Google announced today that it has made energy produced by wind farms more viable using the artificial intelligence software of its London-based subsidiary DeepMind. By using DeepMind’s machine learning algorithms to predict the wind output from the farms Google uses for its green energy initiatives, the company says it can now schedule set deliveries of energy output, which are more valuable to the grid than standard, non-time-based deliveries.
According to Google, this software has improved the “value” of the wind energy these farms are providing by 20 percent over a baseline where no such time-based predictions are being performed. We don’t know exactly what that value is in monetary terms or in terms of energy output. We also don’t know where exactly this is being deployed, although Google works with wind farms largely in the Midwest, where some of its US data centers are located. Google was not immediately available for comment.
Last year, Google said it had finally reached the milestone of offsetting its energy usage with 100 percent renewable sources. That’s largely thanks to energy purchase contracts and investments with solar and wind farms that help power its data centers, as well as with renewable energy certificates that offset standard power grid usage in other markets.
When it comes to wind power, however, making use of that energy can be difficult because knowing how much a given farm will generate and how best to store and then deliver that energy to the grid changes every day. Google says “the variable nature of wind itself makes it an unpredictable energy source — less useful than one that can reliably deliver power at a set time,” due to having to rely on nature to generate the needed electricity demands of the grid.
“We can’t eliminate the variability of the wind, but our early results suggest that we can use machine learning to make wind power sufficiently more predictable and valuable,” write Sims Witherspoon, a product manager at DeepMind, and Will Fadrhonc, Google’s Carbon Free Energy program lead, in a co-authored blog post. “This approach also helps bring greater data rigor to wind farm operations, as machine learning can help wind farm operators make smarter, faster and more data-driven assessments of how their power output can meet electricity demand.”
This isn’t the first time DeepMind’s AI expertise has been used in this way. Back in 2016, Google announced that it had cut the power costs of its data centers by 15 percent thanks to the AI lab’s help. In 2018, Google went further and gave these AI systems even more control. And there were reports in 2017 that DeepMind was in talks with the UK’s national electricity grid agency to help it balance supply and demand.
This sort of work helps Google in an obvious way, but it also helps DeepMind. The company has done phenomenal work from a research perspective, but has yet to find substantial revenue streams. It loses a lot of money ($368 million in 2017), which has reportedly contributed to tensions between DeepMind and the mothership. If the company’s software can be put to use in real-life scenarios outside the research lab, DeepMind could become a revenue-generating segment of the business that justifies its high costs.