Guide
Site plan data for architects: geoportal, OpenStreetMap or CAD ATLAS?
There are three ways to get the data for a site plan: official geodata from the geoportal of your region, raw data from OpenStreetMap, or CAD ATLAS, which delivers buildings, streets, terrain, trees and water as a DWG file with sorted layers. Which one fits depends on whether you want to process raw data or start working on the plan right away.
The three ways compared
| Source | What you get | Formats | Steps to a plan |
|---|---|---|---|
| Geoportal / survey office | Parcels (ALKIS in Germany), aerial images, terrain model, 3D buildings | NAS, Shapefile, CityGML, GeoTIFF | Join tiles, convert, set up the layers |
| OpenStreetMap | Streets as centre lines, buildings, paths | OSM, GeoJSON | Draw street widths, surfaces and layers yourself |
| CAD ATLAS | Buildings with heights, true-width streets, contour lines, trees, water, shadows | 2D: DWG. 3D: DWG, 3DM, 3DS, OBJ, SKP, SAT | Open the file and start working |
What is in a CAD ATLAS file
- Streets at their real width, with continuous edges instead of centre lines (more)
- A height for every building. Where none is published, our own AI adds it.
- Vector shadows that give the plan depth and make it easy to read (more)
- Contour lines from terrain models, measured at the ground and not at the treetops (more)
- Trees from OpenStreetMap, city tree registers and AI detection
- Every element on its own layer, clipped to your site
Where the data comes from
A competition entry or a planning submission can't afford a licence problem. A source goes into CAD ATLAS only if it is open data, allows commercial use and can be credited the way its licence asks.
- OpenStreetMap for roads, railways, water, land use and mapped trees
- Overture Maps for building footprints and heights
- National mapping agencies for high-resolution aerial imagery and elevation
- Cities worldwide for their official tree registers
- Global bare-earth elevation for terrain at any site on Earth
On top of these we run our own technology: AI height prediction, canopy detection and the geometry engine that turns raw data into clean drawings. We leave out commercial satellite basemaps, data under non-commercial licences and anything with unclear terms.
Good forCompetitions · Design · Presentation plans · Figure-ground plans · Urban analysis · 3D massing models · Context models