A Multimodal Study To Assess Hazard Recognition Ability On Construction Site

Hazard recognition is a critical process for safety management and accident prevention in the complex and dynamic occupational environment.In order to improve the efficiency of job site safety management and carry out targeted management measures and related training projects,the current literature have tried a variety of quantitative methods to measure the hazard recognition ability of workers and individuals.Although previous studies have struggled to quantify hazard recognition ability,the results are potentially optimistic because of limitations in their experimental settings,unquantitative hazard classification method and/or single data collection channels of eye movements.These defects lead to some errors in the quantification results of the hazard recognition ability and indirectly lead to inadequate comprehension of the cognitive mechanism of the hazard recognition process of the construction workers.In order to overcome the weaknesses in the previous research,this study aims to re-develop a compound HRA index that utilizes the experimental technology of multimodal real-time monitoring and incorporates microvascular function in the brain,to make a more comprehensive measurement of the workers’ psychophysiological state during the construction hazard recognition.First,the authors identify critical indicators of hazard recognition and design an experiment to be conducted in a real scene.A total of forty-eight students majoring in civil engineering were invited to conduct the experiment.Data are then collected through questionnaires(experience and risk tolerance),eye-tracking devices(eye movement),and near-infrared spectroscopy.In the data analysis stage,the analysis of variance proved that the four independent variables in this paper--prefrontal cortex cognitive activation,eye movement features,experience,and risk tolerance,could well explain the variation of hazard recognition ability index of the dependent variable,and thus represent the correlation between the independent variable and the dependent variable.Finally,discriminant analysis is applied to develop an HRA index.The prediction accuracy of the proposed HRA index is shown to outperform previous approaches,indicating that the index development in this study is of advantages and the utilization of the near infrared reflectance spectrum is effective in construction safety research.Theoretically,this research signals a new perspective(changes in hemodynamic properties of the prefrontal cortex)in the assessment of hazard recognition ability.Quantifying the level of brain cognitive activity in the process of hazard recognition by measuring changes in the hemodynamic characteristics of the prefrontal cortex is proved to be effective and feasible.The experimental and analytical methodology in this study can serve as a reference for further studies.Practically,the proposed HRA index can be used for onboard assessment of workers or safety inspectors,reducing human errors and undetected occupational hazards.In addition,the cognitive process of hazard recognition for construction workers,the evaluation results and the sub-index performance of hazard recognition ability can be analyzed to deepen the understanding of cognitive mechanism and redesign safety training measurements from a cognitive perspective.