Smart building technology will help enterprises to use their resources and data more efficiently. Fittingly, the same principles can be applied to Intel® Rack Scale Design in datacenters.  

I’m usually the last person to leave at the end of the day. At one particular job, my office was far away from the elevator lobby. Every night, a security guard would step off the elevator, look around, see no one, and flip off the lights for the entire floor.

“I’m still here!” I would shout. If he heard me, he turned the lights back on -- for the entire floor.

The need for smart building solutions

So there was a resource problem: Was it better to keep one guy working all night, but wasting electricity for the whole floor? Or was it better to save electricity, but accept that not as much work would get done? Leaving the lights on, but only in my office, was not a realistic option. Sure, someone could go around flipping all the other lights off, but that process would be a waste of time in itself, and it wouldn't scale.

Once, at the same company, I walked into a room that was the workspace for four people. All of them were dressed head to toe as if they were going skiing: scarves, wool knit caps, and puffy jackets. The room was indeed freezing. Also, it was July. The employees had complained, but the whole floor was on one thermostat, and that thermostat was far away from that room, so the complaint wasn’t taken seriously. They had given up.

The downside of non-automated buildings

The problem behind these absurd (but real) situations is that buildings are not usually designed in such a way that their utilities (electricity, air, water, and so on) can be managed at a micro level. To do that, the traditional thinking goes, would be to give employees more control of their environment than would be reasonable. Lights would still get left on. Heat and air conditioning would still be wasted. But the waste would happen in a haphazard way that would be impossible to track or control.

Recently, a third option has emerged: smart buildings. Usage of services - and the apparent need for those services - is tracked, recorded, and analyzed. Check out the video below called "Harnessing the Power of IT with OT" from Christoffer Ramm at the Intel Developer Forum 2016 to see how a new Ericsson building is wired in such a way that all the problems I described above could be easily avoided.



Smart building, smart workers

Maybe the lights and the air conditioning only need to stay on at night where people are still working. Or maybe the scheduling gets more regimented, and everyone has to leave at the same time. That’s up to management, but this way, management has the tools they need to make the right decision.

My friends in the cold room would also fare better in a smart building. The temperature of each room could be tracked, and that data would be juxtaposed with data on where the employees were sitting. Of course, it’s technically possible to do that now, in a “dumb” building, if you have someone constantly walking around, checking thermostats, noting which rooms are empty at which times of day. But it’s not reasonable, because that type of manual effort doesn’t scale. Even if you do like that solution (maybe you want to give your cousin a job), that person can’t be everywhere at once. But a smart building is gathering and analyzing data all the time.

Smart building design and Intel® Rack Scale Design

Interestingly, there is both a functional and an ideological relationship between smart buildings and Intel® Rack Scale Design (RSD), the model of software-defined infrastructure that informs various Ericsson projects, such as the Hyperscale Datacenter System (HDS) 8000.

In a traditional datacenter, each server has the exclusive use of the components associated with it, such as hard drives and memory modules. So if one computer is engaged in a particular process, every part of that computer is considered unavailable for other work, whether that’s actually true or not. This is a bit like the floor of my office building in which all the lights were either on or off. One of the resources (me) might still be available to work, but that doesn’t matter, because that resource is required to function as part of that floor, even though my ability to work actually should have had nothing to do with my physical location.

The smart-building scenario in which I get to stay in my office, with the air conditioner and the lights on, even though the rest of the floor is dark and arid, is more like RSD. RSD decouples components from motherboards, instead placing resources into pools, from which they can connect to any process in the datacenter that might need them. So the lonely hard drive, like me, can work on into the night, even if the other hard drives next to it are still and quiet.

To explore some of our ideas on how your datacenter must be embraced as the digital foundation of business in the 21st century, please download our e-book: Transform your datacenter into your competitive edge:

Download the eBook

Cloud Infrastructure Partnerships Data & Analytics

Michael Bennett Cohn

Michael Bennett Cohn was head of digital product and revenue operations at Condé Nast, where he created the company's first dynamic system for digital audience cross-pollination. At a traditional boutique ad agency, he founded and ran the digital media buying team, during which time he planned and executed the digital ad campaign that launched the first Amazon Kindle. At Federated Media, where he was the first head of east coast operations, he developed and managed conversational marketing campaigns for top clients including Dell, American Express, and Kraft. He also has a master's degree in cinema-television from the University of Southern California. He lives in Brooklyn.

Michael Bennett Cohn