A lot of the work that goes into making a t-shirt or a pair of jeans has been automated, except for cutting and sewing. Now machines are taking over one of these last surviving manual jobs, which humans have performed for hundreds of years.
Some forty years ago, Ijaz Khokhar’s father set up a factory in Sialkot, a dusty industrial town of Pakistan, to make karate and martial arts uniforms. Over the years, as the country faced several political and economic upheavals, Khokhar's clothing company, Ashraf Industries, prospered.
It has now become one of the the country’s largest exporters of karate and martial arts outfits with loyal clientele from Japan to Europe.
Along the way, Khokhar saw profound technological shifts in the way business was done. The just-in-time inventory system, cross-border outsourcing, internet and cloud computing—innovations that have drastically shaped his business since the 1960s.
Every year, Khokhar travels to textile machinery exhibitions to see the latest innovations in manufacturing processes.
“I try to keep myself up to date with new trends,” he told TRT World. “There’s no escaping it.”
Yet he never imagined that one day he’d be talking about the possibility of robots becoming a threat to the way garments are sewn in his factory.
“I remember watching a documentary on it, but never thought it could become a reality.”
The automation onslaught
From harvesting cotton in the farms, to making thread, then weaving it into cloth in looms, followed by the stage of printing—the textile manufacturing cycle has largely been automated in the past two hundred years.
What remains in the hands of humans is when fabric is sewed into the clothes that we wear.
That might change as tech start-ups, encouraged by lightning technological changes, make robots that can imitate humans.
Softwear Automation, a company based in Atlanta, in the US, has built an entire assembly line manned by robots that can pick a piece of garment, arrange it properly and then sew it. This technology is called the Sewbot.
Just picking up a piece of fabric is a massive step forward for robots. Sewing and stitching has eluded machines because cloth is floppy and crumbly, difficult to handle even for humans who are not trained tailors.
Nimble finger movements can quickly adjust a piece of fabric under the needle of a sewing machine. It’s a grueling job for a worker to continuously adjust the garment under the striking needle, making sure the seam stays straight and smooth.
It’s a skill that garment factory workers in developing countries such as Pakistan, Bangladesh and India acquire over many years of mentorship.
What has come to drive them out of the factories is a combination of powerful algorithms, fast computing speed and the ever-decreasing cost of technological products.
The Sewbot work-line robots rely on high speed cameras, which see the individual threads in fabric, pinpointing the exact location where a needle strikes and adjusting the garment accordingly.
Softwear Automation sees this as a disruptive technology, which will have a lasting impact on how apparel, home textiles and garments are made. And it can do that without workers.
“Our Sewbot work-line can produce nearly twice as many finished t-shirts in an eight-hour shift as manual sewing can run 24 hours a day,” Softwear Automation’s CEO Palaniswamy Rajan said.
“It’s 80 percent more efficient.”
Rajan, a venture capitalists, invested in Softwear Automation after the company was founded in 2012.
Multiple studies by organisations such as the OECD and the World Bank have warned that automation can leave millions of people jobless, not just in developing countries but also in advanced economies.
The use of industrial robots across the automotive, electronic and others industries is at its highest, said the International Federation of Robotics.
The growth in industrial robot sales is led by Asia. Between 2011 and 2016 robot sales increased by an average of 12 percent.
An inevitable transition
Besides Softwear Automation, another start-up working to automate the repetitive sewing job, is called Sewbo, the brainchild of 30-year-old New Yorker Jon Zornow.
He devised a method to stiffen fabric by using a dissolvable chemical solution. This allows cardboard-like hard patches of material such as denim to be handled by robots. Once sewed the garment is washed with water without compromising its quality.
Zornow’s inspiration came from 3D printing. “That technology also uses water-soluable scaffolding to temporarily support objects as they are being created on the 3D printer,” Zarnow told TRT World.
His robots come off the shelf, making it easier for companies to find replacements when needed.
That’s a tell-tale sign of how one technology is giving birth to another. It also explains why prices are continuously coming down.
