BIM Spatial Models for Construction Dependency Inference and Scheduling Optimization

This dataset contains three Building Information Modeling (BIM) projects in IFC format, along with their corresponding processed semantic data used for spatial reasoning, dependency inference, and scheduling research. Each BIM model folder includes both the original IFC file and several CSV files that provide structured geometric, spatial, and semantic information extracted from the IFC model.The dataset consists of:
1. BIM_meeting_room_data
A small-scale architectural BIM model representing a typical meeting room.
Included files:
BIM_meeting_room.ifc – the original IFC model
ifc_family_data.csv – component family/type attributes
ifc_space_centroids_data.csv – centroid coordinates of spatial units
ifc_space_data.csv – geometric and semantic space-level attributes
lowest_points.csv – elevation and placement reference points
spatial_relationships_detailed.csv – detailed adjacency, containment, and interaction relationships
2. BIM_residential_building_data
A medium-scale multi-room residential building model.
Included files:
BIM_residential_building.ifc – the original IFC model
ifc_family_data.csv – families and object classification
ifc_space_centroids_data.csv – geometric centers of all rooms
ifc_space_data.csv – IFC-derived geometric features
lowest_points.csv – base elevation and alignment information
spatial_relationships_detailed.csv – full spatial interaction matrix
3. rac_advanced_sample_project_data
A large and complex BIM project commonly used in industrial design workflows.
Included files:
rac_advanced_sample_project.ifc – the complete IFC model
ifc_family_data.csv – component types and metadata
ifc_space_centroids_data.csv – centroid positions for rooms and areas
ifc_space_data.csv – extended geometric/semantic features
lowest_points.csv – elevation anchors
spatial_relationships_detailed.csv – multi-type spatial relationship details
This dataset supports research in:
BIM-based spatial reasoning
Graph neural network (GNN) dependency inference
Construction scheduling optimization
Automated semantic enrichment
IFC-to-graph conversion workflows
All models and CSV files in this dataset were used to evaluate the proposed TSG–GNN–DLSA scheduling framework in the associated publication.