Prerpints.org logo
Preprint
Review

Construction Site Hazard Identification and Worker Adverse Reaction Monitoring Using Electroencephalograms: A Review

Altmetrics

Downloads

112

Views

66

Comments

0

A peer-reviewed article of this preprint also exists.

Submitted:

24 November 2023

Posted:

28 November 2023

You are already at the latest version

Alerts
Abstract
Construction safety is especially important because the construction industry is so important to a country's development. Significant research and practice have been conducted to mitigate potential risks during construction and improve worker efficiency. With the rapid advancement of cognitive neuroscience and the incorporation of medical technology in recent years, various wearable monitoring devices have been widely used in the construction field for real-time monitoring of workers' physical and mental status. Among these, the use of EEG (Electroencephalogram) in construction environment research allows researchers to gain insight into the physical and mental states of construction workers while performing construction tasks. This review introduces EEG technology and portable EEG devices, followed by an overview of their use in both monitoring workers' adverse reactions and identifying hazards on construction sites, providing an effective guide for EEG research in the construction field and on-site safety management.
Keywords: 
Subject: Engineering  -   Civil Engineering

1. Introduction

1.1. Research Background

The construction industry plays a critical role in the development of nations and societies. Over 350 million people work in frontline building construction around the world [1]. Construction accounts for approximately 7% of the total labor force in the United States, but construction workers account for approximately 20% of all industrial fatalities. With 15.2 deaths per 100,000 construction workers, construction is the third most dangerous industry after mining and agriculture. [2]. Construction is a dynamic process with complex and changing working conditions as well as the possibility of unforeseen circumstances at any time [3,4], necessitating a great deal of effort from frontline workers to ensure that safety incidents do not occur [5]. It is also distinguished by a heavy workload, long continuous working hours, unhealthy working postures, an unsuitable temperature and humidity of the working environment, and easy fatigue [6]. Furthermore, due to a lack of sleep or mood swings [7,8], construction workers may already be fatigued when they begin working [9]. Typically, 20-40% of construction workers exceed recognized physiological thresholds for physical labor [10]. As a result of being unable to concentrate for extended periods of time or frequently experiencing mental fatigue, workers are frequently unable to respond appropriately to potential safety hazards, which can lead to accidents [11]. It has been discovered that adverse reactions such as increased mental fatigue and cognitive decline in construction workers can result in hazards [12]. Recognizing and predicting the occurrence of construction hazards can be accomplished through methods such as the creation of scene graphs with interaction-level scene descriptions [13]. As a result, we can effectively reduce the occurrence of hazards during construction by monitoring both the subjective situation of construction workers and objective identification and prediction of hazards in the construction environment.
However, our research on the impact of environmental personnel is still dominated by subjective evaluation, which is the main technical tool used to study the impact of indoor environmental quality on occupants’ indoor comfort [14]. Typically, people assess indoor comfort by completing various evaluation questionnaires, such as thermal comfort [15], visual comfort [16], acoustic comfort [17,18,19], and perception of indoor air quality [20,21]. The same is true for outdoor building construction environments [22], where we obtain current physiological conditions of construction workers primarily through subjective questionnaires [12], and construction hazard identification primarily through personal inspections by the project manager [23,24], which are overly subjective [25] and closely related to the project manager’s personal work status. Subjective evaluation allows us to obtain a large amount of data for research in a short period of time more conveniently, but it also has drawbacks, such as limited topics that can be designed; a wide range of investigations but insufficient depth; variable quality of survey results; and susceptibility to the subjective thoughts of the subjects [9]. As a result, there is an urgent need for effective probes that can monitor the real-time status of building construction personnel in complex building construction environments. EEG, as a noninvasive and noninvasive neuroimaging technique, can provide accurate measurements of brain activity directly [27]. The brain can plan and execute autonomous movements, and many purposeful actions and behaviors are accomplished through various computational sequences within the brain [28], and EEG signals can also directly respond to nervous system activity [29]. Currently, physiological monitoring, such as EEG, is widely used in indoor environmental research [30,31]. Hu et al. [30] investigated the effects of various indoor lighting conditions on work efficiency by monitoring subjects’ EEG signals. EEG has a variety of applications in the field of sleep monitoring. PSG (polysomnography), for example, is used to determine sleep staging [32,33] and the presence of sleep-related disorders (e.g., OSA (Obstructive Sleep Apnea) [34,35], CSA (Central Sleep Apnea) [36], RBD (Rapid Eye Movement Sleep Behavioral Disorder) [37], and so on) in personnel [38].
With the extensive development of cognitive neuroscience technology in recent years, EEG technology can now be used not only for rational monitoring of indoor people, but also for outdoor environments, such as monitoring outdoor people’s movement [39,40], observing pedestrians’ avoidance behaviors in dangerous situations [41], and exploring outdoor thermal comfort and optimal outdoor environment [42]. People who work outside for long periods of time are exposed to more complex scenarios that consume more energy than those who work indoors, making them more susceptible to physical and psychological fatigue, as is the case in the construction industry. The physical and mental health of construction workers is a source of concern due to their long hours of outdoor work [6]. Many studies have been conducted in recent years on the application of EEG technology to the identification of hazardous behaviors and the monitoring of workers’ adverse reactions on construction sites, and the application of EEG in the field of construction is conducive to the in-depth understanding of the physical and mental state of construction workers during construction tasks, as well as the prediction and identification of hazards on construction sites, in order to elucidate In this paper, we first provide an overview of EEG technology, then summarize recent applications and research on the monitoring of workers’ adverse reactions and the identification of construction hazards in the field of construction, and finally look forward to the future development of this field.

1.2. EEG Technology

1.2.1. Four Functional Areas of the Brain

The human brain is the most complex structure known to man, with trillions of organized cells. The human brain is divided into two hemispheres, left and right, each of which controls the response body and receives information from it. Each hemisphere’s cerebral cortex is divided into four distinct lobes - frontal, parietal, temporal, and occipital - that are separated by deep sulcal fissures and have distinct functions. We’ve described the function and location of each lobe in Figure 1. Understanding these will aid us later in understanding of the EEG technique.

1.2.2. Five Brain Wave Frequencies

Hans Berger discovered EEG activity in 1929 and invented a technique for measuring EEG with the goal of providing “a window to the brain.” There are five main brain waves in the human brain, as shown in Table 1, ranging in frequency from low to high. These waves are closely related to states such as sleep, thought, cognition, arousal, and increased coordination when the brain is processing tasks. A typical EEG is made up of different frequency bands, and depending on the state of consciousness in which it is located, a specific brain wave will dominate, implying that different frequencies of brain waves correspond to different brain activities [43]. Different EEG frequency bands correspond to different subjective feelings and tasks, and the energy or work Power Spectral Density (PSD) of the EEG waves in each frequency band can indicate that different parts of the brain are activated, reflecting different physiological states [44].

1.2.3. The International 10-20 Electrode Placement System

The International 10-20 System electrode placement method, as shown in Figure 2, is a standard electrode placement method prescribed by the AASM (American Academy of Sleep Medicine) that is designed to maintain a standardized EEG testing methodology to ensure that the results of a subject’s study can be compiled, replicated, and validly analyzed and compared using the scientific method.
Electrode placement is primarily cranial in reference and does not differ based on individual differences in head circumference or head shape. The sagittal line is the anterior-posterior line from the root of the nose to the external occipital ridge, and the coronal line is the left-right line between the anterior recesses of the ears. The focal point of the two lines is at the top of the head, where the Cz electrode is located. The sagittal lines were Fpz, Fz, Cz, Pz, and Oz from anterior to posterior, and the spacing between the points was 20% of the sagittal line length except for the distances between Fpz and the root of the nose and Oz and the extra-occipital ramus, which were 10% of the sagittal line length; and along the coronal line, from 10% of the left anterior recess of the left ear, T3, C3 and Cz. The other points’ locations are shown above. Arabic numerals were used to represent the electrodes, with the left hemisphere being odd and the right hemisphere being even, A1 and A2 representing the right and left earlobes, respectively, and the numbers decreasing from the lateral to the midline.