In his seminal book, The Rise of the Robots, Martin Ford writes that advanced vision technology being used in industrial robots emerged from gaming consoles.
The US government has played a crucial role in the promotion of this technology as well.
Softwear Automation’s Sewbot concept was conceived by Dr. Steve Dickerson, a Professor Emeritus of Mechatronics at Georgia Tech.
The US Department of Defense supported him with $1.7 million, seeking locally manufactured military uniforms for the US soldiers in return.
His firm was later acquired by Rajan’s investment fund, CTW Venture Partners.
Made in USA—by the robots
Softwear Automation and other proponents of robotics say that automation will help bring manufacturing back to the US.
But unlike what US President Donald Trump has been telling his supporters, this won’t necessarily translate into more jobs.
Up until the 1990s, the US met most of its garment and apparel needs from domestic factories. But then the tide turned.
Production was outsourced to Mexico, China and India, almost eliminating all local production, and leaving thousands of people jobless.
Between 1990 and 2012, the US textile and apparel sub-sector lost 1.2 million jobs, or more than 76 percent, as imports replaced domestically produced garments, according to The New York Times.
The companies that do remain in business face ageing workers on sewing machines, high wages and an uninterested workforce that views textile factories as sweatshops.
And it’s not just the wages that pose a challenge.
“In the firms I visit, people have forgotten how it is to run this business,” said Zornow.
Even though some manufactures want to bring apparel and garment production back to the US, chances of that happening on any large scale are slim, he said.
“People ask me if it's possible for factories to move back to the US, and they want the answer to be yes. But I am not sure. Garment manufacturing is a complicated process. You still need humans to analyse every piece that comes off the line.”
“Take silk or chiffon. How do you think the machine will handle that?”
The supply chain plays a crucial part as well, he said. “Where do you source the cotton from, do you have all the materials? All these things come into play,” Khokhar said.
Perhaps the most encouraging point for bringing back the automated sewing work to the US is proximity to the consumer market.
Big retailers such as Walmart, which source garments and apparel from developing countries, have to wait months before they can put a new line of clothes on the shelf. This is because clothing samples are sent back and forth for approval.
Softwear Automation has received funding from Walmart as the retail giant wants to meet the demand of its customers before fashion wears out.
What about the people?
Both Rajan and Zornow are focusing primarily on the US market for their robots. They see a timeline of at least five years for the technology to start making any marked change.
“A brand or manufacturer that’s willing to commit can have up to 10 percent of its manufacturing to the US within five years of setting up a Sewbot factory,” Rajan told TRT World.
However, the change, they said, is inevitable. And it won’t necessarily be bad for the workers.
“Sewbots are not meant to replace all of garment manufacturing,” said Rajan.” Using automation for high-volume basic apparel enables sewing machine workers to focus more on complex garments, while advancing their skill sets and commanding higher wages all around the world.”
The shift to e-commerce and on-demand manufacturing makes this even more important, he said.
For Zornow, automation is a necessary step in the cycle of industrialisation. “Two hundred years ago, how many people were living in poverty and dying of starvation? Has industrialised technology helped? The answer is definitely yes.”
He said he went public with his plans at an early stage to provide an opportunity to people in textile-dependent countries such as Bangladesh to prepare for the technological onslaught.
“I read Bangladeshi newspapers everyday. There is so much fanfare every time there is investment in the garment sector. But I don’t wanna see that because we know that economic landscape is changing and that money could be invested in other sectors.”
The fear of automation replacing humans has already pushed industry leaders like Elon Musk to debate this incoming disruption. "AI is a fundamental risk to the existence of human civilisation and I don't think people will fully appreciate that," Musk said in July 2017.
Here again, it would be the workers in developing countries that are most likely to be hit by automation.
Pakistani businessman Khokhar said it won’t be easy for governments of developing economies to deal with large scale job losses in the textile industry.
“I’m still skeptical. Automaton is not going to take jobs away from humans any time soon. This work is too complicated for machines to handle,” he said.
“But if that happens, the consequences would be devastating.”