This incredible map shows the world’s 2.75 billion buildings

From the latest skyscraper in a Chinese megalopolis to a six‑foot‑tall yurt in Inner Mongolia, researchers at the Technical University of Munich claim they have created a map of all buildings worldwide: 2.75 billion building models set in high‑resolution 3D with a level of precision never before recorded.

Made from years of satellite data analysis by machine‑learning algorithms, the model reflects a sustained effort to capture the built world in three dimensions. The result now provides a crucial basis for climate research and for tracking progress toward global sustainable development goals, according to the scientists behind it.​

Professor Xiaoxiang Zhu, who leads the project and is the chair of data science in Earth observation at TUM, says the real achievement is that the new map is a three‑dimensional picture of how much space people actually inhabit. “3D building information provides a much more accurate picture of urbanization and poverty than traditional 2D maps,” she explains. With 3D models “we see not only the footprint but also the volume of each building.”

At the heart of this work is the GlobalBuildingAtlas, an open dataset that describes individual buildings across the planet both as 2D outlines and as simple 3D objects. In total, it contains 2.75 billion building footprints—polygons tracing the edges of each structure—covering every building the satellites could detect in satellite imagery from 2019.

At first glance, there are some interesting takeaways from the map, like the distribution of building volume clusters around major metropolitan regions—with particularly dense concentrations in East Asia, Europe, and North America. Meanwhile, many parts of the Global South show vast numbers of buildings that are small and low‑rise, especially in Africa, which has more buildings than Europe and North America, but far less total built area and volume.​

The ability to map building height and volume reveals disparities that conventional 2D maps tend to hide: A dense informal settlement and a carefully planned neighborhood of multi‑story buildings can look similar in a flat, area‑based statistic. But if you have accurate 3D buildings, experts can understand that they offer radically different housing conditions and require different infrastructure. Their proposed metric of building volume per capita turns the GlobalBuildingAtlas into a lens for spotting where housing and infrastructure lag behind population and, therefore, where urban policy and investment should concentrate.​

How they made it

The scientists used machine learning algorithms to identify one billion more buildings than any previous global database, creating simplified 3D “shoebox” models for 97% of them. That’s 2.68 billion 3D buildings, compared to Google Open Buildings, which has 1.8 billion building outlines.

The team started with daily satellite images from the PlanetScope constellation, which photographs the Earth at roughly 9.8 feet per pixel. Then they stitched together about 800,000 cloud-free scenes from 2019 into a seamless global mosaic, and taught a neural network to recognize buildings by training it on known building outlines from OpenStreetMap and other sources.​

To add height to these flat building outlines, the team used laser measurements (LiDAR) from airborne surveys in developed countries to train an AI that can estimate how tall a building is just by looking at a single satellite photo—similar to how a person can judge a skyscraper’s height from its appearance and shadow. This height-prediction model scans the entire global image and assigns a height value to every pixel, even calculating its own margin of error.​

The final step merged the building shapes with the height data. Because no single map source is perfect everywhere, the researchers created a system that picks the best available building footprints for each region—usually OpenStreetMap, or Google’s data in Africa and South America—then fills in gaps with secondary sources like Microsoft’s footprints or their own AI-generated outlines. Finally, they assign each building its height by taking the tallest point within its footprint.

Unbelievable detail

It’s clever and, yes, it’s a crazy intense workflow with a mind-numbing number of processing steps. It’s amazing to know it worked as well as it did. If you try it, you can find anything in the world, anywhere. Even shanties in the middle of the forest or the desert.

When they summarize the results, the atlas reveals a built world that may look just like your regular Google Maps, but it’s so much bigger, more granular, and more accurate. Globally, the 2.75 billion buildings occupy a roughly 195,600 square-mile ground area and add up to roughly 65.32 trillion cubic feet. The authors are so confident that they believe that an earlier United Nations estimate of about four billion buildings worldwide is wrong. 

The research paper says that its continental statistics amount to a numerical X‑ray of global development. Asia leads with roughly 1.22 billion buildings and the largest cumulative volume, reflecting both its population and its tall, dense urban cores, whereas Africa has on the order of 540 million buildings yet only a fraction of the built area and volume, which the authors interpret as a sign that many African structures are small, single‑story or “informal” units.

Europe and North America contain fewer buildings than Africa, but much higher average volume per building, South America and Oceania sit between these patterns, and South America stands out with the largest height and volume errors in the dataset—something the team links to a complex mix of informal settlements and vertical developments that are harder for the model to represent accurately.​

Lessons to be learned

In their paper, Zhu and her colleagues say that the key conceptual move is to go beyond the United Nations’ current indicator for Sustainable Development Goal 11, which aims to make cities around the world safer, greener, more resilient, and more affordable to live in. Right now, the UN evaluates this goal on built‑up land area relative to population growth. This 3D map accurately treats cities as fully three‑dimensional objects and will enable the UN and other organizations to make better evaluations and take better decisions.

Because the data and code are openly available, the atlas is already being pulled into other efforts. In its press note, the university notes that the German Aerospace Center (DLR) is exploring how to use GlobalBuildingAtlas within the “International Charter: Space and Major Disasters,” which coordinates satellite support for emergency response after floods, earthquakes, and other catastrophes.

The scientists emphasize in their work that the same 3D building information can refine models of energy demand and CO₂ emissions, guide decisions on where to add housing or public facilities such as schools and health centers in disadvantaged neighborhoods, and support the planning of heat‑resilient and flood‑resilient green infrastructure as cities adapt to climate change.

So it’s not about doing something cool for the sake of it, but about creating a measuring instrument for climate scientists, urban planners, economists, and disaster managers. Now they will be able to, for the first time, run their models on a global level and anywhere at the national, regional, or local level.

View the full article