This is an article from the fall 2018 issue of LSA MagazineRead more stories from the magazine.

Here’s one for you: What do you get when you put some scientists, laboratories, and equipment in a building together? If you’re a public university committed to research, you hope you get a world-changing discovery. And if you’re Jason Owen-Smith, professor of sociology, research professor, and executive director of the Institute for Research on Innovation and Science (IRIS), you get the chance to investigate ways to make such breakthroughs more likely.

Every discovery is unique, but there’s a general path to aha that feels familiar. There are also moments we don’t normally recognize as part of the process of hypothesis and experimentation. Owen-Smith, working with a cross-disciplinary team, is trying to quantify these moments, too.

“On the one hand, there are the apocryphal stories about innovations that happen because a physicist and a biologist both have kids on the same softball team and spend a lot of Saturdays in the stands,” Owen-Smith says. “On the other hand, there are formal case studies that say something like, for example, Steve Jobs put the bathrooms in the center of the building at Pixar. Pixar is awesome. If we put the bathrooms in the center of our building, then we’ll be awesome, too. That’s the thinking. But of course we don’t actually know that.”

Above: In each of the four examples pictured here, a hallway and two offices occupy the same amount of physical space, but they don’t yield the same opportunities to collaborate. The likelihood of collaboration increases with the amount of time people spend in a space together. When there’s a stairwell at each end of the hall, as in example A, people are more likely to simply use the stairs near their office rather than walk down the hall, so there’s very little chance for overlap. In example B, people do use the same stairwell, but their paths separate as soon as they exit into the hall. Example C creates slightly more overlap because people must briefly walk together to get to either office. In example D, the overlap area is increased because both offices are farther from the stairs.

Architects and spatial scientists have spun skeins of theories about ways to encourage innovation, but the theories tend to take place in abstract space. People interact in actual, physical spaces, Owen-Smith says, and they too develop real patterns there.

“If you watch how people move through their houses,” Owen-Smith says, “there are spaces that are rarely entered and spaces that are used a ton. We move from known point to known point without venturing into other spaces too much.”

Owen-Smith and his colleagues were curious to see if these patterns apply at work, too. He found we mostly go to certain spaces — our desks, the water cooler, the cafeteria, the bathroom — and travel the same paths between them. Across industries, he found employees’ movements were easily understandable in a certain space while others were harder to decipher.

“We radio-tagged a group of corporate engineers and followed their walking paths in real time,” Owen-Smith recalls. “They kept congregating in one corner of the building and we couldn’t understand why until we sent a graduate student there to observe. It turns out the woman whose cubicle was there tended to bring brownies.”

In their biggest study of this kind, Owen-Smith’s team used data on office and lab spaces and the floor plan of bioscience buildings to estimate the logical walking paths they thought biomedical scientists in the building would take. “We marked where their offices were, where their labs were, and the nearest relevant bathroom,” Owen-Smith says. “We marked the stairways and elevators and then we drew the shortest walking path through the building among all of those spaces. Given what we understood about people’s habits and movements, the scientists were probably at one of those points or on one of those paths most of the time.”

Once they marked these paths, they identified areas of overlap between two researchers who followed the same paths every day: You walk around, you wait for the elevator. You start to recognize the people around you. You ask, what do you do? You get that nodding familiarity.

“We found that for every additional 100 feet of overlap in walking paths, there is about a 17 percent increase in the chance they will start a new project together,” Owen-Smith explains. The same amount of overlap also increased the collaboration’s chance at funding by almost 20 percent.

“The work of science is hard, and it requires a lot of back and forth,” Owen-Smith says. “If you’re bumping into people routinely, they can say, ‘Hey can I grab you for a second? I really need to troubleshoot this assay.’ That’s much easier and more efficient than sending an email to ask to schedule half an hour to come to the lab.”

It’s hard to teach people to recognize a breakthrough waiting to happen, Owen-Smith says, and it usually needs to happen face to face.

“Sometimes the search is directed. You know you need someone who’s really good at graph theory to help with a particular mathematical network problem and you find them. It’s much harder when you don’t know what you need. That’s where cultivating the chance for serendipity comes in.”



Illustration by Julia Lubas