The Key Methods To Monitor The Safety Status Of Construction Workers On Site

Construction safety management is one of the critical parts of construction management.Construction worker's unsafe behavior is the main cause for the safety accidents on construction on sites,thus it is necessary to analyze the rules and causes of construction worker's unsafe behaviors,so as to improve the construction safety performance.According to psychological theory and the previous research on unsafe construction safety,mental status is considered as an important influence factor of unsafe construction behavior.However,due to the lack of enough data,relevant research in construction industry has to be based on models and assumptions,limiting the feasibility of conclusion.To solve above problems,a real-time unsafe behavior and mental status data collecting method was proposed.To identify and record unsafe behavior,the method integrated ergonomics and computer vision,and developed an automatic and nonintrusive unsafe construction behavior identification and record method.The method first transferred unsafe behaviors to leading postures,and then simplified leading postures to key joint parameters according to human skeleton model.As a result,the behavior identification problem was transferred to a parameter matching problem,which was much easier to solve and thus facilitate real-time unsafe behavior identification and record.As for automatic mental data collecting,the method combined the latest progress in psychology,measuring mental status with physical indicators with wearable devices.14 kinds of emotion status and 4 kinds of physical indicators were selected.Through collecting and analyzing data from construction site,7 significant relationships were found between the above mental and physical indicators.These relationships were used to measure mental status with physical indicators.In conclusion,this research proposed an automatic unsafe behavior and mental data collecting method by integrating the knowledge of construction management,computer vision,ergonomics and psychology.In addition,a series of experiments were performed to demonstrate the feasibility and efficiency of the method.This research is of high academic and practical value.The method makes it possible to collect large amounts of behavioral and psychological data from construction site,which serve as the foundation for possible new construction safety management knowledge.