Demo: A real-time 3D digital twin built on Microsoft Fabric
Microsoft Developer demonstrates a real-time 3D digital twin of a wind farm built in Microsoft Fabric, combining LiDAR-derived terrain geometry, live turbine telemetry from a Fabric Eventhouse, and a natural-language interface that can identify underperforming turbines and estimate cost impact.
Overview
The video shows a simulated wind farm based on a real site in Sievi, Finland:
- The terrain is built from open laser-scanning (LiDAR) data.
- Turbines are fed by live telemetry coming from a Microsoft Fabric Eventhouse.
- The experience supports interactive 3D exploration (orbiting the site and inspecting turbines component-by-component).
- A natural-language interface (typed or spoken) can be used to ask questions like which turbine is underperforming and what it is costing.
Architecture: three refinement pipelines feeding one scene
For this digital twin, three pipelines feed the final 3D scene:
Geometry pipeline (LiDAR to mesh)
- Starts from LiDAR / laser-scanning data.
- Produces a mesh representation suitable for the 3D runtime.
Telemetry pipeline (Eventhouse to tools)
- Live turbine telemetry is ingested into Fabric Eventhouse.
- That telemetry is exposed to downstream tools used by the runtime experience.
Language pipeline (tools to agent)
- A language interface (typed or spoken) is connected to an agent.
- The agent uses tools backed by the telemetry pipeline to answer operational questions (for example, identifying underperforming turbines and estimating cost impact).
Design principle used across pipelines
The build applies the same discipline across geometry, telemetry, and language pipelines:
- Reduce data as far upstream as possible.
- Keep the runtime thin.
- Generate downstream artifacts from a single source of truth.
Link
Learn more: https://commandline.microsoft.com/fabric-rayfin-lidar-real-time-3d-digital-twin/