A talent pool for the 21st Century: The AI team in the Data Center

Early in June, we went public with plans to expand our Montreal facility by adding two new data halls and an additional 6MW of capacity. Agreements are already signed for two-thirds of available space, reflecting ongoing demand from hyperscalers, high-tech and enterprise customers that are choosing Montreal. The build will be complete by the end of the month, and we’ll put another stake in the ground in ROOT’s growth story.

Our expertise is in fast and reliable expansion and scalability. Since the beginning, we’ve been able to build infrastructure and systems quickly and efficiently while at the same time, scaling our talent pool. This year as we grow, we’re scaling a different kind of expertise, using artificial Intelligence to train a whole new cohort of “co-workers” to support our established team.

In every industry, Artificial Intelligence is in beta testing. Applying it to improve data center operations can look different at every company. Our partner, Litbit, says that most organizations start using their AI by placing “pulse point stickers” – essentially QR codes – on the parts they want to track and then use a cell phone to take its “pulse” by analyzing sounds and vibrations. The technician tells the machine that this is how this particular part sounds with a normal load, and the next time it takes its pulse, it will be able to identify for the human, sounds that are abnormal, triggering an alert. Amazing, right?

At other companies, microphones can be permanently mounted on devices, allowing constant monitoring of machinery. This is how we’re using it in our initial test at ROOT.

With the volume of generators at our growing facility, it would be impossible for technicians to stand beside each one of them listening 24/7 for any noise indicating a potential problem. Instead, we have microphones using AI in each one. They listen for odd sounds. It could be a low fuel level, a bearing rubbing, or a knocking that’s out of place in the normal operation of a generator. In any case, it alerts our team and we’re able to address the issue, deterring what undetected, could become a bigger challenge.

So it’s preventative. And interestingly, it’s cumulative and therefore prescriptive. Traditional monitoring tools largely fail in transferring knowledge, but AI takes the data and feeds it into the system so that there’s a continuous integration of learning that’s accessible to everyone over time. For example, the first time a machine hears a failing motor bearing, it won’t know what it is. But it’ll know the second and every time after that, not just how to identify it but – with the accumulated learning of previous working solutions – how to fix it.

This is healthy for our industry as a whole. Best operational practices have long been established and now exponential growth is demanding innovation. For ROOT in particular, AI is part of a smart growth strategy. Highly skilled professionals are how we are able to consistently deliver on a commitment to 100 per cent uptime, and anything that strengthens our team only makes us stronger as a company.

To learn about more about our testing of AI at ROOT, click Artificial Intelligence in the Data Center: ROOT Data Center First to Use Machine Learning to Maintain 100 Per Cent Reliability.