A Semi-Automated Scan-To-BIM Approach to Fire Compartmentation in Existing Buildings

Abstract: Fire safety compliance checks in existing buildings are often labor-intensive and prone to error, due to the lack of design data and the reliance on traditional/manual inspection tools. This work proposes an alternative to conventional approaches through a semi-automated Scan-to-BIM methodology to create a BIM model directly from point cloud data. This model is specifically designed to facilitate the regulatory checks required for fire safety, particularly at the level of compartmentation, and can be applied to evacuation routes in later stages. The methodology involves four main phases: i) point cloud acquisition and registration; ii) point cloud processing; iii) geometric features acquisition, and iv) creation of spatial entities compatible with the IFC format. The methodology demonstrates robustness when applied to benchmark datasets. The results confirm the efficiency and accuracy of the methodology in developing early design and BIM models suited to support fire safety engineering and verification of compliance with fire safety regulations.