The Case for Letting Robots Make Our Clothes
I visited the garment factory workers who could lose their jobs to automation. But could robots also change their lives?
The more you know about the $3 trillion garment industry, the harder it is to shop. That's why—if you're a (usually failed) conscious consumer like me—you're stuck between the clothing rack and a hard place. Like cheaper blood diamonds, our t-shirts are often the product of poorly paid labor—stitched by underfed hands, and packed in hot and smoky warehouses while toxic dye runs into a river somewhere outside of Dhaka.
It starts with excessive consumption: In the West, we buy at least six times as much clothing as our our counterparts in China. Sourcing material adds an extra layer, since countries manufacturing clothes don't always have local cotton, lycra or leather. And to meet the demand, the majority of our clothes are made in Asia—where workers are paid something like 50 to 90 cents an hour—and then shipped across the world.
Keeping costs low, of course, comes at a human expense: Many of the garment workers who make our clothes experience miserable wages, health care access, unsanitary conditions and forced overtime. Some workers also told Human Rights Watch they experienced physical abuse and extreme pressure to produce high quotas of clothing.
But there's a possible solution to these human rights abuses: robots.
When we think of automation—specifically replacing workers with robots—our reaction is worry about the social burden of laying off thousands of workers. We think of the cruelty of decimating a 60 million person workforce. But what happens when you automate jobs in a system that has largely failed to treat its workers with dignity?
Earlier this year I visited garment factories in Hyderabad, a city in the southern part of India. These facilities produce clothes for international companies like Zara and Banana Republic.
At a four-story factory called Chermas, built at the edge of a dusty road in an industrial neighborhood, there were 800 people at work. Each section of the building handled a range of tasks. There were pattern makers cutting designs, women inserting individual buttons into a little machine that stamped them onto a shirt, men shoveling huge swaths of clothes into industrial washers and dryers, and workers spot-checking zippers and loose threads.
The work was unquestionably painstaking, monotonous, and physically taxing.
"Everything here is done manually," Shaik Abdul, Chermas' merchandiser, told me as we walked through.
The factory was clean and organized. There was a 15-minute chai break, where workers would go to an outdoor cafeteria for little glasses of tea. But India's scorching summer was two months away and there was no air conditioning. Workers toiled side by side on shared tables, either standing all day or hunched over a machine. The work was unquestionably painstaking, monotonous, and physically taxing.
Subash, a 46-year-old worker who makes samples at the factory, told me that his job was a relatively good one compared to others in his community. The company, he said, takes workers to the hospital if they get sick or hurt. He said most workers' salaries ranged from 4,000 ($61) to 10,000 rupees ($61-154) per month—a low to middle income salary for a single person in urban India, just enough to survive.
"I plan to stay here," he said. "Or eventually open my own tailoring shop."
Other workers also told me their goal was to start their own businesses at home. This is common in India, where many people still buy cloth and get their clothes made by tailors, but also competitive.
There are already several companies working to automate garment making and replace the work that people like Subash do every day. Sewbo and Softwear, both of which manufacture robots that make clothes, are hoping their technology will enter the factories themselves.
"It's higher quality production," Palinaswamy "Raj" Rajan, the CEO of Softwear, told me over the phone from Atlanta. "Less failures, less material. So we enable that. With robots we can do things with higher quality."
Rajan said his company, which got a grant from the US military research outfit DARPA, is hoping to make the garment industry more efficient. Softwear's robots—or Sewbots™, as the company calls them—include automated sewing machines. This technology could replace the workers I saw feeding material into sewing machines, or could hem shirts and pants. So far, Raj said, small companies across the US have started using Softwear to make clothes, bedsheets, and other sewn linens.
There are technological limitations to most of the automated "sewbots" on the market right now. Rajan told me his machines, which a spokesperson said cost about as much as conventional garment machinery, can't do the nimble work of seamstresses when it comes to fancy outfits that have frills and layers of lace. And some of the baubles and additions to clothing can't be done by a standardized machine.
Robots would also fall short in a garment factory since most current machines can only handle specific kinds of materials, not the variety of leather and suede and linen that comprise our market, robotics expert David Bourne told me. Cutting different materials would require different devices, whether it's cutting with pressurized water or blades.
We're not close to replacing human garment labor altogether.
"The big problem with cows is they have bee stings, barbed wire cuts, so there's all kinds of imperfections that have to be avoided during the cutting process [of leather]," he said.
Bourne, who works on garment automation at Carnegie Mellon's Robotics Institute, said there's already technology that could make t-shirts and other simple garments, but we're not close to replacing human garment labor altogether given the diversity of our designs.
Even if robots catch up to humans, that brings us back to people like Subash, and his livelihood. The garment industry could be faced with a big labor problem in communities where millions of people could be left without work.
Large clothing corporations started eyeing cheap, emerging markets back in the 1980s. China set up a Special Economic Zone in Shenzhen to attract foreign investment and employ thousands of factory workers. In countries like Thailand and Cambodia, foreign companies started to employ dispersed homeworkers to manufacture clothes in the 1990s. And by 2005, garment factories employed 1.8 million people (1.5 million of whom were women) in Bangladesh.
Today the industry employs between 60 to 75 million people across the globe, including 3 million in Bangladesh. There are entire communities that have been built around the clothing factories, and families who depend on the daily wages for the most basic necessities. There's a strongly held and legitimate concern among labor activists that automating even part of the garment manufacturing process will put not just people, but entire neighborhoods out of jobs.
