• United States

Buildings should behave like humans

News Analysis
Jan 18, 20183 mins
Data CenterInternet of Things

Artificial intelligence, along with real-time IoT, must be incorporated into buildings to make them smarter and healthier, says a major construction firm.

One must be able to walk into a room, including those in data centers, and not only access information about every facet of it, but also importantly, have it automatically solve all of its problems on its own.

Site 1001, which specializes in artificial intelligence-run facilities management systems, says the problem should be achieved through neural networks that copy how humans and animals think.

The company, a spin-off of JE Dunn Construction Co., demonstrated its all-listening, predictive building maintenance at CES 2018 last week. It says its big data, AI-driven system will ultimately produce smarter and healthier buildings.

One example demonstrated in mock-up form at the conference was an apparatus that could automatically determine when bacteria conditions in water systems were ripe for producing Legionnaires’ Disease. The IoT-driven infrastructure autonomously flushes the building’s plumbing system with pathogen-killing hot water, reporting back to the facilities manager that the relatively simple-to-perform — but hard to know when to do — job had been accomplished. It uses heat data from sensors to figure out when pathogens might form.

In traditional systems, water samples would have to be taken manually. And if they weren’t, then reports of illness in building occupants would be the impetus for corrective measures.

Buildings as living structures

“When we look at buildings as living structures, we can understand how various systems are connected and operate together,” says Dr. Filip Ponulak, principal data scientist at Site 1001, in a press release.

Ponulak says all buildings should now be listening for issues. He says it’s an innovative way of managing new buildings. Site 1001 believes its system would also work in older buildings.

Chief Innovation Officer Eric Hall told me one could draw an analogy with an aging car, except that unlike cars, buildings don’t have odometers to help identify failing parts.

In other words, by collecting data on failings, for example, predicting upkeep becomes possible — you know when things are likely to fail and can pre-empt them, like a flexible car service schedule. That lets facilities management “move to an entirely conditional and proactive maintenance schedule,” says Hall on the company’s website.

Data centers fit into this platform, too, the company says. Indeed, I’ve written before about folks who think AI will ultimately self-manage the data center:

Robots are going to be running the show, I wrote in a recent blog post. In that case, robots are making physical cabling connections. But a facet of that article was a company that said it would be cloning employees’ intimate knowledge of their workspace soon.

In addition, HPE says predictive troubleshooting from AI will be used to unravel bottlenecks between applications and data within the data center.

“A building can be compared easily to our own body and working systems,” Ponulak told Builtworlds in an interview this month. He says HVAC is like the respiratory system, the electrics are like the “body’s circulatory system, which supplies energy to the body,” and the musculoskeletal is the building structure. Communication with the original architectural plans, live sensor data, and AI create the brain.

It’s the coordination of IoT that gives “the building the ability to heal itself,” Site 1001 says.


Patrick Nelson was editor and publisher of the music industry trade publication Producer Report and has written for a number of technology blogs. Nelson wrote the cult-classic novel Sprawlism.

The opinions expressed in this blog are those of Patrick Nelson and do not necessarily represent those of IDG Communications, Inc., its parent, subsidiary or affiliated companies.