Dynamic risk assessment of tower crane operations by integrating functional resonance analysis method and Bayesian network

Tower cranes are vital to modern construction but pose significant safety risks. While existing studies primarily focus on risk identification and evaluation, they often neglect the complex interactions and dynamics of these risks. This study proposes a comprehensive framework for understanding and mitigating tower crane operation risks by integrating the Functional Resonance Analysis Method (FRAM) with Bayesian Network (BN). The FRAM model identifies key functions and their interdependencies, which are analyzed through Monte Carlo simulations. The results are transformed into BN nodes, forming a network that employs Bayesian inference to assess overall risk levels. The framework was validated in a real-world construction project, where it revealed that the tower crane operations were generally safe, with critical focus areas identified as "Tower Crane Components," "Tower Crane Installation Acceptance," and "Slings and Hoisting Objects." By combining both static and dynamic data, this framework enhances risk assessment and contributes to safer construction practices. • Identify key couplings in tower crane operations using the FRAM model for analysis. • Improve risk inference by combining expert judgment with case data insights. • Enhance risk evaluation by integrating both static and dynamic operational data. • Apply the proposed framework in a case study to show its practical effectiveness.