The last time a piece of technology significantly overhauled the textile industry, it was 1842 with the invention of the sewing machine. Today, robots can handle parts of the garment-making process–cutting fabric, for instance–but human hands are still needed to feed fabric into sewing machines.
That’s what the web developer and inventor Jonathan Zornow wants to change. He has created a system called Sewbo, which, he claims, is responsible for the first instance of a robot sewing a T-shirt. Sewbo is designed to solve one of the biggest barriers to automating garment manufacturing. According to Howie Choset, a professor in the Robotics Institute at Carnegie Mellon, robots are bad at handling soft, flexible fabric–which is why industries that use hard materials, like cars and electronics, are light years ahead in automation.
Sewbo takes a cue from established robot counterparts. It works by using a water-soluble thermoplastic to stiffen fabric to be as sturdy as cardboard, which can then be manipulated by an off-the-shelf universal robot and fed into a sewing machine. Once the seams are stitched, the stiffened garment is put into hot water, where the plastic melts off and the shirt emerges in its soft, finished state.
In textiles, as in many global industries, automation is the holy grail. So says Eric Spackey, CEO of Bluewater Defense, an apparel manufacturer for the U.S. Department of Defense. Spackey’s factory in Puerto Rico produces 80,000 trousers a month with 523 people (each combat trouser, which has 64 separate pieces, takes about 45 minutes to make). While Bluewater Defense has automated as much as is possible, it still relies on manual labor for much of the sewing. And Bluewater is no exception. Today, clothes must still be hand-fed into a sewing machine.
Because of labor costs, many companies have moved manufacturing overseas to Asia, where workers may be paid very little to sew in dangerous conditions. Automation has the potential to change that, Spackey says. He believes full automation is coming to the garment industry in the next 5 to 10 years. “If we can make it happen where we can develop machines that can actually take woven fabric that’s cut in parts and sew it into clothing, manufacturing is going to move back to the U.S.,” he says.
Whether or not that’s the case, automation is the future, and other startups are working on it. Kniterate, which was founded several years ago, is trying to knit together ready-to-wear garments–almost like 3D printing of fabric, with no sewing needed. Fabrican uses a “spray-on fabric,” where a chemical is applied directly to the body to create a garment that molds to any individual’s physical form. SoftWare Automation uses computer-enabled cameras to track the location of the fabric and determine stitching position. But Sewbo appears to be the only startup using a stiffener to make fabric adaptable to existing robots–a cheaper solution than building a machine that uses cameras and algorithms to manipulate fabric, since Zornow estimates the cost of plastic needed to stiffen one shirt is less than 10 cents.
Zornow was inspired when he read an article describing the properties of polyvinyl alcohol, which is used in 3D printing as a temporary scaffolding so complex shapes don’t collapse as they’re being built. His concept is similar, but uses the plastic in liquid form. Once the cut fabric has been dipped in the plastic, it can be molded, aligned, and welded together.
Right now, Zornow’s prototype doesn’t do all those things, but that’s because those tasks–including cutting fabric from a pattern–have already been automated. He’s focused on creating the right process so that two cut pieces can be sewn together with no human touch needed. For the first three years of the project, which he started working on in 2012, he experimented, researched, and continued to run his idea by people in the industry in order to poke holes in the concept. It was only when he’d surpassed his own skepticism toward the project, he finally quit his job, found investors, and launched Sewbo into the world.
Polyvinyl alcohol is nontoxic and is used in a wide variety of industries, including in textiles, where it’s used to strengthen yarns for weaving. It can be found in everything from paper to contact lens lubricant to the pod-sized packages that hold dishwasher soap and laundry detergent. In Sewbo’s case, Zornow says it can be recovered and reused after it’s been washed out of a garment through an evaporation process.
Right now, Zornow’s prototype takes about 30 minutes in total to make a shirt, with most of that time spent measuring and adjusting the alignments. The actual sewing, he says, is done at “normal sewing speeds”–he believes that the robots would be able to use Sewbo’s technology to put together shirts at the same rate as the most efficient human assembly lines, if not faster.
Zornow envisions a world with no human sewing (at least not on a mass scale). Automation could mean that there would be more factories in the Western world, which would enable fashion companies to turn around new designs faster. Manufacturers in Asia might also be able to produce garments faster and cheaper. Zornow doesn’t believe that factories accustomed to an entirely manual process will have any doubts about switching over to automation, but automating would require a fairly large investment upfront, which some factories may be reticent to make.
It would also mean that millions of workers across the globe would be out of a job. While automation wouldn’t necessarily replace all of the world’s sewers–details or even whole pieces for higher-end companies would likely still be sewn by hand–it could have a serious impact on the global economy.
Since he announced his technology earlier in September, Zornow says he’s received a large amount of interest from investors and manufacturers. He has begun the search for strategic partners to help him test Sewbo in an environment bigger than his lab–as well as employees, since he has run Sewbo as a one-man-show for the last year. Zornow hopes to license Sewbo to factories around the globe.
As conversations with investors, potential hires, and manufacturers begin, Zornow is fine-tuning the assembly process, such as looking at how hot the water should be in order to melt the plastic off different kinds of dyed textiles, which can only be washed at certain temperatures, or figuring out the most efficient way to put a certain piece of clothing back together.
“From my end, it ends up looking like a jigsaw puzzle,” Zornow says. “You have to figure out with these robots, what’s the most efficient way to do this?”
[All Photos: Zack DeZon]