Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More

Around 81 million office workers spend about 11.6 hours per day in sedentary activities in the U.S. That’s not healthy for anyone, as it leads to more obesity, musculoskeletal disorders, and an increased risk of colon cancer. It’s also not easy to change, as things like desks, furniture, and buildings have been pretty dumb for a long time.

But Burcin Becerik-Gerber, associate professor of civil and environmental engineering at the University of Southern California Viterbi School of Engineering, is doing research on smart desks and smart buildings to try to remedy the situation. A smart desk could be as simple as one that sends you a pop-up message telling you to take a break after it senses you’ve been sitting in the same position for a while. Or it could tell you to switch from a sitting mode to a standing mode — and raise the desk height for you.

Becerik-Gerber and her graduate students are working with Arup, an industry collaborator and global corporation that specializes in engineering services. They have created prototypes with motors to raise the desk height, a Raspberry Pi computer to coordinate sensors and make suggestions, and a variety of sensors that measure how you‘re doing throughout the day. While this is a start, Becerik-Gerber says our desks and buildings are a long way from being smart. I interviewed her on a recent trip to USC.

Desks, rooms, and buildings should be able to do things like dynamically control the lighting and climate in a way that supports human circadian rhythms and reduces discomfort. But these adjustments should also be personalized for each individual, with machine learning to determine preferences for lighting and climate, and that’s why Becerik-Gerber is focusing on the desk first. Hopefully, the payoff will come in the form of a healthier and happier work force.

Here’s an edited transcript of our interview.

Above: Burcin Becerik-Gerber and her grad student Ashrant Aryal show off a smart desk at USC.

Image Credit: Dean Takahashi

VentureBeat: Can you tell me more about what you do?

Burcin Becerik-Gerber: I’m an associate professor in the civil and environmental engineering department. My research focuses on what I call human-building interactions, similar to human-computer interaction. I tend to see buildings as the machines. I try to understand how and why people interact with their buildings.

When I say “buildings” I use the world loosely. It could be your thermostat or your desk or your window or the space around you, all that. Why do people interact with these buildings? What triggers their interactions? Through these interactions, how does building behavior change, and how does that influence human behavior in turn? I also try to build new interaction modalities or new interaction modes.

VentureBeat: This is leading to things like smart desks and smart buildings?

Becerik-Gerber: Exactly. “Smart buildings” is a bit of an overused term. I try not to use it, because people always think about energy efficiency and sustainability. My interests are wider: emergency safety, security, health and well-being. Whatever the building’s purpose – in the workplace it’s productivity and satisfaction, at home it’s health and well-being – energy efficiency is kind of secondary. Depending on the purpose of a building and the goals of users, I try to achieve those goals through technology development.

One thing I have to mention is that this research is very user-focused. People’s happiness is very important for me. I always argue that you can make buildings energy-efficient just by taking everything out and making everyone in them miserable. That’s very energy-efficient. So that’s not my purpose. The center of the research for is users. How can I make people happier and more satisfied?

VentureBeat: I always think of game analogies. I talked with Will Wright a long time ago about how he designed The Sims, and he said that he decided to make the people dumb and the objects around them smart. With the objects beckoning to them, the people would only do things that made them happy.

Becerik-Gerber: Absolutely. People ask me why I started this line of research, because it’s different from what I was doing before. I was so miserable in my office. It was always cold. The lighting was horrible. I thought, “My God, I spend so much time in here, and it’s the same no matter who’s sitting here. It doesn’t care if it’s me or anyone else.” I started looking into how I could personalize these spaces. We’re personalizing everything now – our phones, our devices. Why don’t we personalize our spaces that we spend so much time in?

VentureBeat: So you have different kinds of tech prototypes that are illustrating this?

Becerik-Gerber: We’re working on many different things. One thing that was interesting to me, when you talk about personalization in an office or a residence or other buildings, how do you understand what people want? Thermal comfort, which is part of the desk, is one of them. How do you know when people are going to be comfortable or uncomfortable?

We had a progression of different prototypes. We started from smartphones, where people provided input, but that was too consuming. When people provide feedback they’re usually already uncomfortable. We want to catch them as they’re becoming uncomfortable, so we don’t make them uncomfortable. We had these glasses with infrared sensors where we tried to correlate the room regulation system response to people’s feelings of being uncomfortable, so it’s less intrusive.

Above: This smart desk prototype has a built-in Raspberry Pi computer.

Image Credit: Dean Takahashi

Later we moved into thermal cameras. That way you’re not wearing anything. It’s non-intrusive. We have some other wearables we can also show you. The whole idea is, how do you mathematically model people’s comfort preferences? Because we all have different comfort preferences. It’s a range. It’s not one number. It could be from 70 degrees to 75, or 70 to 85. It’s cultural. It’s gender-based. It depends on what you’ve been exposed to.

These centrally controlled buildings that we live in America aren’t very good for exercising your neutrality, because the more comfortable you get with a wider range, the healthier you are. A narrow range causes all kinds of cardiovascular problems, as well as type II diabetes. This is all recent research. We’re trying to get people from here to there through these desk negotiations.

