Wind is the largest source of clean, renewable energy in the US, comprising over 9% of total generation and almost half of all renewable energy generated. Wind turbines produce a total of 3x more energy than solar panels. [eia.gov]
The vast majority of turbines today aren't able to adjust to wind changes until after winds reach turbines — because this critical data arrives too late, turbine control systems are constantly surprised by changes. This results in lost energy, and strong gusts cause expensive damage to turbine components such as gearboxes.
Using patented methods with an array of pressure sensors on the landscape and modern Machine Learning, we provide 15-60 second alerts of wind changes and gusts before they arrive, increasing energy production, turbine life, and project ROI. Wind farm subscribers will see payback in just a few months.
Wind changes at turbine height are caused by and cause pressure changes; these changes propagate outward and are picked up by pressure sensors near the ground, which share their data through a wireless mesh network.
Sensor readings and site anemometer data are collected at a central node and ingested by our machine learning algorithm that models in real time the airflow at each turbine, both at the present moment and, by “nowcasting”, 15-60 seconds into the future.
Nowcasts are sent to wind farm controllers and SCADA systems, enabling turbines to use preview control to produce more energy and revenue and to avoid expensive damage from unforeseen gusts.
CEO / Cofounder
CTO / Cofounder
Lead ML Scientist