Bedrock Robotics' technology—in a partnership with Sundt Construction—is being used to automate excavators' mass earth removal for site preparation of a 130-acre Southwest U.S. manufacturing facility supporting the domestic energy market.

Bedrock's autonomous systems have moved more than 65,000 cu yd of earth and rock by loading human-operated articulated rock trucks with the same workflow used in manual operations—trucks stopping to be loaded by an excavator taking scoops from a stripped pile. The integration demonstrates commercial viability for autonomous construction, Bedrock says. The technology is now installed across multiple excavator models ranging from 20- to 80-ton machines on the large site, adapting its systems from compact units to large-scale earthmovers.

"They're a tool in a much larger toolbox that we have on this," said Dan Green, Sundt senior project manager, noting plans to move about 700,000 cubic yards of rock and earth on the project. "That's the scale of what we're moving," he said. Bedrock excavators "are about 10% of our utilization out there." 

Bedrock, a San Francisco-based autonomous equipment startup, counts Austin Bridge & Road, Maverick Constructors and Haydon Cos. along with previous partners Champion Site Prep, Zachry Construction, Capitol Aggregates and Sundt as partners testing its add-on kits for autonomous equipment.

Moving that amount of earth "is something a team of people would have to be doing, super routine, in the sense that it's a very similar task 12 hours a day, seven days a week," says Eric Cylwik, Sundt director of innovation. "It's a very intense schedule. By being able to take something that's not a desirable, repeat activity for somebody off their task list, it unleashes our operators to do the more challenging and critical associated work."

Large, remote jobsites needing skilled operators where these workers are not readily available comes at a premium cost and has emerged as a fertile testbed for autonomous equipment. Bechtel has used Built Robotics excavators for trenching on solar projects in Texas, and has used autonomous pile-drivers on other projects. Moog's autonomous skidsteers were used to deliver solar panels, just-in-time for installation on a western New York solar project. Green said adding artificial intelligence and machine learning to the process has helped Sundt's operators bring their autonomous co-workers along faster.

"We've taken very experienced operators and they're teaching how a human does that task by using multiple operators," Green explains. "It's getting to feel the difference between them, because every operator is different in how they do [their job]." Using AI efficiency, "you're kind of taking that consolidated knowledge of multiple operators and turning it into one operator, and that's very powerful," he adds.

Bedrock was founded by engineers and executives from commercial autonomous vehicle startup Waymo. CEO and co-founder Boris Sofman said developing Bedrock's technology on active jobsites with real projects and contractors allows the company to quickly scale up the same way Waymo did, solving wayfinding, topography and other challenges that map-based navigation did not initially understand. 

"We've been really happy with how much a pretty sizable group, a part of the industry is embracing the possibility of this," Sofman said. He added that facing challenges—including workforce availability—that limit project capacity for contractors while ensuring that Bedrock uses an intuitive approach that is non-disruptive to partners and customers has enabled the startup to grow quickly. He also pointed to safety protections offered by geofencing autonomous equipment that prevent onsite interactions with humans on sites.

"Just the availability of labor is a huge pain point, so a lot of jobs just don't get done," Sofman said, noting "an expansionary impact on a lot of construction companies." 

With more operators retiring, younger workers may lack skills leaving "safety as one of the biggest pain points and fears in the industry," he said. "You have the ability to compress schedules and move jobs faster, which is very obviously beneficial."