Ex Machina: Machine Learning in the Next Gen Data Center

From day one, Root Data Center has been pushing the boundaries of technology and innovation. We’ve created design and construction methodologies that let us deploy Tier 3 level data centres at a lower cost than our competitors. We leverage Kyoto Cooling. And we’ve developed a methodology to deploy large-scale data centre services in under 120 days, when it can take our competitors up to a year to build the same capacity.

Being innovative is vital in an industry growing quickly and exponentially. From mainframes, then microcomputers, to data center sites acres wide that are the spine of the digital economy, it’s fair to say that the workings of the modern data center have become too complicated for even the most educated humans to manage alone with manuals and spreadsheets. We need a new way forward.

The dream looks like this: the data center of the future is anchored by Artificial Intelligence. It monitors and controls the IT and facility infrastructure, as well as applications, cooling, power, workloads, storage, and networking, adjusting everything required as necessary in real-time and seamlessly to achieve maximum efficiency, productivity, availability and reliability. AI troubleshoots, augmenting maintenance and preventative maintenance. It learns from its experience. And it accomplishes all of this across multiple sites.

While we might be decade or more away from this kind of depth and complexity, the data center industry is making progress toward achieving parts of that dream. It’s no surprise that ROOT is at the forefront of innovating AI in the data center.

Work to-date has been done with AI in the field of power optimization, but we’re focussed on what we believe to be the most critical issue in our business: reducing the risk of downtime. According to IDC, the average hourly cost of an infrastructure failure is US$100,000. In a critical application failure, that number can get to $1 million. Clearly, 100 per cent uptime is crucial.

ROOT is partnering with Litbit, a San Jose, California-based start-up founded and led by Scott Noteboom. He was formerly Apple’s head of infrastructure strategy, design and development, and before that, spent almost seven years as Yahoo’s VP of data center engineering and operations.

Litbit’s AI solutions are aimed at industrial facilities management more broadly, but the data center industry is a priority for them. We’re proud to be their first partner to use machine learning to maintain 100 per cent reliability. This year, we’re reinforcing our core maintenance program by testing Litbit’s system on 14 emergency diesel generators at our MTL-R2 facility. Odd sounds like grinding, dripping or knocking can indicate any number of reasons to examine and diagnose. Having AI inside every generator is the machine equivalent of having skilled engineers standing beside each piece of equipment 24/7. To learn more about the project and how the AI actually “listens,” click Artificial Intelligence in the Data Center: ROOT Data Center First to Use Machine Learning to Maintain 100 Per Cent Reliability, and stay tuned to this space for an update on test results this year.