Explore Architectural Data

Discover and analyze comprehensive architectural datasets with interactive visualizations and downloadable resources.

Project Workflow

This research project presents a comprehensive methodology for enriching architectural datasets with performance data. Our workflow integrates IFC building models with energy and daylight analysis to create enhanced datasets for machine learning and architectural research in multiple formats.

Project Workflow Diagram

Comprehensive workflow for performance data enriched floorplan datasets

IFC Authoring

Using the Geometry Gym plugin, GH trees are mapped to the correct IFC hierarchy (Project → Site → Building → Storey → Elements), ensuring semantic validity and robust relationships for downstream queries.

Geolocation & Climate Context

Using the CityWeft API, models are placed in real-world contexts. 10 climate zones are defined and candidate locations pre-selected. A positioning script matches each building to an optimal site based on typology, window distribution, and height.

Performance Simulations

Solar radiation, daylight, and energy studies are conducted. Results are stored at the appropriate granularity (room, window, apartment).

Data Embedding & Structuring

Simulation outputs are written back into IFC as custom properties and exported in structured CSV/XML to support analytics and ML pipelines. Rich building graphs are also generated from these enriched IFCs.

Models preview

Interactive Visualizations

Explore architectural adjacency graphs showing spatial relationships within buildings. Pan, zoom, and interact with graph visualizations.

Loading Building 1006 Adjacency Graph...

37+ architectural adjacency graphs available

Auto-loading sample visualization on page load

Supported features:

  • Pan and zoom with mouse/touch
  • Search graphs by building number
  • Filter by graph type
  • Random graph exploration

Dataset Explorer

Interactive preview and filtering of architectural data. Choose your ranges.

Select columns above to add filters

Dataset Preview

Loading...

Dataset will load here...

Download Dataset

Access comprehensive architectural dataset for research and analysis.

Full Dataset

Complete architectural dataset with all measurements and metadata.

CSV Dataset

Complete dataset in CSV format for data analysis and research.

Documentation

Data dictionary and usage guidelines.

Swiss Dwellings

Original Swiss residential building dataset used as foundation.

IFC

3D building information models in Industry Foundation Classes format.

Graphs

Interactive adjacency graphs and network visualizations.

Dataset Creation Process

Dataset enhancement workflow consists of multiple stages, each building upon established architectural processes and simulation methodologies.

01

Foundation Dataset: Swiss Dwellings

A large dataset of real apartment models saved in csv which can be recreated as geometry using Python.

  • 3,000+ residential building units
  • Detailed room-level spatial data
  • Authentic Swiss architectural typologies
  • Custom python node for CSV to geometry
View Original Dataset →
02

3D Model

IFC models created with precision and enhanced properties for comprehensive building information modeling.

  • IFC models with Geometry Gym
  • Adding properties to IFC
  • Precise geometry
Visit Geometry Gym →
03

Real World Context

Integration with real-world environmental data to provide comprehensive contextual information for each building.

  • CityWeft API integration
  • 10 climates/cities coverage
  • Matching building characteristics with city
Visit CityWeft →
04

Performance Simulation

Each building geometry underwent comprehensive environmental performance analysis using industry-standard simulation tools.

  • Energy demand calculations across climate zones
  • Daylight factor analysis with high-resolution sensor grids
  • Solar radiation and sun hour modeling
  • Window performance optimization studies
Visit Pollination →
05

Data Enhancement & Integration

Raw simulation outputs were processed, validated, and integrated with spatial data to create our enriched architectural dataset.

  • Automated data validation and quality control
  • Graph network generation for ML applications
  • Multi-format export (CSV, IFC, JSON)
  • Interactive visualization creation
×