Development of Safety Level Assessment Models for Building Projects using Multiple Linear Regression and Bayesian Network

The construction industry accounts for a large share of the industry in various countries. The rate of accidents and deaths in this industry is much higher than in other industries. Safety is a major concern in the construction industry. One of the most dangerous parts of the industry is the construction of high-rise buildings. In this research, an attempt has been made to evaluate the level of factors affecting the safety of the project, to be able to predict the level of safety of future projects based on the records and information of previous projects. Because with a correct understanding of the impact of factors on safety, it is possible to take the necessary policies to improve safety and reduce damage to this part of the industry and prevent the waste of resources. To achieve this purpose, first, the factors affecting safety have been identified by reviewing the literature and reviewing the unofficial statistics of work-related accidents, and then the relationship between these factors and the level of project safety for 95 projects in Tehran and Kish Island is developed. It was conducted with the help of a questionnaire, with the help of a Bayesian network and multiple linear regression. In the proposed model, the effect of each of the groups affecting the safety of the project is determined. “Monitoring and supervising the safety" and "safe process of doing work" are among the most influential factors on the safety level of high-rise building projects; regression coefficients were 0.338 and 0.264, respectively.