3D Tiles Explained: OGC, Open Standards, Cesium & Open 3D Data

If you’ve ever tried to load a big 3D model of a city or infrastructure project in your browser, you know the pain: long load times, laggy navigation and sometimes a browser crash. That’s the problem 3D Tiles were built to solve. They make it possible to stream huge, detailed 3D datasets smoothly online.

In this article, I’ll break down what 3D Tiles actually are, how they came to be, the role of standards like the Open Geospatial Consortium (OGC) and how tools like Cesium make them usable in the real world. I’ll also touch on open 3D data and where things are heading next.

What Are 3D Tiles?

3D Tiles are a format designed to stream 3D geospatial content over the web. They let you display huge datasets like photogrammetry models, LiDAR point clouds and BIM data in a way that performs well. Instead of loading everything at once, 3D Tiles break the data into smaller chunks (tiles) and only load what’s needed as you navigate.

It’s kind of like how Google Maps only loads the part of the map you’re looking at. If you pan or zoom, new tiles load in. Same idea, but with full 3D models.

Why They Were Needed

Before 3D Tiles, most 3D formats were designed for individual objects (like a car or a building), not entire cities or landscapes. Formats like OBJ or COLLADA weren’t built for streaming massive geospatial datasets. You had to download the whole thing upfront, which just doesn’t scale when you’re working with real-world environments captured from drones or satellites.

When we started experimenting with photogrammetry of mine sites years ago, we ran into this issue constantly. Beautiful models, terrible performance. 3D Tiles changed that. They made it feasible to share and explore large-scale 3D scenes online without needing high-end hardware.

Cesium and the Origin of 3D Tiles

3D Tiles were created by the team at Cesium, an open-source platform for 3D geospatial applications. They needed a way to serve up massive 3D datasets efficiently and decided to build a new format from the ground up. That became 3D Tiles.

CesiumJS, their JavaScript engine, was built to display this kind of content. Later, Cesium Ion made it easier for users to upload and convert their own data to 3D Tiles. Cesium open-sourced the format and advocated for it to become a standard, which led to OGC stepping in.

What is the OGC and Why Does It Matter?

The Open Geospatial Consortium (OGC) is the international standards body for geospatial tech. If you’ve worked with formats like GeoTIFF, WMS or GeoJSON, OGC probably had a hand in shaping them.

OGC recognised the value of 3D Tiles and made it an official Community Standard. That matters because it helps ensure interoperability between tools. In plain terms, it means if you publish data as 3D Tiles, there’s a better chance it will work across different platforms and software. That’s a big deal if you want to avoid vendor lock-in or need to integrate with other systems.

Under the Hood: How 3D Tiles Work

At a technical level, a 3D Tiles dataset is made up of a tileset.json file (which defines the structure) and a bunch of tile files. These tiles might contain:

• B3DM for 3D models like buildings
• PNTS for point clouds
• I3DM for instanced features like trees or street lights
• GLTF/GLB for general 3D geometry

The tileset defines how all these pieces fit together. It uses bounding volumes to define which tile covers which area and level of detail (LoD) to decide which version of the tile to load based on your camera position. If you’re zoomed out, you get low-detail tiles. Zoom in and higher detail loads in dynamically.

This hierarchical structure is what makes it scalable. You never have to load the whole dataset at once.

Real-World Uses of 3D Tiles

We’re seeing 3D Tiles used in all kinds of industries:

• Smart cities: digital twins of city infrastructure
• Mining and energy: remote site planning with photogrammetry or LiDAR models
• Transport and utilities: modelling corridors like railways or power lines
• Tourism and culture: online experiences of museums, heritage sites, or walking tours

One of our clients uses 3D Tiles to visualise drone captures of a power transmission corridor through a national park. Being able to stream that data and overlay it with planned upgrades has completely changed how they review and approve work.

3D Tiles Beyond Cesium

Although Cesium pioneered the format, 3D Tiles are now supported across a growing number of platforms. You’ll find plugins or tooling for Unity, Unreal Engine and even some GIS platforms. Some users also convert 3D Tiles to GLTF if they need wider compatibility, though you lose things like the LoD hierarchy and spatial indexing.

That said, adoption is still growing. While Cesium has the best support by far, the open standard helps ensure that other platforms can catch up or build their own implementations.

The Promise of Open 3D Data

3D Tiles work really well with open datasets. Cesium itself publishes free tilesets like OpenStreetMap Buildings. Some governments have started releasing LiDAR scans and photogrammetry as 3D Tiles too.

This opens up a lot of opportunities: better public access to urban data, easier collaboration between planning teams and new tools for education or research. It’s one of the reasons we built Swyvl to support 3D Tiles - because they make it easier to bring together all sorts of spatial data in one place, without needing proprietary formats.

Challenges to Keep In Mind

There’s no silver bullet format and 3D Tiles are no exception. Here are a few things to be aware of:

• The spec is powerful, but not always beginner-friendly
• File sizes can still be large, depending on the source data
• Hosting requires good infrastructure, especially for heavy datasets
• Viewer support is improving but still uneven outside Cesium

We’ve worked with clients who needed to convert Revit files, drone models and point clouds into 3D Tiles. Each one needed different tools or workflows and it’s rarely a one-click process.

What’s Next for 3D Tiles?

The standard continues to evolve. 3D Tiles 1.1 brings better metadata support and other improvements. More open-source tools are popping up to generate and host tilesets. Integration with game engines is improving, especially as digital twins and immersive experiences become more common.

As the ecosystem matures, we’re likely to see more collaboration between open-source communities, standards bodies and industry platforms. That’s good news if you’re trying to future-proof your 3D workflows.

Wrapping Up

If you’re working with large-scale 3D data - whether it’s photogrammetry, point clouds, BIM models, or any spatial dataset - 3D Tiles is probably worth learning. It’s not the easiest format to start with, but the payoff is huge if you need to share or visualise complex environments.

Want to see what 3D Tiles look like in action? We’ve started adding support for them in Swyvl, so you can position them on a map right alongside your 360 photos, still photos and 3D models. It’s a great way to bring together different types of spatial data without needing custom software.

Let me know if you’d like to try it out or have questions about converting your own data into 3D Tiles. Happy to help and we have some additional resources over in the Swyvl Help Centre, including step-by-step instructions for exporting your 3D Tilesets from Reality Capture.

Additional Resources

🔗 3D Tiles Specification (OGC Community Standard) The official documentation from OGC for the 3D Tiles format.

📘 Cesium: Tile 3D Content with Cesium A series of explainers, straight from the creators of the format.

🎥 Cesium’s YouTube Channel Includes tutorials and demos, including how to work with CesiumJS and Ion.

📁 Cesium OSM Buildings Dataset (Open Data) Streamable open dataset using 3D Tiles, great for testing and exploration.

📦 RealityCapture to 3D Tiles Export Guide Helpful if you're working with photogrammetry.

📄 Swyvl Help Centre: Import and Visualize 3D Tiles from RealityCapture Step-by-step instructions for getting your data ready for Swyvl.