The Industry Foundation Classes (IFC) cover a wide variety of subdomains in the construction industry. Model View Definitions (MVD) enable to specify a subset of the IFC schema to assess the content of a model for specific use cases and information exchanges. However, IFC and MVD paradoxically complexify the workflow since it requires a deep understanding of the schema combined with construction knowledge to carry out simple use cases such as quantity checking or data export. This gap between domain specific queries and their expression in a computer-readable language weakens the opportunities provided to the building industry by Building Information Modeling. Our research consists in the implementation of MVDs in a high-level programming language to extract data from building models, an assessment of the extraction results and geometrical processing algorithms to correct the explicit quantities and properties that are supplied as metadata alongside the elements in IFC building models. Geometrical processing can be used to reduce and eventually correct errors on property values. We use a generic geometrical representation of IFC entity instances and apply geometrical transformations on those to obtain geometrical shapes. Boolean operations are used to identify relationships between elements. Eventually, incorrect data values are corrected directly in the IFC models accordingly to the IFC schema. For instance, we authored an MVD to extract data pertaining to external walls from different IFC models and corrected the value of the IsExternal property of the models’ IfcWall entities. This use case is of great importance for the cost estimation of a thermal renovation on a building as it gives a good estimate of the outer surface area of the building envelope.@en

Recent progress in computer vision and the pervasive use of Building Information Models (BIM) in the construction industry provide countless opportunities for decision making in architectural design. It allows capturing real-world geometrical data for assessing design options, such as windows products for replacement on the façade of an existing building. However, capturing information from the real world is a tedious task that involves data acquisition followed by costly processing steps. In this study, we focus on using a single 2D picture containing simple building elements in their context to generate a semantic 3D model of these elements. To that aim, we trained a detection neural network whose input constituted of labeled photographs containing windows. After training, the model is able to locate and segment elements of interest on unseen new data. With the use of Computer Vision methods, we derive accurate geometric outlines suitable to reconstruct photorealistic three-dimensional BIM models.@en