Application of Artificial Intelligence in Cost Overrun Prediction in Construction Projects

Application of Artificial Intelligence in Cost Overrun Prediction in Construction Projects

Seyed Azim Hosseini1 Reza Adiban2

1) Professor and faculty member, Islamic Azad University, South Tehran Branch, Tehran, Iran
2) PhD student in Civil Engineering, Department of Engineering and Construction Management, Faculty of Technology and Engineering, Islamic Azad University, South Tehran Branch, Tehran Province

IOI Code : AACZ-ABGFGD
Publication : The 5th International Conference on Seismic Retrofitting(5icst.ir)
Abstract :
    Construction projects are recognized as one of the main pillars of economic and infrastructural development of countries, however, one of the common challenges in these projects is the occurrence of cost overruns, which can increase final costs, delay project completion, and reduce investment productivity. Several factors such as incorrect cost estimates, design changes, material price fluctuations, poor project management, and unforeseen risks can contribute to cost overruns. Traditional cost forecasting methods, which are mainly based on classical statistical models and expert judgment, have limited accuracy in many cases due to their inability to analyze complex and nonlinear relationships between project variables. In recent years, rapid advances in the field of artificial intelligence have led to this technology being considered by researchers as a new approach to analyzing project data and predicting cost overruns in the construction industry. The purpose of this review is to investigate the applications of artificial intelligence in cost overrun prediction in construction projects and to identify the benefits, challenges, and research trends in this field. In this study, using a systematic literature review method, studies published in recent years on the application of artificial intelligence algorithms in cost prediction of construction projects have been reviewed and analyzed. In this framework, various machine learning and deep learning models, including artificial neural networks, support vector machines, decision trees, random forests, and metaheuristic algorithms, have been reviewed and compared. The results of the studies show that artificial intelligence-based models have higher accuracy in cost overrun prediction than traditional methods and are able to significantly reduce cost estimation errors by analyzing historical project data, identifying key factors, and modeling complex relationships between variables. These models can also be used at different stages of the project life cycle from planning to implementation for cost monitoring and control. However, challenges such as lack of high-quality data, limited interpretability of some models, implementation costs, and organizational barriers are still considered limiting factors for the widespread use of this technology.

Keywords : assessment construction projects risk management risk identification resource allocation