skin temperature and skin conductance, while heart rate remained unaffected. The findings of this study add to the body of knowledge by highlighting that traditional benchmarking of task demands using only physiological measures may not be comprehensive. Rather, it might be suggested that additional psychological measurements are also essential because they might affect physical stress development. Furthermore, the current study has increased our understanding related to the interaction of physical and mental stress by revealing interindividual differences among the participants. Accordingly, by examining each worker separately, practitioners and safety managers can develop better mitigation strategies and individualized training programs, especially for more vulnerable workers, which can enhance overall health and safety on construction job sites.

Safety and health management response to COVID-19 in the construction industry: A perspective of fieldworkers

The COVID-19 outbreak has significantly impacted the construction industry. The pandemic can exacerbate an already dire safety and health situation in the industry and negatively impact construction employees and employers. The present study investigates the safety and health measures implemented by construction firms in the United States (US), their effectiveness and usefulness, and workers’ satisfaction with these COVID-19 measures. A questionnaire survey was developed and distributed to construction fieldworkers in the US to collect their perspectives on the implemented COVID-19 measures in the construction industry. A total of 187 valid responses were received and analyzed to achieve the aim of the study. Results revealed that strategies implemented to increase social distance and minimize group gathering to 10 persons in certain workstations were perceived to be substantially more effective than job-site screening strategies. Furthermore, smaller contractors implemented fewer safety measures and perceived them to be significantly less effective than those used by medium- and large-sized contractors. Fieldworkers were favorably disposed toward using technologies, such as video-conferencing apps and wearable sensing devices, to slow the spread of COVID-19 on construction job sites. The present study contributes to the body of knowledge by identifying safety and health measures to mitigate the spread of COVID-19 in construction. Practically, the study findings provide valuable insights to inform the successful implementation of safety strategies in the construction industry during a pandemic. The results are crucial for industry practitioners responsible for developing and revising pre- and post-pandemic safety and health plans.

Sound-based multiple-equipment activity recognition using convolutional neural networks

Automatically recognizing activities of heavy construction equipment using sound data has recently received considerable attention as a promising research area in construction. Although existing methods are effective, they only focus on tracking the activities of one single piece of equipment. On construction job sites, multiple equipment sound signals are mixed in the environment; Thus, there is a need for a robust method to recognize these activities that are taking place simultaneously. To address this shortcoming, we proposed a multi-label multi-level sound classification method based on Short-Time Fourier Transform (STFT) and Convolutional Neural Network (CNN) that only requires a single-channel off-the-shelf microphone. In addition, we developed a data augmentation method to simulate real-world equipment sound mixtures. We tested the proposed method on both synthetic and real-world equipment sound mixtures. The results of our study showed that this method was effective in identifying activities of multiple pieces of equipment on real construction job sites without the need for separating sound signals in advance. Future studies can focus on other potential applications of sound signal processing in the construction domain, including analyzing engine abnormalities and monitoring environmental performance of the equipment.

Analysis of Events and Exposures Leading to Construction Injuries in Developing Countries: The Case of Lebanon

There is unanimous agreement that the complex and dynamic nature of construction jobsites is coupled with inherent risks. Global occupational and health agencies revealed that the construction industry leads with the number of fatal injuries incurred every year. Lebanon, a developing country in the Middle East, is no exception. Studies on construction safety in the country showed that the industry has failed to implement safety regulations, while also witnessing a lack of safety incentives, training, and education. The situation worsens with the absence of a national experience multiplier that could reflect on contractors’ safety statuses and hold them accountable for their safety performance. This study addresses the problem by proposing a cost model that evaluates construction injuries. An analysis was performed on more than 3,000 accident claims collected from insurance companies to understand the events and exposures that construction labour face. Results showed that there are 17 different types of events and exposures which can be grouped into three clusters depending on their frequency and average cost. This paper presents and expands on the three clusters and identifies the events that fall under each of them. The findings will serve as the base of the cost model in a future study.

AI-based risk assessment for construction site disaster preparedness through deep learning-based digital twinning

Hurricanes are among the most devastating natural disasters in the United States, causing billions of dollars of property damage and insured losses. During extreme wind events, unsecured objects in jobsites can easily become airborne debris, which results in substantial loss to construction projects and neighboring communities. Towards a systematic disaster preparedness in construction jobsites, this paper presents a novel vision-based digital twinning and threat assessment framework. We encode the context of disaster risk into deep-learning architectures to identify and analyze the characteristics and impacts of potential wind-borne debris in construction site digital twin models. Case studies on nine piles of construction materials are presented to demonstrate and discuss the fidelity of the proposed computational modules. The proposed methods are expected to help provide heads up for practitioners to quickly recognize, localize, and assess potential wind-borne derbies in construction jobsites, and thereby implementing hurricane preparedness in an effective and timely manner.