Overture 2023-04-02-alpha is a pre-release version of open map data intended primarily to demonstrate the potential and possibilities for open map data that the Overture Maps Foundation intends to pursue. It is primarily for demonstration and not as a production release.
This release is built on the Daylight Map, a distribution of OpenStreetMap that builds an OSM planet file that is quality checked for errors and vandalism. Daylight also provides additional name translations, building footprints, and other data not in OpenStreetMap in the form of “sidecar” files which add data coverage or other improvements.
License:
All data included in Overture 2023-04-02-alpha that is licensed under ODbL (including OpenStreetMap data) or derived from ODbL data that constitutes a “Derivative Database” (as defined under ODbL v1.0), is licensed under ODbL v1.0.
Data from sidecar files which was not previously licensed under ODbL is licensed under CDLA Permissive 2.0.
The geojson files which contain buildings only are licensed under ODbL v1.0.
Overture 2023-04-02-alpha makes substantial improvements over Daylight in several ways.
Quality Validation
Overture 2023-04-02-alpha leverages the validation and quality checks that were introduced for the Daylight map distribution. This includes both automated and manual validation checks to detect map errors, breakage, profanity and vandalism in source data. This results in a safer version of open map data. Overture 2023-04-02-alpha uses Daylight v1.25 as a baseline. The results of these validation checks can be found in the release notes here.
Integrated Planet PBF File
Daylight introduced the notion of “sidecar” files, OSM change files that allow users to include additional data on top of the base map dataset, such as ML buildings, roads, and administrative areas. Sidecar files require additional processing and tooling in place to use and can be difficult for some users to apply. For Overture we are including all data directly in a single planet PBF file. This first release contains core basemap data along with Microsoft ML buildings, additional building heights (including on ML buildings), and additional administrative area translations all in one file. In total, we added over 177 million additional building footprints. OSM tooling that can process a PBF should be able to use this planet file seamlessly.
New 2.5D Building Data
There are over six million new 2.5D buildings in the database. These were created using lidar data from the USGS 3D Elevation Program (3DEP) combined with open data building footprints from OpenStreetMap and Microsoft. More details are available in this Technical Spotlight. Building heights from lidar data are included in the “est_height” tag on building ways.
Normalized Building Heights
Building heights in OSM data can be specified in different conventions. Some example height tags include: “11 m”, “8.5 meters”, “353 ft”, “124 feet”, and “19’8″”. Downstream users need to be aware of these different conventions and parse them accordingly. To make it easier for data consumers, we are converting any building height value into meters and replacing it in our output. Overture users can treat height values as meters without any further parsing or logic. We omit the height value when it cannot be converted into meters – real examples include ”Unknown” or “half”. While starting with building heights, we plan to provide similar normalizations over other tags in future releases.
Including Only Relevant Tags
Many OSM ingestion tools ignore tags that are not useful to rendering, routing and other applications. We remove any tags like: note:*, odbl*, gnis:*, geobase:*, KSJ2:*, yh:*, osak:*, kms:*, ngbe:*, naptan:*, CLC:*, source:*, tiger, logging*, ‘source_ref’, ‘SK53_bulk:load’, ‘attribution’, ‘comment’, ‘accuracy:meters’, ‘sub_sea:type’, ‘waterway:type’, ‘3dshapes:ggmodelk’, ‘import’, ‘created_by’, ‘fixme’, ‘osak:identifier’, and ‘nysgissam:nysaddresspointid’. This removes approximately 188 million tags from the planet file which helps to reduce file size and improve processing time. We plan to expand this list and remove tags in future releases that are unnecessary for end-user applications.
Attribution
- © OpenStreetMap contributors available under the Open Database License (www.openstreetmap.org/copyright)
- U.S. Geological Survey, 2019, USGS 3D Elevation Program Digital Elevation Program
- Building data © OpenStreetMap contributors, Microsoft https://github.com/microsoft/GlobalMLBuildingFootprints, Esri Community Maps contributors
