Abstract:Developing real-time automatic intuitive devices based on electroencephalography (EEG) to improve hazard recognition performance (HRP) is essential in construction engineering management. However, existing studies generally developed devices based on unimodal data, failing to distinguish the role played by brain activities in different visual areas and ignoring the possible existence of other psychological mechanisms. Therefore, this study aims to use multimodal data based on brain-eye synergy to predict HRP and reveal the cognitive mechanism of hazard recognition to develop intuitive hazard recognition devices. Accordingly, we designed experiments to investigate this problem: (1) 53 construction workers completed a 120-trial hazard recognition task in a laboratory environment (N=6,360), (2) an EEG recorder collected the participants' EEG activities and an eye-tracking device recorded the data of eye movement, and (3) the power spectral density (PSD) of EEG was calculated and then imported into Matlab2018a in combination with EEGLAB for analysis. The results showed that the prediction results of hazard recognition have a time-variant effect. For fall-related hazards, EEG signals within 0-320 ms after the onset were significant in predicting HRP; for electric-related and fire-related hazards, EEG signals within 560-640 ms after the starting point were the most valuable. Also, compared with fall-related hazards, it took a significantly longer time to recognize electric-related and fire-related hazards, which consumed more attentional resources. Finally, the frontal lobe requires more attentional resources to finish cognitive activities than the occipital lobe to finish visual information processing activities, and the occipital lobe needs to process visual information first, and then transmit it to the frontal lobe for analysis and judgment. For the body of knowledge, this study, finding the time-variant effect of HRP, improves the theory of construction hazard recognition from an eye-brain synergy perspective and provides guidance for developing intuitive hazard recognition devices and improving management performance. This study combined eye-movement data and EEG data to predict the performance of hazard recognition using multimodal data. Not only did this study reveal the mechanism of hazard recognition, but it also proposed suggestions for project management and device development. Managers are suggested to implement refined management and regularly monitor the physiological signals of employees. Some construction companies started to use smart devices to measure workers' arousal, emotional valence, emotional regulation ability, fatigue, workload, or safety attention level before work every day. And it is recommended to provide safety reminders to employees based on whether there are any abnormalities in their physiological indicators. In addition to project management, the results of this study can serve the development or optimization of hazard recognition equipment. Developing devices cannot only consider EEG information, but also combine EEG information with eye-movement information. Meanwhile, developers can also consider different types of hazards and different brain regions to improve device performance.

Mixed reality-based active Hazard prevention system for heavy machinery operators

Real-time mixed reality-based visual warning for construction workforce safety

Efficiency of VR-Based Safety Training for Construction Equipment: Hazard Recognition in Heavy Machinery Operations

Motion Recognition Method for Construction Workers Using Selective Depth Inspection and Optimal Inertial Measurement Unit Sensors

Real-time smart video surveillance to manage safety: A case study of a transport mega-project

Varification of a Detectability of Dangerous Scenes Felt by Civil Workers and Construction Site Workers with Biological Measurement Using VR-Video

Real-time simulation of construction workers using combined human body and hand tracking for robotic construction worker system

Computer Vision-Based Human-Machine Collision Warning System on Construction Site

Combined Method Comprising Low Burden Physiological Measurements with Dry Electrodes and Machine Learning for Classification of Visually Induced Motion Sickness in Remote-Controlled Excavator

Presence, Mixed Reality, and Risk-Taking Behavior: A Study in Safety Interventions

AR-Facilitated Safety Inspection and Fall Hazard Detection on Construction Sites

Investigating hazard recognition in augmented virtuality for personalized feedback in construction safety education and training

Revealing the "invisible Gorilla" in Construction: Estimating Construction Safety Through Mental Workload Assessment

EEG-Based Performance-Driven Adaptive Automated Hazard Alerting System in Security Surveillance Support

Improving Construction Equipment Operation Safety from a Human-centered Perspective

Monocular 2D Camera-based Proximity Monitoring for Human-Machine Collision Warning on Construction Sites

Toward an Intuitive Device for Construction Hazard Recognition Management: Eye Fixation–Related Potentials in Reinvestigation of Hazard Recognition Performance Prediction

Proactive analysis of construction equipment operators’ hazard perception error based on cognitive modeling and a dynamic Bayesian network

An AR-assisted Deep Reinforcement Learning-based approach towards mutual-cognitive safe human-robot interaction

Sharing Construction Safety Inspection Experiences and Site-Specific Knowledge through XR-Augmented Visual Assistance