In a recent development, DeepMind revealed that it has used AI to enhance wind energy production. Acquired by Google in 2014, DeepMind is now under Alphabet, Google’s parent company.
In 2018, Google and DeepMind began the application of machine learning algorithms to generate 700 MW wind power. Further, the trials took place in the central United States. DeepMind’s Sims Witherspoon and Carl Elkin along with Will Fadrhonc from Google spoke about how it began in the United States. Also, they explained the working of the neural network and the way it was trained on historical turbine data and weather forecasts.
20% Increase in Value of DeepMind’s Wind Energy
DeepMind has developed a system that can predict 36 hours of wind power output. This way, one can predict the requirement of wind energy turbines for the necessary output. After looking into the production quantity, the model can recommend optimal hourly delivery to the power grid a day in advance. As a result, this model is highly valuable as it can deliver specific amount of electricity at a particular time.
Further, application of machine learning in wind farms has increased value of DeepMind’s wind energy by 20%. Moreover, these results are after comparing output with the baseline scenario of no time-bound commitments to the grid.
Using artificial intelligence in the renewable energy sector is likely to become crucial in the coming years. As wind and solar cannot promise constant stream of power, utilizing renewable energy is optimally is crucial.