AI can make wind power considerably more predictable and valuable and help wind farm operators make improved assessments on how to meet electricity demand

Google recently confirmed that it has used the artificial intelligence software of DeepMind, its London-based subsidiary, for making the energy produced by wind farms more viable. The IT giant claims that it is now able to schedule set deliveries of energy output, which are more valuable than standard and non-time-based deliveries to the grid.

Apparently, by the company has deployed DeepMind’s machine learning algorithms to forecast wind energy output from farms that are used by Google to support its renewable power initiatives. Google says this software has enhanced the value of wind power the farms produce by almost 20% over a baseline where no similar time-based projections are being obtained.

Google had informed last year that it had reached the goal of offsetting its energy use with 100% renewable sources. This was achieved primarily due to investments and energy purchase contracts with wind and solar farms that help to power its data centers, along with renewable energy certificates which offset standard usage of power grid in other markets.

Google’s Carbon Free Energy initiative lead, Will Fadrhonc and a project manager at DeepMind, Sims Witherspoon mentioned in a post that machine learning can be used to make wind energy sufficiently more valuable and predictable. They added that this approach also allows brining greater data rigor to wind farm operations, with machine learning helping farm operators to make faster, smarter and more data-driven assessments of how electricity demand can be met by their power output.

Reportedly, Google had announced in 2016 that it was able to cut power costs for its data centers by 15% owing to the help from the DeepMind’s AI. Google went further in 2018 by giving these AI systems greater control. Allegedly, DeepMind was also in discussions with the national electricity grid agency of U.K. in 2017 for helping it balance power supply and demand.