VentureBeat: I’m thinking of the things that smart buildings are supposed to do, like turn on the lights for you or adjust the climate when you’re in the room. But if it’s not anticipating you, you’re uncomfortable right off the bat until it does something.

Becerik-Gerber: I know those sorts of buildings. They’re annoying to me, quite honestly. They turn off things when you don’t want them to, and you wind up waving your hands or jumping up and down. The blinds are constantly closing and opening in the background.

Our idea isn’t binary on and off. It’s like knowing your customer, basically. It’s difficult, because it’s not like I have one preference and it stays like that. From summer to winter, that changes. It’s also task-based. You have to understand what’s happening to people – if I’m stressed, if I’m sick. I’m looking into lighting and other things as well. Lighting should be different if I’m in a meeting or working at my computer. It’s not like I’ve just figured you out. I have to figure you out in your situation, in the context of what you’re doing and how you feel today.

As you say, it’s not just on and off. What do you want and how can I provide that? What’s your schedule? What would you rather have in morning or afternoon? That’s why I don’t like to call it “smart.”

Above: Hopefully future smart desks won’t have so many wires, but there’s a lot of tech and sensors built into this prototype.

Image Credit: Dean Takahashi

VentureBeat: Cost is a factor. There’s that joke about how the future isn’t evenly distributed. You look at something like Bill Gates’s house. Some people can afford things that are “smarter.” That’s something to overcome.

Becerik-Gerber: The first prototypes are always more expensive, but cost is something we have in mind. It also depends on the benefits. If adding $95 to my bill doesn’t create much of a benefit, that’s one thing, but if I’m healthier as a result, or more satisfied, or more productive, or I stay in the office longer getting more work done—some benefits are difficult to quantify.

VentureBeat: You were talking about posture and things like that. There’s something to consider in the work-life equation, the cost to the company if someone has–

Becerik-Gerber: Right, back issues, or absenteeism. Maybe someone is at work, but they’re not doing as well as they could be because they’re constantly bothered. The desk as we have it now attends to thermal comfort, visual comfort, and postural comfort. It has sit-stand, which isn’t new, but the idea is—this is the new part of it, not just the sensors or modeling of preferences or learning the user. Once it learns you and knows what you’re doing, there’s the ideal or the optimal or the healthiest regimen. How does the desk take you from your baseline to someplace healthier?

The whole negotiation between the desk, an everyday object—we’re trying to rethink it. There are 80 million office workers in the United States. Imagine how many desks there are. How does a regular object negotiate over time and co-evolve with its user to make the user healthier and more productive? This is the whole idea. It’s not just the sensing.

There are things I can reject outright. I don’t want to think about my desk going up and down as I write an important email. But some people want that. They want the desk to tell them what’s best. We’re not about one-size-fits-all. We’re trying to create these desk personalities that go along with the user’s personality.

Above: Ashrant Aryal shows off this smart desk, which senses how you are doing.

Image Credit: Dean Takahashi

VentureBeat: What do you need to measure or sense from the user?

Becerik-Gerber: How long you’re sitting. How long you’re standing. Too much standing can be as unhealthy as too much sitting. Your posture, if you’re hunching over. Desk and chair height are important. The distance of your monitor and your keyboard. All that sort of stuff.

VentureBeat: Something like an iPhone or an Apple Watch can’t do that, right? Does that mean there’s some kind of camera trained on you?

Becerik-Gerber: We don’t use cameras because of privacy issues. People hate cameras. We have distance sensors. We can add pressure sensors, although we don’t have those yet. We have height sensors. They’re simple sensors. They’re not super expensive or sophisticated. The idea is using multiple data points and pulling them together to make sense of everything.

VentureBeat: Does it all go into the desk, or is it set up elsewhere in the room?

Becerik-Gerber: We’re thinking around the desk and the chair, the work station. This is going to be the space. Some of the sensors are in the chair and some are in the desk. Everything is in the cloud. We have our algorithms running in the cloud.

VentureBeat: And you just start up an app and it starts working?

Becerik-Gerber: Right. We’re working on different ideas. We’ve put the desk together. It’s all CNC cut. We’ve installed the sensors. But we’re now thinking about whether it has to be in the desk itself. We might have a sort of sensor box or other addition, because nobody’s going to throw away their old desk to buy an entirely new one. We’re thinking about how to customize this to fit an existing desk.

Above: The components of a smart desk.

Image Credit: USC

VentureBeat: Do you need a power connection for both the desk and chair?

Becerik-Gerber: Right. But there’s usually power right around you, so that’s not too difficult. It’s another cord, but it’s not that difficult. They can also be battery-powered. The sensors are wireless. It doesn’t necessarily have to be connected to the wall.

VentureBeat: I think about things like VR. They’re in an interesting state now where they can do very interesting things, but they’re still cumbersome.

Becerik-Gerber: I use VR in my research as well. We can’t keep people in as long as a half hour with what we’re doing. The whole idea, though, you just don’t want to dread coming in and sitting in your chair because it’s going to do something to you or monitor you. We want to make this seamless. It shouldn’t interfere with your work or your daily life. It’s just there in the background, working for you. But it has to really be seamless.