In a laboratory in Pittsburgh, a four-legged creature stands staring at a window. In this case, “at” is the operative word. While you might expect “Spot” to be watching for cars or squirrels, he’s actually looking for defects—scratches, dings and other damages that might occur in the installation and handling of doors and windows.

In a construction market that faces severe labor shortages, Spot, an autonomous dog-like robot, is one company’s answer to the need for human inspectors.

“Customers have high expectations for flawless window frames after installation,” said Jorge Vasquez, a Ph.D. student at Carnegie Mellon University (CMU). “Defects, such as scratches, dents and bends can negatively impact those expectations … however, most companies still rely on human inspectors, which are subjective, inconsistent and typically slow.”

Spot, an autonomous dog-like robot, is one company’s answer to the need for human inspectors.

There are also fewer of them. According to Associated Builders and Contractors, the U.S. construction industry needs to hire an estimated 546,000 additional workers on top of its normal pace in 2023—and that’s just to keep up. In Japan, matters are even worse, said Hidemitsu Hori, chairperson of YKK AP Inc.

“One of the problems we’re facing in Japan … we can see in the very near future that the shortage of labor will be more and more serious. That’s why we have to take action,” Hori told [DWM] in a private press conference.

Amid a grand opening ceremony for a new 7,576-sq.-ft. research and development center, officials for Hori’s company and other invited guests looked on as researchers brought to life another alternative. In several demonstrations, Spot, and other autonomous technologies, navigated mock jobsites to identify and inspect fenestration products for damages—some too small for even the human eye to detect. Others tracked human activities to ensure compliance with safety protocols. Through a new research arm, dubbed YKK AP Technologies Lab (NA) Inc., and a partnership with CMU, YKK AP Inc. is investing in those and other technologies to aid in functions once reserved for humans. As the industry faces a worsening labor shortage, the company is seeking answers among autonomous robots, unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) programmed to navigate and work on jobsites.

Unmanned aerial vehicles (UAVs) are programmed to perform tasks and to monitor the whereabouts of workers to stay out of the way and avoid collision.

Located in an old steel mill that’s been revamped into a LEED Gold-rated facility, YKK-funded researchers are now working to crack the code on everything from jobsite safety to robotic installers. Led by Shiori Fukada, YKK AP Inc.’s chief information and digital officer, the operation includes more than 50 students and researchers, working in tandem with YKK employees as visiting scholars. The group will also focus on “the digital transformation of YKK AP Inc.’s business and the construction industry as a whole,” company officials said. Through a “virtual factory,” researchers aim to improve logistics, supply chain management and global management systems. The end goal includes digitalizing and automating the door and window industry—the full door and window industry, not just manufacturing.

As the company moves forward with a new manufacturing facility in Georgia for residential products and robust plans for market expansion, the same labor issues strike home for YKK AP America.

“We hear about the labor compression here in the U.S., but it is much more urgent in Japan,” said Oliver Stepe, president and CEO of YKK AP America. “As you may know, the population is actually declining and based on the stack of age brackets, there are a lot of skilled, field workers who are leaving and retiring, and no one is coming in behind [them].”

So far, most of the industry’s focus for robotics has been on automated production, Stepe said. But, “The real critical area now, especially in Japan, but soon coming our way in the U.S., is in the field,” he said.

Before machines and humans can coexist on jobsites, first they must learn to work together. In student-led demonstrations, researchers showed how devices equipped with cameras and lidar can be programmed to track human activities. In the event that someone forgets to wear protective equipment, or they enter a restricted work area, the system will automatically detect those mistakes and sound notifications. The same system also analyzes the size of doors and windows, warning workers who attempt to pick up and install products that are too heavy for one person to handle.

From left to right: Hidemitsu Hori, chairperson of YKK AP Inc., Shiori Fukada, YKK AP Inc.’s chief information and digital officer, and Oliver Stepe, president and CEO of YKK AP America

Autonomous robots have also been equipped to track stationary and moving obstacles across jobsites to stay out of the way of humans. In one of the company’s demonstrations, an aerial vehicle tracked the movement of a person, choosing to pass behind them rather than enter their projected pathway.

Behind all of these accomplishments are deep learning algorithms, researchers said. And they have a way to go, but the company plans to develop and deploy robots that can install doors and windows—not decades down the road, but within the next five to six years, Hori said.

“We will need to make improvements to robots, but also we need to improve our products to make it easier to implement [these systems],” he said. In the process, it’s possible that the company will tweak its doors and windows to optimize them for robotic handling. “It’s intuitive to think that will be one outcome,” Steppe said.

At the same time, YKK isn’t giving up on humans.

“Nothing happens without all of us,” Stepe suggested. “Equally important to automation and machine technologies is human capital.” That’s something that YKK companies have always held dear, he said, and that isn’t expected to change, even with the arrival of robots.


Leave a Reply

Your email address will not be published. Required fields are marked *