How AI Streamlined The Novus — Durham’s Newest Residential Complex

Durham, North Carolina, is home to Duke University and a thriving medical job market, and it is part of the state’s Research Triangle area, including Raleigh and Chapel Hill. As the city grows and attracts more diverse residents in the tech and health care sectors, they need places to live, work and shop. That’s where The Novus comes in. Designed by the award-winning architect firm LS3P, the 670,000-square-foot, 27-story mixed-use development features 54 for-sale condominiums, 188 rental apartments and five levels of parking.

The Novus, which also features a 25-meter swimming pool, pickleball court and putting green, is a partnership between Durham-based Austin Lawrence Partners and Global Holdings Management Group, seeking to develop and improve the Durham area. Samet Corporation, which has partnered with Global Holdings on other projects, is overseeing the construction.

How AI Helped Build the Novus

Samet has partnered with the artificial intelligence (AI) construction company Buildots to help streamline construction of The Novus. Buildots uses videos of the site to spot potential problems, such as a misplaced column or plumbing pipe, and it has tracking software to keep projects on track.

“It keeps people accountable,” said Chris Hargrove, construction manager at Global Holdings Management Group. “It’s a reporting tool to see how progress is being met, and we look at it together. We look at it and see where a certain subcontractor could be improving [and if] subcontractors are going ahead or falling behind.”

Hargrove said the video aspect is valuable for stakeholders who aren’t in North Carolina, as they can see progress happening on a regular basis without having to travel. “It’s exceptionally valuable information as you’re building the project, and we [may have] made a change, we have to move this bathroom over and, ‘Oh, my God, what’s in the walls?’” he said. “And even though the drawings show [what should be there], the photos and the photo information, and the history of it shows exactly what’s behind those walls or above the ceilings.”

Clayton Simmons, senior project manager at Samet, said that Buildots helps facilitate conversations with structural engineers when a change needs to be made to a completed section of the building. The previous scans show clear areas behind concrete or drywall to add new features without damaging the structure.

How Tech Can Aid Collaboration

The software also helps with cost and design decisions, Simmons and Hargrove said. For example, if a condo buyer wants to move the kitchen island or the master shower, the design team can go into Buildots to produce a more accurate design and let the condo buyer know if the change will affect their final costs.

“The buyer can make a decision on if they want to make the change or not, versus getting halfway into the change and going, ‘Oh wait, there’s a lot more stuff here than we thought,’” Simmons said.

Aviv Leibovici, co-founder and chief people officer at Buildots, said a major advantage of the software is that it acts as a neutral third party when a project gets off track.

“It allows people to be a lot more collaborative because it takes away the need to fear transparency,” he said. “It’s the thing that is saying something is wrong. It’s not anyone versus anyone.”

Another advantage of Buildots, according to Leibovici, is tracking performance metrics in a way that helps construction companies understand the root cause of a delay.

“[Buildots] is a tool that measures not where you are against your schedule, but how you got there,” he said. “You might be two weeks delayed on installation, but what’s really going on is that you’re losing a half day every week. So, you can see what that would look like over the next 10 months of work and maybe get back on track.”

The data Buildots has gathered from 200 projects around the world can now be applied in the planning phase of a project, pointing to potential risks and delays before the golden shovels come out to break ground, Leibovici said.

“You might want to do a few things, change the schedule and gear up for the problem before you reach it,” he said. “We’ve had projects that when we point out the potential risks, they end up being more efficient.”

Simmons said Buildots has helped with exactly this issue on 3D models of the Novus.

“We’ve been doing 3D models for a while, but it’s kind of up to everybody in the field to check the model and discover discrepancies, which doesn’t always happen as effectively as you’d like,” he said. “This has allowed us to catch a lot of issues early on, when it’s that big of an issue to fix. It catches items that aren’t following the model, and we can resolve those things right away before we get too far down the line.”

Buildots’ AI capabilities also help keep The Novus on track when materials get delayed in shipping, Hargrove said.

“There are a lot of parts to a project like this, and things happen,” he said. “Buildots can say something is delayed, but we can look at the project as a whole and see what we can do to get ahead of it. Samet would look at the schedule, then enter the revised schedule into Buildots.”

Keeping the Human Touch

Artificial intelligence has grown in recent years, and with it, fears that technology will take away people’s jobs. Hargrove said that hasn’t been the case for Global Holdings or Samet.

“You need people to run Buildots,” he said. “You’re not taking jobs away from the industry, you’re just shifting jobs, and you’re shifting responsibilities, and you’re using technology to do it. You may say, ‘OK, we don’t need five supers in the field. Maybe we need three.’ But you also need people to record the information, download the information, analyze the information and generate reports. It’s not a job eliminator; it’s just shifting and allocating our personnel into different positions.”

Hargrove added that Buildots relies on human input to track the project. If the software isn’t fed good data, it can’t output good predictions, he said. “Buildots forces you keep the information updated, which leads to better and more efficient outcomes.

Simmons agreed, adding that Buildots is making people more efficient on The Novus build site. “Before, you might have [had] to go out and take a bunch of photos on your phone, come back and create your punch list and assign it to everyone” he said. “This makes the list. You still have to spend time walking around with a camera, but Buildots compiles the list for you. Instead of taking 12 hours to do the thing, it maybe only takes six hours, and now they have time to get other things done.”

Leibovici sees Buildots less as a competitor to human power and more as a sidekick to the people doing the planning. He sees a future where people won’t plan or make changes without consulting an AI system for advice.

“I think it will happen over time; it cannot happen in a day no matter what,” he said. “I think it will happen over the next five years. In between, there will be companies that start doing this earlier, there will be those that will do it a bit later. That is true for any innovation in the world.”