Spurred by experiments in remote site management during the pandemic, Manos has spearheaded an effort to use Boston Dynamics robots equipped by Field AI to 3D-scan jobsites, and feed that information through AI algorithms to assess work progress. “We’ve piped some of that data into a new column in our scheduling applications,” he said. “The superintendent has a column next to their reported progress with what the robot says the progress is, so we’re excited.” As with so many of the new approaches, Manos said the challenges are less technical and more about finding the right use cases.

Sometimes AI can show up in unexpected places. “What we do is we take in massive amounts of data. We have a large data set already we can use to model the performance of concrete [and] we can actually start to predict the performance of a brand new concrete design based on the raw material makeup,” said Ryan Henkensiefken, vice president of business development at Concrete.ai. “This is what AI is really, really good for—it runs hundreds of thousands of different formulations and calculations to find that unique set of raw materials that go together to meet all requirements for the project.” 

Concrete.ai’s system works with ready-mix suppliers to optimize concrete mixes, including those in pursuit of lower carbon. That has paid dividends for sustainability goals of Compass Datacenters, said Nancy Novak, its chief innovation officer. “We were able to nominally reduce the cement content by close to 20% in our mixes, which made a huge difference for our carbon footprint, since one 220,000-sq-ft facility is 17,000 cubic yds of cast-in-place concrete, and we can have 10 of these facilities on one campus,” she said.

Oprizon CEO Hiroto Sato (right) explains the challenges of construction tech investment.
Photo by Jeff Rubenstone/ENR

State of the Startups

Having a good idea is only the first step to bringing it to market, Kaustuhb Pandya, partner at Brick & Mortar Ventures, said during a construction tech investment panel. “Picking the right horse is not necessarily the right answer,” he said, adding that a successful pilot of a startup firm’s technology has to be followed by the hard choice of whether it’s a good fit for a firm at scale.

Hiroto Sato, CEO of Oprizon, a technology joint venture of Obayashi and Hitachi Solutions, said it is hard for tech companies to understand construction’s needs due to the insular nature of the industry and the figurative “temporary wall” industry firms erect around their work while it’s going on. “During the construction period, nobody knows what we are doing except for construction engineers,” he said. “So that [creates] this distance between the Silicon Valley tech companies and the construction industry.” Sato added that sometimes the most fruitful way to cultivate startups’ interest is to show where the pain points actually are located in the process. Pandya offered a similar take on what he looks for in startups. “I don’t think we care about the business model as much as we care about ‘is this a solution that’s going to solve a problem that the industry is looking to solve?’” he said.

Unleash the Drones

Drones helped improve accuracy and safety in the placement of a 2,200-piece prefabricated timber roof for Portland, Ore.-based Portland International Airport’s new terminal, explained presenters from the project's Skanska team. The 9-acre, 9-million-ton curved timber roof needed to be placed within a 3/8-in. tolerance.  

A drone inspected the construction of the elevated curtainwall, parapet and expansion joints and the curtain wall at up to 90 ft, said Brooke Gemmell, Skanska emerging technology manager. “We want to be able to get the drone so that we can see that really well without having to put someone up in a lift.” 

The team used an app called Luma that takes imported photos and videos and creates a low-fidelity model. Gemmell added that co-presenter Larry Curran, Skanska superintendent, “was able to go out and fly this curtainwall mock up, and then ingest that into Luma. And it created 3D models that we could send to the designers so they could really easily analyze this. We can rapidly make this 3D model and have them be able to ask questions and just have a better understanding of what's going on without needing to be on site.” 

Drones provided even more benefits for the airport, Gemmell added. “One thing that the port does twice a year is a runway inspection. So they'll actually have someone manually walk the runway and take photos. And it's really timing-intensive. And it results in the runway being closed down for up to a week at a time. We were able to schedule a [drone] flight that took less than a day and capture video and photos across the entire runway that we can then share with the team. So they didn't have that runway closure.” 

Curran added that drones can be a valuable tool for emergency response teams. “We actually talked with the local fire department … they came out and visited our site to get an understanding of where potential hazards are. We set up what is called a ‘high angle rescue practice.’ Because this roof is up 90 ft in the air, they needed to understand how they were going to be able to get an injured individual off the roof structure. I said ‘hey, why don't we fly the drone, I'll be standing next to the incident commander. And he can be looking at my screen. And I can show him exactly where his people are, where the injured individual is, or any obstacles that may be in the way of them trying to get where they need to be. Once I showed them this technology, this drone, they started a whole other conversation of ‘maybe we can use this sort of technology for actual accidents like a plane crash.’”