Falls from heights are common construction site accidents that can lead to serious injuries and fatalities. Nowadays, Building Information Modeling (BIM) offers an effective solution for planning and designing safety measures during both the design and construction phases. In Vietnam, safety plans are typically developed manually, relying on the expertise of HSE engineers. This study proposes a framework for developing a tool that utilizes BIM models to improve risk management and enhance safety design efficiency. BIM collects building geometry and environmental data to identify potential risks. Additionally, the framework effectively identifies and mitigates potential hazards through automated detection and visualization within a 3D environment by utilizing BIM, while Bayesian Networks (BNs) analyze this data to assess fall risks across various scenarios. The results are then processed by an optimization algorithm to develop effective safety plans. The anticipated outcomes and recommendations associated with the proposed framework can serve as a foundation for further research, enabling project managers or HSE managers to effectively review and refine the safety plan. Thus, minimize the occurrence of fall accidents, enhancing overall safety on construction sites.