WeatherSens AI
WeatherSens creates local wind predictions for wind energy producers using drones and physics-informed ML.
Summary
WeatherSens builds local wind predictions for wind energy producers. VTOL drones capture atmospheric profiles over the region, feeding a physics-informed ML model that corrects NWP baselines. We sell forecast accuracy as a service to wind operators and BRP providers in Finland, where forecast error directly drives imbalance costs.
Problem
NWP models (HARMONIE, ICON, Euro1k) fail at local resolution. Wind operators face systematic forecast errors that translate directly into balancing market losses. Median BRP imbalance cost for wind is -11 €/MWh vs spot, with no good local correction tool available.
Solution
VTOL drones fly vertical atmospheric profiles over the region, providing initial conditions to a physics-informed ML model that produces local wind predictions, aiming to beat operational NWP baselines and other models on market on hub-height accuracy. Delivered as a forecast API.
Market
Onshore wind operators in Finland and Nordics with own electricity trading (BRP not outsourced). Primary targets: >50 MW portfolio operators. Secondary: independent BRP providers managing wind imbalance risk.
Customers & users
Validation interviews completed with Ilmatar Energy, EPV Energia, Neoen, Fortum, OX2, Vattenfall, Exilion. LOI sprint in progress.
Competitive advantage
End-to-end local wind prediction pipeline: drone-based atmospheric profiling feeding physics-informed ML. Competitors sell either forecast services or measurement tools, not a closed-loop prediction system built from first principles.
Traction
~10 active wind operator validation discussions. Aalto Prototyping Grant awarded (4K€). Accepted to Ignite accelerator (June 2026).