Analysis of Productivity Variation by Accident Risk in Construction Industry

The construction industry is characterized by continual changes, bombardment of varying technologies, poor working conditions, the involvement and the need for coordination of different trades and operations. Due to the hazardous nature of work, the industry has had a disproportionately high rate of accidents for its size. Accident statistics have played an important role as a prime indicator for measuring safety performance as well as a framework for evaluating accident prevention programs. However, the current system of statistics collection is based upon post-accident analyses. These data provide factual information regarding the post accident situation, but ignore conditions that exist prior to accidents.This study aims at identification of factors affecting safety performance for substructure operations. Then their effect on the risk of accidents will be analyzed. The probabilistic occurrence of accidents will be estimated a fuzzy-based approach. The variation in productivity resulting from the probabilistic occurrence of accidents will be simulated using MicroCYCLONE to assess the impact of accident on productivity and any potential savings. This will provide insights into the relationship between the accident risk and its associated productivity variation. Finally, this study will adopt artificial neural networks (ANNs) to forecast the variation in productivity prior to the occurrence of accidents.