This project was developed during a hackathon focused on interoperability in 3D geospatial data workflows. The main goal was to design and implement an automated process that converts Industry Foundation Classes (IFC) building models into CityGML, ensuring compliance with OGC standards and cross-platform visualization compatibility.
Building Information Models (BIM) are often stored in IFC format, which prioritizes construction semantics but is not readily usable in 3D city modeling or geospatial applications. The challenge, therefore, was to create a streamlined pipeline that bridges the gap between building-centric IFC data and city-scale CityGML representation, enabling seamless integration into 3D city databases, web visualization tools, and urban digital twin platforms.
Tools and Environment
Software: FME 2025 Desktop
Source Format: Industry Foundation Classes (IFC, versions 2x3 and 4)
Target Format: OGC CityGML 2.0 and 3.0
Auxiliary Tools: FME Data Inspector, Cesium ion, 3DCityDB Importer/Exporter
Platform: Windows 11
Workflow Summary
The workflow was implemented in FME Workbench as a modular translation pipeline consisting of seven main steps:
IFC Reader
The IFC_ reader was used to import architectural elements such as walls, slabs, roofs, doors, and windows. Feature types like IfcBuilding, IfcWall, IfcSlab, and IfcRoof were selected to capture both external and internal structures. Parameters were configured to read IfcSpace geometries and ensure full representation context.
Geometry Processing
Raw BRep geometries from IFC were converted to meshes using the Triangulator transformer to ensure compatibility with CityGML. Multiple geometry fragments were unified via the MeshMerger, simplifying building representation.
Attribute Creation
A AttributeCreator transformer added mandatory CityGML attributes such as:
gml_id = @UUID()
citygml_class = e.g. Building, WallSurface, RoofSurface
lod = 2 or 3
name and description from IFC attributes
CityGML Geometry Setting
The CityGMLGeometrySetter (FME Hub version, 2025 update) defined geometry roles and Levels of Detail:
CityGML LoD Name: lod3Geometry
Feature Role: boundedBy, outerBuildingInstallation, or consistsOfBuildingPart depending on feature type.
IFC geometry (often in millimeters) was scaled to meters using Scaler, and reprojected using CsmapReprojector to a spatial reference (e.g., EPSG:25832 – ETRS89 / UTM zone 32N).
LocalCoordinateSystemSetter was used for proper georeferencing.
CityGML Writer
The OGC CityGML Writer exported the translated model to .gml format. Writer parameters were configured to:
Match schema version (CityGML 2.0 or 3.0)
Preserve gml_id and citygml_class
Output to output/citygml/building_model.gml
Validation and Visualization
The resulting GML file was validated and visualized successfully in multiple platforms:
FME Data Inspector
CesiumJS (via 3D Tiles conversion)
3DCityDB Importer/Exporter
Global Mapper
The output displayed complete geometry with LoD2/LoD3 semantics, confirming structural correctness and schema compliance.
Results
The final GML model accurately represented the original IFC features, including roofs, walls, and openings, while adhering to CityGML structure. All previously encountered validation errors (invalid lod3MultiSurface traits) were resolved by adopting correct geometry roles (lod3Geometry), ensuring compatibility with standard CityGML readers and visualization platforms.
Impact and Future Work
This project demonstrates the feasibility of automating IFC-to-CityGML transformation in FME without requiring external scripts. It enables faster creation of 3D city models from BIM data and supports workflows in:
Urban digital twin development
Smart city planning
Semantic 3D database integration
Future work includes extending the workflow to support CityGML 3.0 ADEs, semantic enrichment from IFC properties, and direct 3D Tiles output for real-time visualization.
Acknowledgement
Developed as part of a hackathon challenge on BIM–GIS interoperability, this project showcases the power of FME 2025 and open geospatial standards in advancing 3D city modeling workflows.
Would you like me to make this report downloadable as a formatted PDF (with logo space, title page, and section headers)? It would be perfect for including in your portfolio or GitHub project page.