1.2.4. Portable EEG Monitoring Devices

Portable monitoring devices, such as smartwatches, have grown in popularity in recent years and can be directly connected to a person’s cell phone, making it simple for the user to view various data and understand his or her current physiological state (e.g., heart rate, blood pressure, skin temperature, sleep quality, etc.), in order to better understand his or her health.
The following portable monitoring devices are commonly used, as illustrated in Figure 3: electroencephalogram, eye movement meter, accelerometer, skin temperature sensor, heart rate monitor, inertial measurement unit, and so on. People’s physiological signals can be monitored by various portable devices. Electroencephalograms (EEGs) can be used to monitor EEG signals, which are critical for judging construction unsafe behaviors and workers’ adverse reactions; construction workers can also wear portable eye-tracking devices to determine risks during the work process [46]; and skin temperature sensors can monitor the skin temperature of construction workers at the moment to obtain the thermal sensation situation at the moment. Millions of people in various industries, including construction workers, use personal portable devices on a daily basis to monitor their heart rate and other health-related physiological parameters to ensure their well-being [48]. Nnaji et al. [49] demonstrate, using data from the National Institute for Occupational Safety and Health (NIOSH) fatality data, that the likelihood of accidents can be greatly reduced by the prudent use of intelligent portable monitoring devices.
The use of intelligent portable monitoring devices has the potential to improve construction safety and efficiency. Because the healthcare industry has been at the forefront of implementing this type of technology, there are now an increasing number of cases of it being combined with the construction field [49]. Portable EEG monitoring devices that can monitor construction workers’ physiological states in real time without causing discomfort to their physiology are a novel idea in current research. Through application examples, this section introduces the Portable EEG monitoring device from two perspectives: Traditional Scalp EEG and Ear-EEG.

Traditional Scalp EEG

Wang et al. [50] created a Scalp EEG with a set of electrodes and a microprocessor installed in a standard helmet to collect EEG data from eight different parts of the wearer’s brain. The mental fatigue of construction workers was objectively monitored using electroencephalography (EEG) signals, and the EEG signals of 16 construction workers were recorded while performing their tasks, and the time-frequency-energy data of the acquired EEG signals were processed using WPT (Wavelet Packet Transform) and CNN (Convolutional Neural Networks) to recognize their current mental fatigue state. The framework provides a cognitive fatigue state classification that matches the self-reported fatigue state with an accuracy of 88.85%, which can be useful in reducing construction risky behaviors and providing assistance in fatigue management for workers.
Chen et al. [51] proposed and tested an EEG-based method for quantifying the mental load of construction workers. PSD (Power Spectral Density) was used to calculate the subjects’ post-experimental mental load. The results were consistent with the NASA-TLX (NASA Task Load Index) mental load score. A portable EEG helmet based on the Neurosky Think Gear module (NeuroSky, San Jose, California) was also developed to collect four sensing channels at different sensor locations, Fp1, Fp2, Tp9, and Tp10. The location of Fp1 was associated with logical attention; the location of Fp2 was associated with emotional attention. The two frontal EEG channels are compared to Tp9 and Tp10, which can be used as cross-channel references. In addition, an accelerometer was installed on the microcontroller to capture the three-axis motion of the helmet, as shown in Figure. In this experiment, each subject was asked to (1) sit in a chair and relax for 5 seconds; (2) climb a ladder (1 m high, requiring 3-4 s to reach the top); (3) select the appropriate bolt (2-3 s); (4) install the bolt (4-5 min); and (5) climb down the ladder and then rest. The installation task required each subject to select the appropriate nut and then tighten the bolt with a wrench. This task had to be completed three times by each subject. At the end of the experiment, all subjects were asked to complete a questionnaire to assess task load. Subjects wore helmets fitted with instruments and wirelessly connected to a laptop via Bluetooth during the experiment. Figure 4 depicts a schematic diagram of the experimental subjects and equipment. Figure 4a depicts the subjects with the experimental equipment, Figure 4b depicts the EEG monitoring chip, Figure 4c depicts the nuts and bolts used in the mounting activity, and Figure 4d depicts the ladder that the subjects had to climb.
Aryal et al. [52] created a sensing system that used infrared sensors attached to a helmet to monitor skin temperature at four different locations on the face, as well as heart rate and EEG signals. Physiological data from 12 construction workers were collected, and analysis revealed that the combination of skin temperature, heart rate, and EEG signals predicted worker fatigue with an accuracy of up to 82%.Li et al. [53] created a quantitative method for assessing the level of mental fatigue in subjects based on traditional scalp-based EEG measurements by examining and analyzing EEG spectra, such as gravity frequency. By collecting EEG signals from relevant brain regions of the subjects using traditional scalp-type EEG, Xing et al. [54] determined the positive effects of Progressive muscle relaxation and Trigeminal nerve stimulation sessions on the adverse emotions of construction workers at high altitude.

Ear-EEG

The traditional scalp EEG example was discussed above. Workers are bound to feel uncomfortable when wearing the helmet for an extended period of time due to the extremely limited space inside the helmet and the need to place the electrodes inside the helmet. Furthermore, workers will secrete a lot of sweat when working continuously outdoors, which will affect the electrode impedance and lead to inaccurate monitoring, making continuous EEG monitoring of workers inconvenient. As a result, we summarize another emerging and more popular EEG monitoring device, ear-EEG [55,56].
As shown in Figure 5, Looney et al. [57] developed the world’s first Ear-EEG in 2012, which drew widespread attention at the time. In general, the benefits of using Ear-EEG for monitoring include the fact that it does not obstruct the field of view and that the Ear-EEG is usually fixed in the ear canal for measurement, making it securely positioned and less likely to fall off. It is also less likely to sweat, avoiding the effects on electrode impedance caused by excessive sweating. While scalp EEG may require the assistance of experienced assistants, the Ear-EEG device can be simply placed in the subject’s ear, saving labor and improving the stability of continuous monitoring and monitoring efficiency.
Ear-EEG can be used to monitor facial expressions and body movements, and this research can aid in emotional recognition and determining the physical and mental states of construction workers. Matthies et al. [55] created an in-ear headset based on Neurosky EEG sensors that can control various cell phone functions using human blinking motions and ear wiggling. In a later study, Matthies et al. [58] used multiple electrodes on a foam earbud to detect 25 facial expressions and gestures using four different sensing techniques. The results showed that five gestures had an accuracy of more than 90% and 14 gestures had an accuracy of more than 50%. Athavipach et al. [56] designed and used an Ear-EEG to achieve basic emotion categorization and demonstrated a high level of accuracy. The studies above have shown that Ear-EEG has a good performance in monitoring and acquiring EEG signals, which, combined with its headphone-like convenience, confirms that it can be widely used in fields such as medical monitoring [59,60] and sleep monitoring [61].

1.3. EEG Monitoring of Worker Adverse Reactions and Construction Hazard Identification

According to the Occupational Safety and Health Administration, more than 6,000 fatalities or injuries of varying severity related to construction projects occur each year, accounting for approximately 20% of all fatalities in the United States. In just one year, 773 housing and municipal safety accidents occurred, resulting in 904 deaths [62]. Approximately 65% of these accidents are caused by falls from great heights, which are frequently caused by non-compliant interactions between workers and other building components. Aside from the enormous loss of life and the impact on injured people, there is also a massive economic loss. To avoid such losses, we must monitor construction workers’ adverse reactions and identify construction jobsite hazards so that they can be reduced or eliminated.
Current researchers have shown a link between brain waves of different frequencies and mental fatigue. Jap et al. [63] found that the energies of EEG signals (α + β)/θ, α/β, (θ + α)/(α + β), and θ/β can indicate driver fatigue. The (α + θ)/β energy of EEG signals, in particular, can indicate different degrees of mental fatigue [63,64,65]. Li et al. [66] used a two-channel EEG model with 92.3% accuracy to monitor driver fatigue.

