This audio was created using Microsoft Azure Speech Services
If you walk around the white space in a typical data center, where all the IT gear lives, you’ll more than likely see very few people. In many edge data centers, there are no IT people at all, a so-called “lights-out” environment. That’s because over the years we’ve become adept at automating IT management, and doing so remotely, such that fewer people can manage more systems without sacrificing availability or performance with data center operations.
We’re heading in the same direction for the data center grey space, which houses all the supporting infrastructure, such as chillers, rooftop cooling units, switchgear, transformers, breaker panels and the like. Just as with IT systems, the day is coming when more of this equipment will be managed remotely, requiring fewer staff on site. Call it the dimming of the data center lights.
Improve efficiency and lower costs in data center operations
Dimming the lights is an important concept because it promises to deliver more efficient data center operations at lower cost. The idea is to use technology and better processes to streamline operations and improve reliability, while requiring fewer human beings on site.
Getting there starts with instrumenting and connecting more data center infrastructure. It amounts to applying Internet of Things technology to all the gear in the data center grey space, such that each component can report on its status and health. The data goes to a remote service bureau, where artificial intelligence can be applied to detect anomalies that suggest a problem may be brewing. Service bureau staff can then direct those on-site on how to address the issue.
Improve data center operations with big data and artificial intelligence
This is by no means some futuristic pipe dream. Schneider Electric is using it now to collect data from numerous customer data centers and deposit it in a data lake, to which we apply AI applications that identify the tell-tale signs that a component may be in danger of failing. Essentially, it’s the concept of predictive maintenance, which is far more effective and less costly than traditional schedule-based maintenance.
The gating factor in dimming the lights is instrumenting and connecting all those systems. While newer gear will likely support instrumentation, much of the existing medium- and low-voltage systems don’t. It’ll take some time to get it all upgraded and connected.
And for the immediate future, at least, we’ll still need some staff on site in our data centers. Someone has to perform repairs the service bureau identifies and it’s also useful to have people looking for issues that can’t yet be identified remotely, such as an odd smell, or to escort visiting vendors on site, for example.
That said, I can envision a day when we have robots and improved sensors that can take on some of those jobs, enabling us to dim the lights even further.
EcoStruxure IT leads the charge in optimizing data centers
In the meantime, you can get started in your data center by employing EcoStruxure IT, the cloud-based Schneider Electric architecture that delivers intelligence on your data center. It’s how Schneider Electric is leading the charge in terms of dimming the lights in data centers, including those for customers with mission critical facilities, all the while improving data center operations and management.