AI进军建筑业，能否解决“用工荒”？

When Logan Farrell places one of his Rugged Robots on the floor of a commercial construction project, it knows exactly what to do: draw the layout of the building.

The robot references 3D computer architectural and engineering designs to place details like lines, columns, and the diameter of water pipes throughout the construction zone. This task was traditionally done with a set of plans, tape measure, and chalk line by various trades, including architects, plumbers, structural engineers, and electricians. With a Rugged Robot, that’s all marked up with precision—and no human interaction.

“It is ‘intelligent’ in the sense it can interpret computer-assisted design data and bring it into the real world,” says Farrell, who founded Rugged Robots with Derrick Morse and also serves as the company’s chief technology officer.

Rugged Robots is just one of the technology companies leveraging emerging artificial intelligence technologies to solve construction challenges in the built world. Farrell explains that many of his peers are looking at modeling-stage and early construction pain points, and it’s just the beginning for innovation in the construction industry.

Companies like Rugged and Dusty Robotics are tackling this stage by using robots to help draw life-size building model layout plans at construction sites to improve efficiency and limit errors. They have developed robots that print building layouts on the floor of construction sites, so workers know where—and where not—to build. Similarly, companies including BotBuilt and Canvas are using robots to help with the actual building process, by measuring more precisely, making use of alternative materials, and installing laborious components like drywall.

The motivation for many of these founders is to use technology to assist human labor in the process. A lot of it has to do with the labor shortage in construction; experienced workers are retiring out of various construction-focused industries, and there are less new skilled workers taking their places. Morse, Rugged’s chief executive officer, says he saw the workforce not only leaving the ability to complete the task, but taking valuable knowledge with them. The greener employers often took longer to complete the same task.

“Right now, we’re retiring two people for every one customer we bring in, so we have a huge need to bring new workers up on skills quickly,” says Kevin Albert, founder and chief executive officer of Canvas, a robotics company that installs walls at construction sites.

Canvas robots, Albert explains, take over a labor-intensive trade, drywall installation, and put the human power in programming a robot. The robot, equipped with the plans and materials, then plots its own course to install drywall using a suite of sensors for mapping and a robotic arm for material application. It’s trained to use all the necessary hand tools too.

Albert says that the accuracy of the machine reduces the number of steps in the traditional process and can speed up work. What may take human labor five days to complete will take Canvas robots about two days to do.

While the machine is installing the drywall, the human labor has transitioned to higher skill tasks, overseeing the robots but also managing teams, adjusting the scheduling, moving forward to the next step of the construction process, and ensuring safety protocols. Albert notes that it also creates opportunities for people, such as those with disabilities, to be involved in the construction trade who may not otherwise be physically able to do some of the building tasks.

“We’ve seen it level the playing field,” Albert says.

These founders are also keeping construction culture in mind; instead of coming in as a replacement, they position themselves as partners. Brent Wadas, cofounder and chief executive officer of BotBuilt, spent years getting to know builders to understand how robots could help with the manufacturing process for critical building materials. His company builds residential home frames using robots, a time frame that typically takes four to five weeks without the assistance of technology. With robots, it’s done in two to five hours.

Wadas says that he wanted to find a way to decrease residential home prices, referencing the housing affordability crisis in the U.S. One big way to do that is to lower the cost of construction. By speeding up the process, and with more precision, the small-scale developers and custom home builders don’t have to pass along as high of a cost to the consumer. They can also complete homes more quickly, allowing them to build more in the same amount of time.

For Wadas, that included identifying and developing a lesser-used but more sustainable type of lumber, stud wood, that his robots could use in the framing process. It costs 40% to 60% less than imported Scottish pine and other commonly used lumber for residential homes, he says. Previously, it was often discarded because it sometimes has imperfections that are tough to catch in traditional manufacturing lines as a part of massive factories. But with high-touch robots, which can be programmed to notice inconsistencies with a computer vision system, the stud wood is easily adapted for framing.

Wadas, like other robotics company executives, views A.I.-driven machines as essential equipment for construction teams, as you would an excavator, a forklift, or a crane—all machines operated by a human.

“Our goal is to create a ‘power tool’ that can be operated by a human at an easy pace in a safer, controlled space,” Wadas says, noting that he has lectured at community colleges and vocational schools on robotics and entrepreneurship to mechatronics students. “Hopefully, we are educating the next generation that there can be great joy in carpentry and construction.”

“I think it will do great things for recruiting numbers,” he adds: “By the way, I play with robots. What do you do?”