Safety Cost Estimation of Building Construction with Fuzzy Logic and Artificial Neural Network

Abstract

The number of cases of construction work accidents continues to increase every year. To prevent work accidents, especially the fatality rate, Occupational Health and Safety (OHS) must be implemented in the construction project. Implementation of OHS can be a success if the availability of the budget is allocated explicitly for implementing OHS in construction projects. However, respondents’ current OHS budget is currently insufficient when referring to the guidelines regarding OHS costs applied in Indonesia. This condition will increase the initial budget and cause financial losses. So it is necessary to develop a cost estimation model that can estimate costs quickly and accurately. The results from this study will help estimator make OHS cost estimates quickly and accurately so that an estimator does not need a long time to do cost estimations at the beginning of the project, and the results of the estimated cost estimates are accurate. The fuzzy method uses to classify the contract value. The output of the fuzzy process will use as input in learning with the neural network. The Mean Absolute Percentage Error of tested data set for the adapted model is highly accurate (9.906%). The model obtained has a better MAPE value than the estimated cost by using regression analysis.