1.4. Contributions of the Review

Based on the research background presented above, we know that in recent years, EEG research and application have played an important role in the aforementioned fields. With the comprehensive development of cognitive neuroscience technology, it has a greater potential for future development, providing broader ideas for research in various fields both indoors and outdoors, and promoting the advancement of times. This paper examines recent EEG applications in the monitoring of adverse reactions in workers and hazard identification in building construction, providing directions for future research and development.

2. Review Methodology

2.1. Literature Research

The study employed a content-based literature review methodology based on content analysis was used. The exact process of searching for and selecting literature is shown below. Google Scholar, Science Network, and Science Direct were among the databases searched during the review exercise. Journal articles, conference papers, and relevant books were among the target paper types. Finally, a collection of EEG-based literature or books on the monitoring of risky behaviors and workers’ adverse reactions in construction was gathered by examining the titles, abstracts, and keywords for human source identification.
To search the literature, Google Scholar, Web of Science, Science Direct, Pubmed, and Keywords were used. To ensure the relevance and high quality of the selected literature, keywords were used in conjunction with the Boolean operators “AND” and “OR” to conduct a comprehensive search of the literature on EEG in the identification of construction hazards, the use of EEG in adverse reactions in workers, and portable EEG. In addition, the search results were examined. In addition, references in the search results were scrutinized. Although this percentage was small, the review also included highly relevant literature where milestones appeared in the references, despite the fact that it was likely to be outside the search time and English language constraints.
Keyword co-occurrence analysis is a text analysis method for determining co-occurrence relationships between keywords in a given collection of text sets [67]. The basic idea behind keyword co-occurrence analysis is that if two keywords appear frequently together in a text, they may be related in some way. This can be accomplished by calculating the co-occurrence frequency of keywords, the co-occurrence matrix, the co-occurrence network, and other factors. The co-occurrence frequency indicates the number of times two keywords appear together in the same text, whereas the co-occurrence matrix and co-occurrence network can visualize the correlation between keywords. Researchers can use keyword co-occurrence analysis to uncover patterns and associations hidden in textual data, which can help them better understand textual content, discover related topics, create concept maps, and so on. This analysis method is useful for processing large amounts of text data, discovering domain knowledge, and assisting decision-making.
In this paper, we also use VOSviewer [68] to analyze the related keywords; the core idea of VOSviewer software is “co-occurrence clustering,” which means that two things appearing at the same time are related to each other; there are many types of such correlations, and their strengths and directions differ; different types of groups can be found based on the clustering of measures of the strengths and directions of the relationships. Clustering based on measures of the strength and direction of the relationship can be used to identify different types of groups. The co-occurrence visualization graph, shown in Figure 6, displays scientific terms that appear at least three times in all titles and abstracts. The size of the nodes indicates the frequency of a phrase in the entire literature, the thickness of the line indicates the strength of the association between the phrases, and the color of the same line indicates the proximity between the phrases.
Using these two approaches, this paper reviews a total of 22 articles, primarily focusing on the period between 2019 and 2023, with a small selection of important papers with representative studies prior to 2019.

2.2. Selection Criteria

The process was as follows, as illustrated in Figure 7, which describes the screening and adoption criteria of the literature for this review:
  • Consider combining EEG monitoring with subjective monitoring during the monitoring process.
  • In terms of monitoring adverse reactions in construction workers through EEG, consider mood monitoring, fatigue monitoring, distraction monitoring, and vigilance monitoring of workers.
  • Aspects of the identification of hazardous behavior in construction through EEG include monitoring at the construction site and simulation of the construction site environment in the laboratory through VR technology.

3. Worker’s Adverse Reaction Monitoring and Construction Hazard Identification

Traditional state identification is heavily reliant on subjective ratings derived from subjects’ responses to a series of body and mind-related questions [69,70]. As a result, respondents’ assessments of their actual state are frequently biased, and a large sample size is required to compensate for inter-subject variability and difference, preventing real-time state assessment and identification [71]. As a result, a real-time state monitoring method is required to accurately capture the subject’s true state in the moment. The recent boom in EEG technology has aided in the realization of such an approach, and EEG technology is increasingly being used to identify construction hazards and monitor construction workers’ adverse reactions on the job, providing new ideas for improving construction safety.

3.1. Workers Adverse Effects

EEG-acquired EEG signals have been widely studied in the construction field to analyze construction workers’ cognitive state and work status. This includes monitoring workers’ moods [54,56,73], fatigue [53,74,75,76], distraction and mental load [77,78,79,80], and vigilance in the face of construction hazards [81,82,83]. For your convenience, I will describe these aspects in the sections that follow.

3.1.1. Emotional Aspect of Workers

EEG has been shown to monitor and recognize different emotional states in people [72]. Workers’ emotional states are especially important during the construction process. The emotional state of construction workers is recognized as a significant factor influencing their performance. Measuring their emotional state should be prioritized to better understand the changes in workers’ moods while on the job in order to prevent negative effects on job performance. Among the many methods for measuring mood states, EEG has the greatest potential for quantitative measurement by overcoming the potential bias of subjective mood assessments based on surveys.
Because of the thin air and cold climate all year, keeping construction workers safe is especially important at high altitudes [54]. Positive interventions are especially important in preventing negative emotions from causing poor outcomes. Xing et al. [54] proposed a multi-component intervention that included PMR (Progressive Muscle Relaxation) and TNS (Trigeminal Nerve Stimulation). The subjects’ emotional and psychological states were measured using a portable EEG sensor, and the PMR intervention and TNS intervention were shown to be useful in adjusting the adverse emotions of high-altitude workers by different indicators. The VAD (valence-arousal-dominance) model was used to quantify the workers’ emotional characteristics. It was demonstrated that this intervention plays a key role in modulating the emotional state from negative to advanced based on the exponential trend of valence (significant increase), arousal (maintaining high level), and dominance (significant increase). Athavipach et al. [56] used Ear-EEG to classify basic emotions based on the modeling of valence and arousal emotions. Mir et al. [73] discovered that different types and levels of noise had a significant effect on the amplitude of the EEG signal, which in turn affected the emotional state. The negative emotional impact of noise generated by a saw lift hammer and a jackhammer was greatest in the frontal lobe, and the temporal lobe responded differently to noise types with different acoustic characteristics.

3.1.2. Work Fatigue Monitoring

In terms of monitoring worker fatigue, Aryal et al. [74] used portable EEG to simulate construction site conditions and concluded that classification accuracy based on features extracted from average skin temperature data was 9% higher than that based on heart rate data, and that combining the two resulted in an optimal accuracy of up to 82%, and also demonstrated that monitoring thermoregulation from the temples was more effective than other methods.
Li et al. [53] developed and experimentally validated a quantitative method to assess the level of mental fatigue of subjects by analyzing and comparing EEG spectral parameters such as gravity frequency and power spectral entropy for the pre-work fatigue of construction workers, and effectively screened the pre-work fatigue of workers by comparing experimental performance and workers’ reaction time. Xing et al. [75] designed a variety of manual processing tasks to stimulate the physical fatigue state, arranged the cognitively required risk identification task to induce mental fatigue, and tested this using the EEG integrated analysis method of rhythmic changes, pilot method for experiments, so that the results of this study provide a reference basis for the management of fatigue in construction workers. Tehrani et al. [76] investigated the effect of working at height on the development of mental fatigue in the context of preventing the risk of falling. Using wavelet transform and sample entropy two methods, the mental fatigue in the VR environment is evaluated, and the conclusion that the high intensity work group reflects a higher level of mental fatigue is reached, and the conclusion can also provide reference for the actual construction environment is reached.

