The atmosphere, measured.
Atmospheric measurement and physics-informed forecast correction for wind energy.
One grid cell
One site
01Problem
They model the atmosphere. They don't measure it.
Worst at Hub height Complex terrain Icing
02The Insight
The data gap isn't a model problem. It's a measurement problem.
You can't correct what you can't see.
We fly through the atmosphere and measure it directly.
A curtain of measurements every hour
03Solution
One pipeline, three moves.
From the air above the site to a corrected forecast.
-
01Measure
VTOL drones profile the vertical wind layer that regional models only approximate: near site, at hub height.
-
02Model
Windwall, a physics-informed neural operator, detects when HARMONIE or ICON is failing at a site and anchors the forecast to real measurements.
-
03Deliver
A corrected intra-day forecast, delivered via API, priced as a share of measured savings against the free IFS baseline.
04Why we're different
Three moves, one owner: sensor → model → forecast.
The market today
- Hardware or data vendors: no scalable vertical stack available.
- AI weather labs: model the atmosphere without measuring it.
- In-house tooling: “basic ML correction”, how operators describe what's offered today.
WeatherSens
- Owns the measurement: VTOL drones profile the vertical wind layer at the site itself.
- Owns the model: Windwall, physics-informed correction, anchored to its own observations.
- Owns the delivery: intra-day forecast via API, priced on measured outcomes.
05Windwall
Physics-informed from the start. AI-native by design.
Models drift
Measurements anchor them
06Vision
Mapping the whole world.
To make the atmosphere measurable everywhere it matters, starting with the turbines that power the world.
From air to action.
Every place deserves weather intelligence as precise as its own air.