A Novel Framework for Estimation of the Maintenance and Operation Cost in Construction Projects: A Step Toward Sustainable Buildings

Abstract

Building maintenance and operation costs represent a significant portion of the life cycle costs (LCC) of construction projects. The accurate estimation of these costs is essential for ensuring the long-term sustainability and financial efficiency of buildings. This study aims to develop a novel framework for predicting maintenance and operation costs in construction projects by integrating an emotional artificial neural network (EANN). Unlike traditional models that rely on linear regression or static machine learning, the EANN dynamically adapts its learning through synthetic emotional feedback mechanisms and advanced optimization techniques. The research collected input data from 313 experts in the field of building management and construction in Ha’il, Saudi Arabia, through a comprehensive questionnaire. The integration of expert opinions with advanced machine learning techniques contributes to the innovative approach, providing more reliable and adaptive cost predictions. The proposed EANN model was then compared with a classic artificial neural network (ANN) model to evaluate its performance. The results indicate that the EANN model achieved an R2 value of 0.85 in training and 0.81 in testing for buildings aged 0 to 10 years, significantly outperforming the ANN model, which achieved R2 values of 0.78 and 0.72, respectively. Additionally, the Root Mean Squared Error (RMSE) for the EANN model was 1.57 in training and 1.60 in testing, lower than the ANN’s RMSE values of 1.82 and 1.90. These findings show that the superior capability of the EANN model in estimating maintenance and operation costs. This led to more accurate long-term maintenance cost projections, reduced budgeting uncertainty, and enhanced decision-making reliability for building managers.