3.1.3. Distraction and Psychological Burden of Workers

Ke et al. [77,78] investigated worker distraction in noise environments in terms of monitoring workers’ distraction and mental load. The cognitive performance and ability of workers in various noise environments were studied by having them wear portable EEGs in order to identify the hazards of different noise exposure conditions. The experimental results revealed that the degree of noise exposure had a negative correlation with subject performance, and intrinsic cognitive states such as attention, stress, and mental load were recovered with varying degrees of negative impact on the Big-O. Evaluating the effects of noise on cognitive functioning aids in explaining the psychological impact of poor mental performance, whereas neurocognitive monitoring of EEG lays the groundwork for predicting task performance under various noise conditions. Ke et al. [77] used sustained concentration on a response task and a dual-task paradigm to induce distraction and noise interference in workers, and discovered that, using feature calculations, the beta frequency of the left temporal lobe and right prefrontal cortex, as well as the frequencies, can distinguish these two states, and that these metrics can be used as an objective evaluation of an individual’s sustained focus and inability to sustain focus.
Chen et al. [79] assessed the construction process using mental load in order to develop a hazard assessment measure using neural time and frequency domain analysis. The experimental results demonstrated that EEG signals, particularly those in the low frequency band, were effective in assessing the mental load of construction workers. Then, for the problem of task allocation in construction, an EEG method for task mental load assessment based on the main frequency band PSD was introduced. Experiments were designed and carried out in order to validate the proposed assessment method, and the final statistical results and frequency box model revealed that the assessment results were consistent with the NASA-TLX scale scores and had a high reference value [80].

3.1.4. Vigilance Aspect of Workers

Building construction is an occupation where emergencies can happen at any time, and in the face of these emergencies, workers must remain alert at all times to avoid safety accidents, so it is especially important for workers to be alert at work. In an attempt to identify fatigue indicators, Thomas et al. [81] investigated various physiological correlates with fatigue. During monotonous driving, four types of EEG (δ, θ, α, β) were assessed in subjects, and the activity results showed that and activity remained stable over time, activity decreased slightly, and activity decreased significantly. Over time, all four algorithms demonstrated an increase in the ratio of slow-wave to fast-wave EEG activity. The findings are an important resource for monitoring workers’ vigilance at work.
Chen et al. [82] used some vigilance metrics to quantify construction workers’ ability to perceive risk, and proposed using wavelet decomposition to study and validate the proposed metrics, as well as field experiments to analyze EEG signal patterns and benchmarking with the results of other existing vigilance assessment methods. Wang et al. [83] proposed a hybrid motor-EEG data type using WPT to calculate vigilance measures, and the three indicators with the highest correlation coefficients were derived from validation experiments and can be used for vigilance detection. These quantitative vigilance indicators can help to improve the safety management capability of construction sites by providing a new perspective on understanding the risk perception process of workers.

3.2. Construction Hazard Identification

The significance of identifying building hazards cannot be overstated; it is the primary component that determines overall safety management [84,85]. In current practice, hazard identification is solely based on safety manager inspections; they manually inspect larger building spaces based on prior experience, knowledge, and established process systems (e.g., job hazard analysis) [23,24], which is clearly insufficient. Furthermore, such a solitary hazard identification process is overly subjective and prone to errors and mistakes. Different inspectors frequently perform different tasks and have different standards of judgment, resulting in subjective inconsistency [85,86]. To summarize, we introduced EEG technology to develop a new direction and provide new ideas for hazard identification in the construction field through objective monitoring of physiological information such as EEG in order to effectively reduce human judgment errors.
Because the aforementioned EEG signals have the ability to recognize various adverse reactions of workers during construction work, it is also used to monitor and improve worker safety at work, and is particularly useful in the field of hazard recognition in construction to make predictions and judgments about potential hazards during construction and effectively reduce the occurrence of hazardous accidents. [87,88,89,91,92,93,94,95].
To improve workers’ hazard recognition skills, Noghabaei et al. [87] conducted safety training in a Virtual Reality environment. By wearing VR, the subjects were able to almost recognize the hazards in the building construction environment. Workers in a subsequent study [88] wore both EEG and eye-tracking to navigate a simulated virtual construction site and identify safety hazards, and the 13 best features were chosen from 306 features acquired by EEG and eye-tracking to train a machine learning model. The findings show that EEG and eye tracking can be used to identify construction safety hazards.
Wang et al. [89] used real-world construction sites as stimuli. Sixty-one construction workers were chosen and shown 120 images, with their EEG signals recorded while they viewed the images. We extracted the ERP (Event-related Potential) evoked by both safe and dangerous images from the EEG signals [90] and trained the prediction model using a nested cross-validation algorithm. Correlations between EEG predictions and self-reported work experience were calculated to assess the prediction model’s performance. Following that, predicted values were used to predict participants’ behavioral outcomes and casualty experiences in order to assess the model’s external validity. Finally, the predictive model was tested on an independent data lockbox for out-of-sample validation. The model demonstrated comparable predictive accuracy for workers with varying risk propensities and educational backgrounds, eliminated the interference of specific risk types in model performance, and highlighted the feasibility of identifying individual characteristics for hazard recognition based on physiological signals.
Jeon et al. [91] discovered that the presence of hazards in the built environment could be identified by mood changes using portable EEG sensors in hazardous areas. In a subsequent study, Jeon et al. [92] combined a portable EEG device with immersive virtual reality (VR) to develop an EEG classifier to detect the presence of hazards in building construction. According to the study’s findings, the CatBoost classifier performed the best, with an accuracy of 95.1%. Furthermore, three critical channel locations (AF3, F3, and F4) and two frequency bands (Beta and Gamma) were found to be closely related to hazard perception. In 2022, Jeon’s team [93] correlated EEG signals with the types of hazards that are likely to occur in construction, as well as conducted experiments in the VR environment, to develop another EEG classifier that simulated construction in the VR environment with different hazardous situations, such as worker falls and circuit tripping. During the experiment, EEG signals from subjects wearing EEG and VR devices were recorded at the same time. For training and testing, two types of EEG features (time/frequency domain features and cognitive features) were extracted, and the EEG classifier was built using 18 advanced machine learning algorithms. Based on a 7-class categorized set of cognitive features, the LightGBM classifier achieves 70.1% accuracy. Jeon et al. relabeled the input data and designed and tested three strategies to improve performance even further. The two-step integrated classification method achieved 82.3% accuracy, according to the results. As a result of the preceding research, we can recognize and distinguish the hazards present in buildings by combining EEG, VR, and ML (Machine Learning) methods. In the same year, Jeon et al. [94] synthesized previous research and proposed an EEG-based framework for universal hazard identification and active safety management. The framework, as shown in Figure 8, is made up of three parts: (a) the creation of an immersive EEG hazard classifier; (b) multi-sensor, real-time hazard mapping; and (c) behavioral intervention. The framework’s feasibility is validated by focusing on the first component (immersive EEG hazard classifier), which is based on an indoor laboratory experiment. The overall framework is depicted in Figure. The binary classifier was found to be capable of classifying hazard-related EEG signals with an accuracy of 93.7%, while the multi-class classifier was capable of classifying EEG signals into five different hazard types with an accuracy of 79.3%. It greatly advances the process and development of construction safety management by providing a toolbox that can better identify construction hazards and helps to reduce the occurrence of hazardous events on construction sites.
Liao et al. [95] investigated the directional and time-varying information flow of observed brain activity in a recent study by collecting validated EEG data from 71 construction workers while performing 120-image-based hazard recognition and simulating the directional and time-varying information flow of observed brain activity using adaptive directional transfer functions based on the (Granger Causal Relation Test). During hazard recognition, brain connectivity is observed. The findings of this groundbreaking study show that behavioral targets from the dorsal attentional network play a top-down moderating role in hazard localization. The sensory cortex, which primarily serves as an information egress center, collaborates with the frontal lobes and visual cortex to provide an attentional redirection mechanism for top-down processing of potentially dangerous stimuli.
Efferent information flow was the strongest in the central parietal area (Cz, CP1, CP2, and Pz electrodes) in more than 70% of the subjects, while efferent information flow in the prefrontal lobe was weak. The areas around the Cz, CP1, CP2, and Pz electrodes in the parietal lobe are thought to be “sources” of brain activity related to hazard recognition. Table 2 depicts the information exchange in various brain regions at various times following stimulation.
From the standpoint of cognitive psychology, the study extends the modules of brain regions to effective brain connections. This is consistent with current research trends in psychophysiology, specifically exploring the relationship between brain neurological indicators and psychological factors while incorporating hazard recognition scenarios during building construction, and is an effective contribution to medical-industrial integration research.

