Levi Strauss & Co. participated in a project spearheaded by industrial manufacturing conglomerate Siemens focused on investigating robotics use cases in apparel production. A spokesperson for the denim giant confirmed to Reuters that the company participated in the project’s early phases but declined to offer further details to Rivet.
Siemens’ ambitions to automate apparel manufacturing emerged from initial efforts to create software to guide robots that could handle all types of flexible materials, such as thin wire cables.
According to Eugen Solowjow, head of the research group at Siemens, the company soon realized that clothing was a prime target for this kind of robotics technology.
“Clothing is the last trillion-dollar industry that hasn’t been automated,” Solowjow told Reuters.
For some sectors, robotics remains in its relative infancy when it comes to deployment. According to a recent study from supply chain robotics technology provider Berkshire Grey, only 13 percent of executives say they are currently using robotic automation. They are aware though of where the industry is headed, as evidenced by 51 percent of executives in the process of adopting robotics or planning to.
But of the 200 senior-level supply chain executives Berkshire Grey surveyed, apparel manufacturing isn’t exactly on their priority list. As many as 62 percent say they are likely to use automation to support packaging/labeling, while 59 percent would use it for item sortation. Fifty-eight percent would leverage the tech for returns and goods retrieval.
Siemens has been taking the time to team up with businesses willing to make automated apparel manufacturing a reality. The company partnered with the Advanced Robotics for Manufacturing (ARM) Institute, a Department of Defense-funded organization aimed at making robotics, automation and AI more accessible to large and small U.S. manufacturers.
The teams’ early work integrated sewing machines with collaborative robot systems and designed an end effector capable of lifting and controlling a single large ply of fabric. Recent projects have built upon these developments to be able to robotically conduct more advanced operations like hemming, fabric fusing, pocket setting and curved stitches.
The tech behind robotic sewing
The two firms then turned to Sewbo, a company that wants to address a common problem that prevents robotics from meshing with apparel production—the technology often has difficulty trying to handle limp, flexible or floppy fabrics, and thus can’t start the sewing process.
“Almost all of the other existing approaches to automation in this space are super complicated, both in the technical approach, and also the actual machines,” Sebo founder Jon Zornow told Rivet. “They are mechanically complicated or digitally doing a lot of work, and it leaves so many places for things to break, and to become over-expensive and too engineered. Also, paradoxically, they are very constrained in their capabilities because most of the solutions out there are hyper-focused on certain specific tasks like pocket setting.”
Because the machines are also expensive, according to Zornow, the upfront investment and maintenance costs are also high. To make matters tougher, the downtime can be substantial, he said.
“Consequentially, you sort of find this paradigm where although a lot of the tools do exist, they’re not really getting used,” Zornow said.
Rather than teach robots how to handle cloth, Sewbo temporarily stiffens the fabric with a nontoxic polymer, enabling off-the-shelf industrial robots to build garments from rigid cloth, just as if they were working with sheet metal.
Zornow told Rivet that the use of the stiffening agent was the “big breakthrough” that made the technology innovation possible.
“The sort of extreme complexity of handling fabrics was always the obvious problem that the robots were going to struggle with,” Zornow said, noting that the water-soluble thermoplastic polymer’s ability to hold up during 3D printing was the inspiration to use it in the garment stiffening process.
The fabric panels can be molded and welded before being permanently sewn together. Once complete, the finished garment is washed to remove the stiffening agent, leaving a soft, fully sewn article of clothing.
Beyond the work with Sewbo, Siemens’ research into automated apparel manufacturing eventually grew to include Levi’s and Bluewater Defense LLC, a small U.S. maker of military uniforms. They received $1.5 million in grants from the Pittsburgh-based AMR Institute to experiment with the technique.
As the use cases for robotic sewing remain few and far between for now, Zornow believes several things need to happen before the technology truly takes over apparel manufacturing. First and foremost, Sewbo and other industry players will have to bring costs down to the point where a manufacturer would be willing to buy the technology en masse and deploy it on a larger scale.
“We need to be able to expand the capabilities so that we can provide that flexible functionality. Right now, it’s an R&D demonstration. So when we do something, it’s a demo that we’ve set up and we’re going to sew pockets,” Zornow said. “We show up with our machine that’s been programmed to sew pockets, and it can do it—and that’s great—but eventually manufacturers are going to want to do this without us present. And they’re just going to need very simple systems that are reliable, cheap and easy to use.”
Zornow believes that Sewbo has reached that inflection point, where “we have a lot of basic skills—enough that we’ve now reached the threshold where I think that a U.S. manufacturer could actually profitably employ these machines. Now, we’re shifting gears for commercialization, taking this, encapsulating it in a product and getting it onto the factory floor.”
Saitex’s seeks to capitalize on robotics growth
Sanjeev Bahl, founder and CEO of Vietnam-based denim manufacturer Saitex, is preparing to install his first experimental Sewbo machine in the B Corp’s 52,000-square-foot jeans factory in downtown Los Angeles.
Zornow said that he hopes the robotics integration at Saitex will prove out by the end of 2023, before thinking of potentially expanding on the limited run.
“We’re going to have the robots do a portion of the work and they’re going to hand off to the workers to do the rest,” Zornow said. “And that’s going to actually have a finished product at the end. If we’re improving the productivity of the workers by X percent, then, there’s actual value there, and they’re ending up with products that they can sell.”
“If [the machine] works,” Bahl told Reuters, “I think there’s no reason not to have large-scale [jeans] manufacturing here in the U.S. again.”
Currently, Saitex uses robotic arms in the L.A. facility to spray each pair of jeans during the denim fading process. The robotic arms are controlled by a proprietary software that is built to spray each pair of jeans without any mistakes. An AI-powered function records a human sprayer, analyzing the process and using the data to enable the robot to copy the same spraying technique.
Other players are entering the automated apparel manufacturing space. One such example is Germany-based Robotextile, which wants to eliminate the challenges textiles can bring into the manufacturing process, particularly when a worker manually loads a sewing machine. The startup uses robots and grippers to automatically remove layers of fabric from a cut-to-size stack and feed them individually to the next production step, without picking up the bottom layer of fabric in the process.
Jury is still out on robotics’ labor impact
The question, as all automation tales go, is how any scaling of robotics would impact the labor force.
Estimates by the International Labour Organization suggest that robotics will replace 64 percent of textile, clothing and footwear workers in Indonesia, 86 percent in Vietnam and 88 percent in Cambodia.
But at a time where there’s a current labor shortage, robotics may be a welcome addition for businesses in the nearer term. More than half (57 percent) of executives believe labor shortages have hindered their ability to meet demand, according to the Berkshire Grey study. And as fewer younger people apply for these supply chain jobs, 71 percent of executives believe robotics automation is necessary.
That’s not to say employees and robotics cannot coexist. Fifty-one percent of executives believe implementing automation will increase employee satisfaction, and 43 percent believe it will reduce employee turnover, the study said.
“Almost every time, all managers want to talk about is the challenges that they have staffing their factory,” said Zornow. “A lot of these factories have high turnover rates, so a sort of a typical number that that I’ve heard is 10 percent monthly. If a factory has 30,000 sewists working for them, that means they have to go and recruit and train 3,000 workers every month, just to keep the lights on maintain their existing business. That’s a story I’ve heard many times over now, so it seems the short-term impact is probably going to be relieving some of these woes.”
Editor’s note: This article was corrected to clarify that Levi’s participated in the Siemens and Sewbo project and isn’t directly piloting the technology itself.