Simplification and enrichment of IFC models for cost estimation

Abstract

Building Information Modelling (BIM) has become a cornerstone for data-driven project management. However, the information requirements for BIM models used in cost estimation are often too high, making them difficult to meet, especially during the early stages of a project. This study explores the possibility to reduce these requirements by using semantic enrichment (SE) for extracting additional insights from low-detail models. To address this, key information necessary for cost estimation was identified, distinguishing between the parameters best extracted directly from the model and those that can be computed algorithmically through geometric and topologic analysis. A custom algorithm was then developed to calculate these properties automatically by either generating new construction parts or recalculating existing ones and evaluating their properties (such as complexity). The approach was tested on 30 sample models and evaluated by a panel of 16 architects and BIM/digitalization experts. While the solution works with complex models, its core approach is the intelligent analysis and enrichment of volumetric models. To derive necessary information from higher level models, simplification and abstraction was adopted. The outcomes confirmed that the code successfully enriched models with varying levels of detail and the enriched BIM models could provide the required data for the cost estimation effectively. Detailed cost estimation uses 180 parameters: currently 45% are extracted from volumetric models, 45% from low-detail models, and 10% added manually. The algorithm will be integrated into a productive web-based cost estimation tool. This method significantly accelerates BIM based cost estimation during early design stages, reducing manual effort and reliance on detailed models. Additionally, the procedure is adaptable to other use cases, such as life cycle cost calculations, further enhancing the practicality and adoption of BIM workflows in early-phase decision-making.