Designing buildings that learn

By | 5 minute read | November 1, 2017

leadspace image for wired article on data and analytics

Our buildings are snapshots of our ideas and culture, physical representations frozen in place and time. The Empire State Building, the Roman Forum or even the Los Angeles Forum, were designed and built to serve a purpose, a population or a team of the moment. But things change. Economies shift and empires fall—yet our buildings don’t make those transitions. We tear them down because they can’t adjust to new energy requirements, an influx of people or new ways of working. At least they couldn’t until now.

Homes and offices will get better at whatever we need them to do

Buildings, even beautiful ones, are brainless—but the Internet of Things coupled with machine-learning systems like Watson promise to change all that. Our buildings are gaining cognitive abilities that will allow them to learn to be better at whatever it is they are designed to do.

So what kind of brain might your home or office building soon possess? Instead of nerves, there are myriad sensors deployed in walls, lights, heating and cooling equipment, even the faucet in a sink. Creating a structure’s consciousness is software that ties it all together. But these brains, like any brain really, don’t just blink on and immediately understand how to function in the world. Like any growing person, the “brains” in our roads, office buildings, homes and factories will need to be taught how to think, how to problem solve, how to arrive at smart decisions.

The building’s digital twin

Some buildings today are built from the ground up with nearly one IoT-enabled sensor per square foot—monitoring temperature, humidity, the weight in the trash cans, how many people are in a room, and on and on. Managers can then “see” the building on a computer, in what’s known as a “digital twin.” It’s that digital twin that, like any thinking being, will build a model of the world it inhabits and find ways to adapt and adjust to whatever the priority or emergency that is thrown at it.

“Buildings will know the context of how space is being used and predict the kind of load that will be required—is it a Friday when many people work from home, does tomorrow’s weather mean there will be extra heating or cooling required?” says IBM Research’s Dr. Joern Ploennigs. High-powered systems such as Watson weave that data together to make buildings work better for occupants and to find ways to trim costs.

That’s a far trickier challenge than it might seem on the surface, but one easily handled by the cognitive abilities of the “digital twin.” For instance, a floor in the average New York 80-story tower costs $6 million per year. Imagine your business owns or rents 10 of them. Your real estate team might think every floor is maxed out, but the building—based on data from tracking people flow—knows better.

“If you talk with most real estate teams, they will tell you they absolutely need all 10 of those floors,” says John Smart, IoT and Smarter Workplace Strategist at IBM. “We’ve seen space utilization at companies who think they are maxing out their real estate running at 38 percent. They don’t always believe it at first, but then the data sinks in.”

Serving employees more efficiently

As Eva Fors, CEO of sensor maker Yanzi, explains, a business might have a conference room that’s occupied 95 percent of the time during working hours. With usage that high, it’s very likely that people are being turned away when they need a room to meet. The conference room may hold 12 people, but the sensors say the average meeting has just three. Cognitive analytics will suggest installing a wall so that the room can serve more employees more efficiently.

The savings can get as granular as desks. People typically work at the desk or cubicle each day. But sometimes people need quiet, to read or really think through a problem. They need a customized workday. Deloitte’s Edge building in Holland analyzes an employee’s calendar and reserves the needed desks accordingly. Someone may start with a collaboration space in the morning to work through the latest client presentation; then have the board room reserved for that big lunch all-hands; and then shift to a quiet reading space to catch up on the new report for a second client. With so much of our work documentation stored digitally, Fors says the idea of an assigned, personal desk will soon be a thing of the past: “You take a seat based on what you’re actually going to do.”

Consuming the work environment a la carte can also go for tenants. Businesses often pay for bathroom cleaning, but what if no one’s using the one on the fourth floor, yet it gets cleaned every day? Managers only want to pay for the resources they’re receiving. The future powered by IoT offers a model of renting based on consumption. We see this today in the rise of co-working spaces. Similarly, if an entire company has a sense of its exact need for physical space, they can negotiate with the landlord knowing precisely what they require, rather than paying for unused facilities. That changes the landscape of the entire lease management market.

No more guesswork

The savings move up the supply chain as building managers optimize their own operations. Autodesk senior business line manager Brook Potter’s team designs the modeling software for building managers. With the insight the data provides, Potter says building managers know how to tune their teams and vendors to perfectly suit the building’s needs. For example, sensors alert the building manager when something’s going to break before it goes. Metrics now show that half the repair staff sits idle each day, so now managers can see that they need more of one type of employee versus another. With a full picture of what’s going on inside their properties, owners don’t rely on guesswork and history to make decisions, but real, actionable data.

Managers can even analyze power usage across their portfolio properties’ assets—lighting, security, HVAC systems, among others. Not only can they fine-tune energy consumption across all their real estate, they can also sell that information back to the supplier, so the supplier can do analysis and improvements on their own products.

No doubt there is a degree of trust when listening to the data—especially for managers used to relying on gut feelings. Sensors may deliver information that feels counter-intuitive, like that a bathroom on the fourth floor could be closed and no one would notice. But as Fors points out, it will take certain courage from managers and tenants to follow through on the analysis. Like most innovations in technology, though, it will be the cost and time savings from the bold users that catalyze the change for the future.

Learn more about what the IoT can do for your buildings.