"If automation comes in, it will have huge consequences to millions of people, as currently, the textile industry is fed by huge quantities of unskilled labor," said Leigh McAlea, a spokesperson for TRAID, a London-based fair trade advocacy organization. "And in poor countries, there is an abundance of people willing to work for extremely low, subsistence level wages."
And countries like India are hardly looking to automation. In fact, the conservative Narendra Modi administration has doubled down on old-school, labor-intensive manufacturing.
"We're actually preparing a huge skillforce," said Sudha Rani, the CEO of Abhihaara Social Enterprise, an organization that channels government funds to run a 30-person garment factory and training institute.
When I visited her organization in Hyderabad, there were young women learning how to sew garments—mostly Indian style kurtas and salwar kameezes. Rani said she receives money from the Ministry of Textiles to run the program, part of a countrywide effort to train more people in viable industries.
Can't we just pay people better wages?
Unfortunately, the pipeline for such work usually leads to the same, bleak situation: sprawling factories with low wages and unregulated labor practices. Unless people are able to start their own side hustles or tailoring shops, or work in places like Abhihaara, which employs a few of their trainees in a less intense setting, they're vulnerable to larger threats of the industries. Threats that can lead to fires, illness, and factory collapses like the infamous Rana Factory collapse in Bangladesh.
There is, of course, the question of fair trade clothing and Abhihaara is looking to these spaces to sell their wares. I've wondered this a lot in my own hunt for ethical clothing: Can't we just pay people better wages and start buying more expensive (and fewer) clothes?
There are plenty of businesses that have cropped up around this premise, with varying degrees of success. Everlane, for example, is a stylish online retailer that swears by transparency about the factories it uses, some of which are in China. And larger companies like H&M have attempted to clean up their reputation of poor labor practices with a commitment to pay better wages.
But most of these efforts have happened on a small scale, and these practices very rarely affect the clothing most often purchased by middle- and low-income people in the West. Some of the companies that claim to have safer and better practices, like TOMS shoes, end up doing what other standard companies are doing—making their shoes in China, for comparable wages, and perpetuating an inefficient supply chain.
Meanwhile, there's spotty evidence for relying on a conscious consumer. As Alden Wicker argued in Quartz, consumers who were trying to make more conscious choices about what they bought had almost little to no impact. "Making series of small, ethical purchasing decisions while ignoring the structural incentives for companies' unsustainable business models won't change the world as quickly as we want," Wicker writes.
Leila Janah, a social entrepreneur, is the CEO of SamaSource, a company that links low income communities across the world to better jobs, often in the digital sector. She also has a company called LXMI, which has introduced artisanal, fair trade makeup to larger storefronts like Sephora. The products are made by women in the Nile River region who are now earning up to three times as the average wages.
SamaSource has been dealing directly with this issue for years—the social enterprise works with some of the poorest people in countries like Ethiopia and Kenya to move people away from taxing, low-paid jobs into other skilled work. Janah thinks technology could indeed wipe out certain types of work, but she said it doesn't make sense to fight that kind of progress.
"Anything that is not skill building—anything that is emotionally taxing and draining, if we can automate those unfulfilling types of work, all the better," she said.
This doesn't work without systemic changes. Companies without human workers, Janah said, should be paying higher taxes that support more human social goals, such as safety nets, job retraining, and universal basic income. And people should be rerouted to work that could be done with dignity—she gave the examples of artisanal design, digital work and elder care.
The garment industry is primed for that kind of shift in human labor, said Carnegie Mellon's Bourne, who has been working on both the engineering and economic aspect of automating the garment industry. His vision for the industry includes the robots he's helping to create, but also humans.
"These are the kinds of things people are so good at."
Bourne said the right model would be to decentralize factories altogether by creating small manufacturing hubs across the world. Then, people could order clothes online close to their homes, which would be made predominantly by both robots and people in these hubs. The robots, in these instances, could do some of the larger stitching and cutting tasks, while people work on finer details, customer service, and design.
"Some of the steps are not really good problems for robotics, maybe ever," he said. "I'm sure in 20 years I can make robots for $100,000 that can do something else. But these are the kinds of things people are so good at."
The model that Bourne describes seems almost utopian, but could do what many countries, including the US, are looking to do: stabilize local economies, fight against dangerous labor practices, and leave room for human innovation. That last part is something I kept coming back to in the garment factories I visited. We often think that taking away jobs is a dead end, miserable practice, but building capacity for people to take their skills into new projects could offer a solution.
The case for automating certain garment factory jobs isn't a case for erasing workers, it's the idea that we should be investing in innovation and ingenuity. And it would only work with a conscious and massive shift in the industry.
Back in Hyderabad I visited a small garment factory called SQube Fashions run by Suresh Yedla, a thin 30-something guy who had left a life of poverty behind in his village to pursue a career in design. With help from his father's tailoring shop and government scholarships, Yedla attended the National Institution of Fashion Technology and started working in the garment industry soon after. But he was quickly bored by the mass production of hundreds of garments, and the painstaking factory work.
When he saved a little bit of money, Yedla got a government loan to set up his own small shop that handles some outsourcing work to companies like Zara, employs about 30 workers, and produces 200 to 300 garments a day.
While the everyday operations keep the shop running, Yedla said his main interest was never pumping out garments, it was to make innovative designs. His visible exhaustion from managing the room of workers disappeared when he showed me some of the things he made—tunics for people in wheelchairs, more fashionable hospital pajamas for the chronically ill.
"Whenever I don't have work to manage I start making my own goods," he said. "That's what I always wanted to do."
Coincidentally, that's not a job for a robot.