Using Network Analysis to Model Fall Hazards on Construction Projects

Safety risk quantification, integration, and management are emerging preconstruction strategies that have significant potential to impact construction safety performance. Unfortunately, current risk assessment strategies have limited practical application because every new infrastructure feature and construction method must be individually evaluated using laborious research processes and data from previous failures. In order to address this limitation, this research tested the hypothesis that the risk of worker injury in dynamic construction environments is the direct result of the temporal and spatial interactions among a limited number of identifiable and quantifiable task and object attributes. To test this hypothesis, a content analysis was conducted on 105 National Institute of Occupational Safety and Health Fatality Assessment and Control Evaluations reports. Attributes that contribute to incidents were identified and their relative risks were quantified according to frequency and severity of accidents that they have caused. Clusters and interactions among attributes were also modeled using Social Network Analysis (SNA) method. Ultimately, these research results can be used to improve the integration of safety information with building information models, project schedules, and architectural design. It is expected that the flexibility of the proposed approach will overcome most risk integration barriers that have been observed in the past decade.