The release showcases new features planned for integration into previous open-sourced releases. The updated map data now includes over six million building heights across 34,000 square miles in major U.S. cities such as Boston, Chicago and Santa Clara. This data was developed through a combination of efforts using several sources: the United States Geological Survey (USGS) 3D Elevation Program; building footprints from OpenStreetMap; Microsoft open building footprint data; and community sources.
The foundation’s goal is to build a map dataset that caters to the needs of map service providers in an evolving mapping market.
“The new data will support thousands of mapping and location-based applications, unlocking many new use cases,” said Marc Prioleau, head of business development, mapping and location at Meta, who is involved in the project. “This release in April is the first output from the project, whose work began in January. This is not a production release; it is intended to demonstrate the types of data Overture plans to build.”
AWS, Microsoft, Meta and TomTom launched Overture in December under the Linux Foundation. The intention is to build a community of companies, government agencies and other parties that collaboratively build high-accuracy, high-fidelity open map data as a shared asset that anyone can use for location-based applications.
The results should include a common, well-structured data schema to help make maps easier to use, and a global entity reference system to link entities from different datasets to their corresponding real-world entities.
Enhancing traditional maps through 3D data
The data provides a glimpse into the power of combining resources from the public and private sectors to build new and unique datasets.
The latest data also includes new open map data sources from Microsoft ML buildings, additional building heights, and added administrative-area translations. Overture Maps Foundation aims to integrate all these open data sources into a cohesive and easily usable set of data.
“In the April release dataset, we used building footprint data, aerial LiDAR data and a digital elevation model to infer building heights,” said Prioleau. “These heights can then be added to the building footprints to create a rough 3D shape. The benefit of this method is scale; it can be used to add 3D shapes to millions of buildings across the U.S. (or anywhere with the necessary open datasets). Individual buildings can then be edited with more detail as needed.”
Overture engineers used building footprints from Microsoft and OpenStreetMap to identify buildings’ locations. They then looked at the USGS 3DEP LiDAR data to get the absolute elevations of those buildings’ rooftops and subtracted the local ground elevation to get the building heights.
The core dataset in the April release is the Daylight Maps Distribution v 1.25 of OpenStreetMap (OSM). OSM is a worldwide crowdsourced map that millions of individual mappers around the world continuously update. It changes every minute as new features are mapped or corrected.
Providing street-accurate maps through real-time data
The foundation has implemented commercial-grade quality checks and multiple validation checks to ensure the data is error-free, stable and easy to use. Overture aims to expand on these quality checks to further enhance the data’s usability.
“The Daylight Maps Distribution of OpenStreetMap is a monthly release of OSM data that has been put through approximately 60 quality validation checks. As new data is added or edited, these checks ensure that the map does not break,” said Prioleau. “The system also checks for profanity and vandalism that individuals can sometimes introduce in a crowdsourced process. All of the results are documented in release notes that go with each new version of Daylight.”
Map service providers can now enjoy a more streamlined and predictable experience when using Overture’s data, all thanks to the brand-new data schema and global entity reference system. This innovative system allows additions of supplementary data to the map, increasing accuracy and adding context.
The data schema uses a standard data model to create software that can render data into maps, develop local search and navigation applications, and create data visualizations. The schema also informs developers on how the data is structured and organized, allowing them to build their software accordingly.
The global entity reference system provides a way to accurately integrate other spatial data into the map. For example, logistics companies can estimate precise delivery ETAs by adding real-time traffic data to roads, while restaurants, bars and other businesses can build richer profiles by adding social data such as reviews, pictures and ratings.
The system can even append temporary data such as road closures or construction to affected locations, providing a real-time snapshot of mobility within the city.
This rigorous schema not only offers developers clear structures, it is also extensible, able to accommodate new types of map data as they emerge. In essence, the new data schema and global entity reference system enable map service providers to offer users more accurate and comprehensive maps.
A future of opportunities for 3D map data
Overture intends its models to address the challenges inherent in building cooperation across many entities. “There is very strong market interest in building an open map dataset that can be a shared asset for all companies,” explained Meta’s Prioleau. “The costs and complexity of building the quality of map data expected by users have grown beyond the capabilities of any one company. That strong interest and alignment give the motivation to overcome challenges. By building Overture Maps within the Linux Foundation, we learned the best practices from the leading open-source collaborations.”
In future releases, the foundation plans to introduce a structured data schema that will simplify the use of maps, and a global entity reference system that will link different datasets to real-world entities.
Overture acknowledges the challenges ahead but is optimistic about the progress it has made and the potential for growth. The foundation’s ultimate goal is to build an open map dataset that anyone can use and that will expand the usability of location as a critical part of the world’s shared infrastructure.
“In the near future, by further collaboration through information and resource sharing, the map data will be more accurate, more complete and more up-to-date than it could be if any single entity developed it,” Prioleau added. “By sharing a common spatial framework, services can interact and interoperate with each other, rather than being confined to one mapping service or another.”
For those interested in learning more or contributing to the project, the foundation encourages feedback and welcomes new members to help build the open map dataset of the future.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.