4. Discussion and Limitations

EEG is a method of recording the electrical activity of the brain that can effectively capture how active the brain is at any given time [96]. In recent years, EEG technology has been used in a variety of fields, including medicine, psychology, education, recreation, and sleep health. EEG technology has also begun to be used in the construction field to monitor workers’ physical and mental states and to predict potential hazards, which is critical for improving safety in the construction field. Workers’ safety and physical and mental health are of utmost importance in a construction environment, and EEG technology can be used to assess their current emotional state, cognitive ability, and fatigue level by monitoring their EEG activity in real time. To prevent safety accidents, if a worker’s EEG signal shows signs of exertion or cognitive decline, timely measures can be taken to halt the work and allow the worker to take a break. EEG technology can also be used to anticipate potential hazards on construction sites. For example, by monitoring worker EEG activity and combining it with VR equipment [89,90], the working environment of a construction site can be simulated to identify potential safety hazards. It is also possible to predict whether a worker’s error or negligence will occur in the next few seconds or over a period of time, allowing for timely measures to be taken to prevent safety accidents. Furthermore, EEG classifiers can be developed in conjunction with machine learning algorithms to identify and classify hazards on construction sites [93,94].
Since the application of cognitive neuroscience to outdoor building construction has only recently emerged, the primary limitation of this review is the small number of referable articles and the small number of research directions categorized, with only two directions reviewed: monitoring of adverse reactions in workers and hazard identification during building site construction. Three studies focused on workers’ [54,56,73] emotional state, four studies focused on workers’ fatigue [53,74,75,76], four studies investigated workers’ distraction and mental load [77,78,79,80], and three studies focused on workers’ vigilance [81,82,83]. Eight studies focused on hazard identification on construction sites [87,88,89,91,92,93,94,95]. These studies provide a wide range of perspectives on the use of EEG in the field of building construction, but there are more directions to be explored in future research. The second issue is that there is no categorization of the causes of workers’ adverse reactions and building construction hazards, and future work can further analyze the causes of construction hazardous behaviors and workers’ adverse reactions, such as the effects of light environment, sound environment, extreme temperature, and extreme dryness or humidity during the construction process. The third flaw is that it does not address the numerous consequences of wearable devices, such as privacy concerns. The application of EEG technology is still in its early stages, and more research and development are needed to improve its accuracy and usefulness. Because of the limitations described above, relevant researchers should proceed with caution when referring to the review’s findings.

Conclusions and Future Perspectives

EEG is an emerging method for monitoring workers’ personal status during building construction, as well as a new way to improve the ability of safety management on construction sites, and has a broad development prospect that merits further research and exploration in the future. Existing studies included in this review have preliminary confirmed that the use of EEG technology can not only measure the emotional highs and lows, fatigue, distraction, mental load, and vigilance of construction workers, but can also be used to identify and judge construction hazards, greatly reducing the dangers in the construction process. Construction researchers can use current research to develop and refine EEG-based experiments that can help improve safety management in the construction industry. Neuroscientists, on the other hand, can use techniques such as filtering, artifact removal, and signal processing to obtain more high-quality EEG waveform data and improve the accuracy of EEG analysis.
Future research can improve the combination of EEG with other physiological signal monitoring (e.g., EOG (Electro-oculogram), EMG (Electromyogram), ECG (Electrocardiogram), and so on) to respond more accurately to the workers’ real-time status and understand their health conditions. Simultaneously, we investigate additional Machine Learning methods for combining physiological monitoring with Machine Learning to optimize the working environment and improve worker efficiency while preventing and reducing the occurrence of hazards. Meanwhile, as research and technological progress continue, EEG technology is expected to be applied in an increasing number of fields, bringing greater convenience to human life.

Author Contributions

Conceptualization, B.Y. and H.F.; methodology, B.W.; formal analysis, B.W., W.Z., P.L. and B.Y.; investigation, B.W.; resources, Z.L.; data curation, B.W.; writing—original draft preparation, B.W.; writing—review and editing, B.Y. and F.W.; visualization, Z.L.; supervision, B.Y., H.F. and F.W.; project administration, B.Y.; funding acquisition, B.Y. All authors have read and agreed to the published version of the manuscript.

Funding

The authors are grateful for the support of the National Natural Science Foundation of China (No. 52278119).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

The authors are grateful for the support of the National Natural Science Foundation of China (No. 52278119).

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

List of abbreviations
EEG  Electroencephalogram
MEG  Magnetoencephalography
EMG  Electromyogram
ECG  Electrocardiogram
EOG  Electro-oculogram
EDA  Electrodermal Activity
TCV  Thermal Comfort Vote
TSV  Thermal Sensation Vote
SVF  Sky View Factor
BCI  Brain-Machine Interface
ERP  Event-related Potential
SVM  Support Vector Machines
IVE  Immersive Virtual Environment
WPT  Wavelet Packet Transform
CSA  Center Sleep Apnea
MSA  Mix Sleep Apnea
PMR  Progressive Muscle Relaxation
TNS  Trigeminal Nerve Stimulation 		NALA-TLX  NASA Task Load Index
ML  Machine Learning
DL  Deep Learning
CNN  Convolutional Neural Networks
MMN  Mismatch Negativity
HMD  Head-mounted Device
RBD  REM Sleep Behavior Disorder
OSA  Obstructive Sleep Apnea
PSG  Polysomnography
TSV  Thermal Sensation Vote
MTSV  Mean TSV
TCV  Thermal Comfort Vote
PSD  Power Spectral Density
VAD  Valence-Arousal-Dominance
α  Alpha
β  Beta
θ  Theta
δ  Delta
γ  Gamma

