No abstract

The construction industry has become one of the most hazardous industries worldwide. How to prevent construction accidents on site is attracting increasing and extensive attention. Traditional accident prevention methods depend primarily on safety training and site management. These are demonstrably not enough to support safety management, evidenced by the high rates of construction accidents. As worker behavior is a key factor influencing accidents, this paper proposes a warning system of unsafe behavior by integrating Building Information Modeling (BIM) with Radio Frequency Identification (RFID) positioning technology. Through the investigation and classification of unsafe behaviors on site, the information demands of the warning system are analyzed and a prototype of the system is developed. Its functions are also tested in test bed trials, where it identified workers' locations on site in real time and responded to workers in unsafe situations, indicating that the system has a clear potential to improve safety management on site.