Risk Management in Construction Projects using Bayesian Networks

The construction industry is inherently exposed to a wide range of risks, including cost overruns, schedule delays, safety hazards, and quality issues, which can significantly impact project performance. Traditional risk management approaches, such as probability-impact matrices, fault tree analysis (FTA), and Monte Carlo simulations, often fail to capture the interdependencies and dynamic nature of construction risks. To address these limitations, this study proposes a Bayesian Network (BN)-based framework for managing risks in construction projects. Bayesian Networks offer a probabilistic graphical model capable of representing cause-and-effect relationships between risk factors, making it possible to perform dynamic risk assessment and scenario analysis. By integrating historical data, expert knowledge, and real-time project updates, the BN model enables continuous monitoring and prediction of risk probabilities. The results demonstrate that the BN-based system improves risk prediction accuracy, enhances decision-making capabilities, and supports proactive risk mitigation strategies, ultimately leading to more resilient and efficient construction project management.