References

  1. Birhane G E, Yang L, Geng J, et al. Causes of construction injuries: a review[J]. International journal of occupational safety and ergonomics, 2022, 28(1): 343-353. [CrossRef]
  2. Hinze J W, Teizer J. Visibility-related fatalities related to construction equipment[J]. Safety science, 2011, 49(5): 709-718. [CrossRef]
  3. Xing X, Zhong B, Luo H, et al. Effects of physical fatigue on the induction of mental fatigue of construction workers: A pilot study based on a neurophysiological approach[J]. Automation in Construction, 2020, 120: 103381. [CrossRef]
  4. Fang D, Wang Y, Lim H W, et al. Construction of a Bayesian network based on leadership-culture-behavior model to improve owner safety management behavior[J]. Journal of Construction Engineering and Management, 2023, 149(3): 04022177. [CrossRef]
  5. Fang D, Zhao C, Zhang M. A cognitive model of construction workers’ unsafe behaviors[J]. Journal of Construction Engineering and Management, 2016, 142(9): 04016039. [CrossRef]
  6. Aryal A, Ghahramani A, Becerik-Gerber B. Monitoring fatigue in construction workers using physiological measurements[J]. Automation in Construction, 2017, 82: 154-165. [CrossRef]
  7. Marcora S M, Staiano W, Manning V. Mental fatigue impairs physical performance in humans[J]. Journal of applied physiology, 2009. [CrossRef]
  8. Tempesta D, Socci V, De Gennaro L, et al. Sleep and emotional processing[J]. Sleep medicine reviews, 2018, 40: 183-195. [CrossRef]
  9. Li H, Wang D, Chen J, et al. Pre-service fatigue screening for construction workers through wearable EEG-based signal spectral analysis[J]. Automation in Construction, 2019, 106: 102851. [CrossRef]
  10. Abdelhamid T S, Everett J G. Physiological demands during construction work[J]. Journal of construction engineering and management, 2002, 128(5): 427-437. [CrossRef]
  11. Haslam R A, Hide S A, Gibb A G F, et al. Contributing factors in construction accidents[J]. Applied ergonomics, 2005, 36(4): 401-415. [CrossRef]
  12. Zhang M, Murphy L A, Fang D, et al. Influence of fatigue on construction workers’ physical and cognitive function[J]. Occupational Medicine, 2015, 65(3): 245-250. [CrossRef]
  13. Zhang L, Wang J, Wang Y, et al. Automatic construction site hazard identification integrating construction scene graphs with BERT based domain knowledge[J]. Automation in Construction, 2022, 142: 104535. [CrossRef]
  14. P. Ricciardi, C. Buratti, Environmental quality of university classrooms: Subjective and objective evaluation of the thermal, acoustic, and lighting comfort conditions, Build. Environ. 127 (2018) 23–36. [CrossRef]
  15. R. de Dear, Thermal comfort in practice, Indoor Air. 14 (2004) 32–39. [CrossRef]
  16. C.E. Ochoa, I.G. Capeluto, Evaluating visual comfort and performance of three natural lighting systems for deep office buildings in highly luminous climates, Build. Environ. 41 (2006) 1128–1135. [CrossRef]
  17. C. Buratti, E. Belloni, F. Merli, P. Ricciardi, A new index combining thermal, acoustic, and visual comfort of moderate environments in temperate climates, Build. Environ. 139 (2018) 27–37. [CrossRef]
  18. N. Perrin Jegen, P. Chevret, Effect of noise on comfort in open-plan offices: application of an assessment questionnaire, Ergonomics. 60 (2017) 6–17. [CrossRef]
  19. M. Pierrette, E. Parizet, P. Chevret, J. Chatillon, Noise effect on comfort in open-space offices: development of an assessment questionnaire, Ergonomics. 58 (2015) 96–106. [CrossRef]
  20. P. Wargocki, D.P. Wyon, Y.K. Baik, G. Clausen, P.O. Fanger, Perceived air quality, sick building syndrome (SBS) symptoms and productivity in an office with two different pollution loads, Indoor Air-Int. J. Indoor Air Qual. Clim. 9 (1999) 165–179. [CrossRef]
  21. L. Lan, P. Wargocki, D.P. Wyon, Z. Lian, Effects of thermal discomfort in an office on perceived air quality, SBS symptoms, physiological responses, and human performance, Indoor Air-Int. J. Indoor Air Qual. Clim. 21 (2011) 376–390. [CrossRef]
  22. Carvajal-Arango D, Vasquez-Hernandez A, Botero-Botero L F. Assessment of subjective workplace well-being of construction workers: A bottom-up approach[J]. Journal of Building Engineering, 2021, 36: 102154. [CrossRef]
  23. Rozenfeld, R. Sacks, Y. Rosenfeld, H. Baum, Construction job safety analysis, Saf. Sci. 48 (2010) 491–498. [CrossRef]
  24. H.-H. Wang, F. Boukamp, Ontology-based representation and reasoning framework for supporting job hazard analysis, J. Comput. Civ. Eng. 25 (2011) 442–456. [CrossRef]
  25. Pinto A, Nunes I L, Ribeiro R A. Occupational risk assessment in construction industry–Overview and reflection[J]. Safety science, 2011, 49(5): 616-624. [CrossRef]
  26. Pan.Y,Wan.Y,Chen.G,Research on Neural Information Systems: Current Status and Prospects[J]. JOURNAL OF MANAGEMENT SCIENCES IN CHINA,2018,21(05):1-21.
  27. Szafir, D.; Signorile, R. An exploration of the utilization of electroencephalography and neural nets to control robots. In Proceedings of the IFIP Conference on Human-Computer Interaction, Lisbon, Portugal, 5–9 September 2011; pp. 186–194. [Google Scholar] [CrossRef]
  28. Inagaki H K, Chen S, Ridder M C, et al. A midbrain-thalamus-cortex circuit reorganizes cortical dynamics to initiate planned movement[J]. bioRxiv, 2020: 2020.12. 16.423127. [CrossRef]
  29. Wu M, Li H, Qi H. Using electroencephalogram to continuously discriminate feelings of personal thermal comfort between uncomfortably hot and comfortable environments[J]. Indoor Air, 2020, 30(3): 534-543. [CrossRef]
  30. Lu M, Hu S, Mao Z, et al. Research on work efficiency and light comfort based on EEG evaluation method[J]. Building and Environment, 2020, 183: 107122. [CrossRef]
  31. Lang X, Wang Z, Tian X, et al. The effects of extreme high indoor temperature on EEG during a low intensity activity[J]. Building and Environment, 2022, 219:109225. [CrossRef]
  32. Zhang L, Fabbri D, Upender R, et al. Automated sleep stage scoring of the Sleep Heart Health Study using deep neural networks[J]. Sleep, 2019, 42(11): zsz159. [CrossRef]
  33. Boostani R, Karimzadeh F, Nami M. A comparative review on sleep stage classification methods in patients and healthy individuals[J]. Computer methods and programs in biomedicine, 2017, 140: 77-91. [CrossRef]
  34. Ishman S L, Yang C J, Cohen A P, et al. Is the OSA-18 predictive of obstructive sleep apnea: comparison to polysomnography[J]. The Laryngoscope, 2015, 125(6): 1491-1495. [CrossRef]
  35. Colín-Cano C, Ruiz-Morales M, Ledo-Gutiérrez L, et al. Correlation between the clinical inventory of sleep quality (CISQ) and polisomnographic (PSG) characteristics in patients with obstructive sleep apnea syndrome (OSA)[J]. Sleep Medicine, 2017, 40: e150. [CrossRef]
  36. Kritzinger F E, Al-Saleh S, Narang I. Descriptive analysis of central sleep apnea in childhood at a single center[J]. Pediatric pulmonology, 2011, 46(10): 1023-1030. [CrossRef]
  37. Neikrug A B, Ancoli-Israel S. Diagnostic tools for REM sleep behavior disorder[J]. Sleep medicine reviews, 2012, 16(5): 415-429. [CrossRef]
  38. Tabar Y R, Mikkelsen K B, Rank M L, et al. Ear-EEG for sleep assessment: a comparison with actigraphy and PSG[J]. Sleep and Breathing, 2021, 25: 1693-1705. [CrossRef]
