Upstream Tech, a subsidiary of Natel Energy, applied machine learning and satellite images to provide the industry with more accurate streamflow forecasts. Supported by a Small Business Innovation Research (SBIR) Phase I award, the company deployed HydroForecast, an artificial-intelligence-powered forecasting software package, at more than 300 locations, allowing a wide array of hydropower developers to plan, operate, and assess risks of hydropower facilities.

Forecasts from this software will enable hydropower developers and regulators to make more informed decisions that save costs, while also increasing operational flexibility and environmental performance of hydropower projects. To demonstrate the accuracy of its forecasting efforts, Upstream Tech competed in the Streamflow Forecast Rodeo, a forecast skill competition run by the U.S. Bureau of Reclamation, placing first in the all-around competition and winning 23 of 25 performance categories.

Encouraged by these results and an SBIR Phase II award, Upstream Tech will develop additional weather forecast resources using a scalable, resilient, and secure cloud-based system. The company will continue quantifying water power technologies’ technical and financial benefits to continue moving its forecasting technologies toward commercialization.