  39. Scanlon J E M, Jacobsen N S J, Maack M C, et al. Outdoor walking: Mobile EEG dataset from walking during oddball task and walking synchronization task[J]. Data in Brief, 2023, 46: 108847. [CrossRef]
  40. Jacobsen N S J, Blum S, Witt K, et al. A walk in the park? Characterizing gait-related artifacts in mobile EEG recordings[J]. European Journal of Neuroscience, 2021, 54(12): 8421-8440. [CrossRef]
  41. Tang S, Wang J, Liu W, et al. A study of the cognitive process of pedestrian avoidance behavior based on synchronous EEG and eye movement detection[J]. Heliyon, 2023, 9(3). [CrossRef]
  42. Ma X, Song L, Hong B, et al. Relationships between EEG and thermal comfort of elderly adults in outdoor open spaces[J]. Building and Environment, 2023, 235: 110212. [CrossRef]
  43. J.N. Demos, Getting Started with Neurofeedback, WW Norton & Company, New York, 2005. ISBN: 0-393-07553-2.
  44. M.X. Cohen, Analyzing Neural Time Series Data: Theory and Practice, MIT press, Cambrige, 2014. ISBN: 0-262-31956-X.
  45. Sharbrough, F.; Chatrian, G.E.; Lesser, R.P.; Luders, H.; Nuwer, M.; Picton, T.W. American electroencephalographic society guidelines for standard electrode position nomenclature. J. Clin. Neurophysiol. 1991, 8, 200–202. [Google Scholar]
  46. Xu Q, Chong H Y, Liao P. Exploring eye-tracking searching strategies for construction hazard recognition in a laboratory scene[J]. Safety Science, 2019, 120: 824-832. [CrossRef]
  47. Patel S, Park H, Bonato P, et al. A review of wearable sensors and systems with application in rehabilitation[J]. Journal of neuroengineering and rehabilitation, 2012, 9(1): 1-17. [CrossRef]
  48. Nnaji C, Okpala I, Awolusi I. Wearable sensing devices: Potential impact & current use for incident prevention[J]. Professional safety, 2020, 65(04): 16-24. [CrossRef]
  49. Abuwarda Z, Mostafa K, Oetomo A, et al. Wearable devices: Cross benefits from healthcare to construction[J]. Automation in Construction, 2022, 142: 104501. [CrossRef]
  50. Wang Y, Huang Y, Gu B, et al. Identifying mental fatigue of construction workers using EEG and deep learning[J]. Automation in Construction, 2023, 151: 104887. [CrossRef]
  51. Chen J, Taylor J E, Comu S. Assessing task mental workload in construction projects: A novel electroencephalography approach[J]. Journal of Construction Engineering and Management, 2017, 143(8): 04017053. [CrossRef]
  52. Aryal A, Ghahramani A, Becerik-Gerber B. Monitoring fatigue in construction workers using physiological measurements[J]. Automation in Construction, 2017, 82: 154-165. [CrossRef]
  53. Li H, Wang D, Chen J, et al. Pre-service fatigue screening for construction workers through wearable EEG-based signal spectral analysis[J]. Automation in Construction, 2019, 106: 102851. [CrossRef]
  54. Xing X, Li H, Li J, et al. A multicomponent and neurophysiological intervention for the emotional and mental states of high-altitude construction workers[J]. Automation in Construction, 2019, 105: 102836. [CrossRef]
  55. Matthies, D.J.C. InEar BioFeedController: A headset for hands-free and eyes-free interaction with mobile devices. In CHI’13 Extended Abstracts on Human Factors in Computing Systems; ACM: New York, NY, USA, 2013; pp. 1293–1298. [Google Scholar] [CrossRef]
  56. Athavipach C, Pan-Ngum S, Israsena P. A wearable in-ear EEG device for emotion monitoring[J]. Sensors, 2019, 19(18): 4014. [CrossRef]
  57. Looney D, Kidmose P, Park C, et al. The in-the-ear recording concept: User-centered and wearable brain monitoring[J]. IEEE pulse, 2012, 3(6): 32-42. [CrossRef]
  58. Matthies D J C, Strecker B A, Urban B. Earfieldsensing: A novel in-ear electric field sensing to enrich wearable gesture input through facial expressions[C]//Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems. 2017: 1911-1922. [CrossRef]
  59. Nguyen A, Alqurashi R, Raghebi Z, et al. A lightweight and inexpensive in-ear sensing system for automatic whole-night sleep stage monitoring[C]//Proceedings of the 14th ACM Conference on Embedded Network Sensor Systems CD-ROM. 2016: 230-244. [CrossRef]
  60. You S, Cho B H, Yook S, et al. Unsupervised automatic seizure detection for focal-onset seizures recorded with behind-the-ear EEG using an anomaly-detecting generative adversarial network[J]. Computer Methods and Programs in Biomedicine, 2020, 193: 105472. [CrossRef]
  61. You S, Cho B H, Shon Y M, et al. Semi-supervised automatic seizure detection using personalized anomaly detecting variational autoencoder with behind-the-ear EEG[J]. Computer Methods and Programs in Biomedicine, 2022, 213: 106542. [CrossRef]
  62. Ministry of Housing and Urban Rural Development of China. https://www.mohurd.gov.cn/gongkai/fdzdgknr/tzgg/202006/20200624_246031.html, June 2020 (last access August 11, 2022).
  63. Jap B T, Lal S, Fischer P, et al. Using EEG spectral components to assess algorithms for detecting fatigue[J]. Expert Systems with Applications, 2009, 36(2): 2352-2359. [CrossRef]
  64. Barwick F, Arnett P, Slobounov S. EEG correlates of fatigue during administration of a neuropsychological test battery[J]. Clinical Neurophysiology, 2012, 123(2): 278-284. [CrossRef]
  65. Simon M, Schmidt E A, Kincses W E, et al. EEG alpha spindle measures as indicators of driver fatigue under real traffic conditions[J]. Clinical Neurophysiology, 2011, 122(6): 1168-1178. [CrossRef]
  66. Li W, He Q, Fan X, et al. Evaluation of driver fatigue on two channels of EEG data[J]. Neuroscience letters, 2012, 506(2): 235-239. [CrossRef]
  67. Freilich S, Kreimer A, Meilijson I, et al. The large-scale organization of the bacterial network of ecological co-occurrence interactions[J]. Nucleic acids research, 2010, 38(12): 3857-3868. [CrossRef]
  68. Van Eck N, Waltman L. Software survey: VOSviewer, a computer program for bibliometric mapping[J]. scientometrics, 2010, 84(2): 523-538. [CrossRef]
  69. Xing X, Zhong B, Luo H, et al. Effects of physical fatigue on the induction of mental fatigue of construction workers: A pilot study based on a neurophysiological approach[J]. Automation in Construction, 2020, 120: 103381. [CrossRef]
  70. STOP, THAT and one hundred other sleep scales[M]. Springer Science & Business Media, 2012.
  71. Aryal A, Ghahramani A, Becerik-Gerber B. Monitoring fatigue in construction workers using physiological measurements[J]. Automation in Construction, 2017, 82: 154-165. [CrossRef]
  72. Dura A, Wosiak A. Unsupervised emotional state recognition based on clustering of EEG features[J]. Procedia Computer Science, 2022, 207: 3093-3102. [CrossRef]
  73. Mir M, Nasirzadeh F, Bereznicki H, et al. Investigating the effects of different levels and types of construction noise on emotions using EEG data[J]. Building and Environment, 2022, 225: 109619. [CrossRef]
  74. Aryal A, Ghahramani A, Becerik-Gerber B. Monitoring fatigue in construction workers using physiological measurements[J]. Automation in Construction, 2017, 82: 154-165. [CrossRef]
  75. Xing X, Zhong B, Luo H, et al. Effects of physical fatigue on the induction of mental fatigue of construction workers: A pilot study based on a neurophysiological approach[J]. Automation in Construction, 2020, 120: 103381. [CrossRef]
  76. Tehrani B M, Wang J, Truax D. Assessment of mental fatigue using electroencephalography (EEG) and virtual reality (VR) for construction fall hazard prevention[J]. Engineering, construction and architectural management, 2022, 29(9): 3593-3616. [CrossRef]
  77. Ke J, Zhang M, Luo X, et al. Monitoring distraction of construction workers caused by noise using a wearable Electroencephalography (EEG) device[J]. Automation in Construction, 2021, 125: 103598. [CrossRef]
  78. Ke J, Du J, Luo X. The effect of noise content and level on cognitive performance measured by electroencephalography (EEG)[J]. Automation in Construction, 2021, 130: 103836. [CrossRef]
  79. Chen J, Song X, Lin Z. Revealing the “Invisible Gorilla” in construction: Estimating construction safety through mental workload assessment[J]. Automation in Construction, 2016, 63: 173-183. [CrossRef]
  80. Chen J, Taylor J E, Comu S. Assessing task mental workload in construction projects: A novel electroencephalography approach[J]. Journal of Construction Engineering and Management, 2017, 143(8): 04017053. [CrossRef]
  81. Jap B T, Lal S, Fischer P, et al. Using EEG spectral components to assess algorithms for detecting fatigue[J]. Expert Systems with Applications, 2009, 36(2): 2352-2359. [CrossRef]
  82. Chen J, Lin Z, Guo X. Developing construction workers’ mental vigilance indicators through wavelet packet decomposition on EEG signals[C]//Construction research congress 2018. 2018: 51-61.
  83. Wang D, Li H, Chen J. Detecting and measuring construction workers’ vigilance through hybrid kinematic-EEG signals[J]. Automation in Construction, 2019, 1ym uj 00: 11-23. [CrossRef]
  84. A. Perlman, R. Sacks, R. Barak, Hazard recognition and risk perception in construction, Saf. Sci. 64 (2014) 22–31. [CrossRef]
  85. G. Carter, S.D. Smith, Safety hazard identification on construction projects, J. Constr. Eng. Manag. 132 (2006) 197–205. [CrossRef]
  86. S. Hasanzadeh, B. Esmaeili, M.D. Dodd, Impact of construction workers’ hazard identification skills on their visual attention, J. Constr. Eng. Manag. 143 (2017) 04017070. [CrossRef]
  87. Wang M, Zhao Y, Liao P C. EEG-based work experience prediction using hazard recognition[J]. Automation in Construction, 2022, 136: 104151. [CrossRef]
  88. M. Noghabaei, K. Han, Hazard Recognition in an immersive virtual environment: framework for the simultaneous analysis of visual search and EEG patterns, in: Constr. Res. Congr. 2020, American Society of Civil Engineers, Reston, VA, 2020: pp. 934–943. [CrossRef]
  89. M. Noghabaei, K. Han, A. Albert, Feasibility study to identify brain activity and eye-tracking features for assessing hazard recognition using consumer-grade wearables in an immersive virtual environment, J. Constr. Eng. Manag. 147 (2021) 04021104. [CrossRef]
  90. Liu Q. Influence mechanism of construction workers’ safety psychology on their safety behavior based on event-related potentials[J]. NeuroQuantology, 2018, 16(6). [CrossRef]
  91. J. Jeon, H. Cai, D. Yu, X. Xu, Identification of Safety Hazards Using Wearable EEG, in: Constr. Res. Congr. 2020, American Society of Civil Engineers, Reston, VA, 2020: pp. 185–194. [CrossRef]
  92. J. Jeon, H. Cai, Classification of construction hazard-related perceptions using: wearable electroencephalogram and virtual reality, Autom. Constr. 132 (2021), 103975. [CrossRef]
  93. Jeon J H, Cai H. Multi-class classification of construction hazards via cognitive states assessment using wearable EEG[J]. Advanced Engineering Informatics, 2022, 53: 101646. [CrossRef]
  94. J. Jeon, H. Cai, A Framework for EEG-based ubiquitous hazard identification and proactive safety management, in: Constr. Res. Congr. 2022, American Society of Civil Engineers, Reston, VA, 2022: pp. 145–153. [CrossRef]
  95. Liao P C, Zhou X, Chong H Y, et al. Exploring construction workers’ brain connectivity during hazard recognition: a cognitive psychology perspective[J]. International Journal of Occupational Safety and Ergonomics, 2023, 29(1): 207-215. [CrossRef]
  96. Lacan L, Hamoud Y, Nguyen S, et al. Fetal sheep cerebral electrical activity: A new technique to record EEG[J]. Journal of Neuroscience Methods, 2020, 345: 108888. [CrossRef]
Preprints 91412 g001
Figure 1. The main functions of the four functional areas of the brain.
Figure 1. The main functions of the four functional areas of the brain.
Preprints 91412 g001
Preprints 91412 g002
Figure 2. Labels for points according to the international 10-20 electrode placement system.
Figure 2. Labels for points according to the international 10-20 electrode placement system.
Preprints 91412 g002
Preprints 91412 g003
Figure 3. Potential locations and use of portable sensors.
Figure 3. Potential locations and use of portable sensors.
Preprints 91412 g003
Preprints 91412 g004
Figure 4. Experimental apparatus: (a) experiment subject; (b) EEG hardhat and the EEG monitoring chips; (c) nuts and bolts; (d) ladder platform [51].
Figure 4. Experimental apparatus: (a) experiment subject; (b) EEG hardhat and the EEG monitoring chips; (c) nuts and bolts; (d) ladder platform [51].
Preprints 91412 g004
Preprints 91412 g005
Figure 5. The first in-ear EEG prototype [57].
Figure 5. The first in-ear EEG prototype [57].
Preprints 91412 g005
Preprints 91412 g006
Figure 6. Electroencephalography in construction for hazard identification and monitoring of workers’ adverse reactions.
Figure 6. Electroencephalography in construction for hazard identification and monitoring of workers’ adverse reactions.
Preprints 91412 g006
Preprints 91412 g007
Figure 7. Selection criteria.
Figure 7. Selection criteria.
Preprints 91412 g007
Preprints 91412 g008
Figure 8. Overall research framework [94].
Figure 8. Overall research framework [94].
Preprints 91412 g008
Table 1. The subjective feelings and tasks corresponding to the five brain waves.
Table 1. The subjective feelings and tasks corresponding to the five brain waves.
Band name Frequency band (Hz) Subjective feeling state Relevant mandates and behaviours
Delta (δ) 0.5Hz-4Hz dreamless sleep, non-REM sleep, asleep Drowsiness, immobility, difficulty concentrating
Theta (θ) 4Hz-8Hz Intuition, recollection, deeply relaxed Be creative and intuitive; Distraction, lack of concentration
Alpha (α) 8Hz-13Hz relaxed, not irritable, not sleepy Meditative, no movement
Beta (β) 13Hz-30Hz Alert, excited, focused Conduct mental activities
Gamma (γ) 30Hz-Up High performance Advanced information processing and information-rich task processing
Table 2. Information exchange in different regions at different times after stimulation.
Table 2. Information exchange in different regions at different times after stimulation.
The time after being stimulated Information exchange in different regions
200 ms The parietal lobe has a relatively active exchange of information with the whole brain.
240-300 ms There is relatively active information outflow from the lateral parietal region.
200-500 ms Strong information outflow was observed in the left temporal lobe region.
400-600 ms Strong information outflow was observed in the right parietal lobe region.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

facebook logotwitter logolinkedin logo
MDPI Initiatives
Important Links
Subscribe

Disclaimer

Terms of Use

Privacy Policy

© 2024 MDPI (Basel, Switzerland) unless otherwise stated