LONDON, United Kingdom — In the 1990s, fashion’s relationship with robots was the stuff of fantasy. On the runway of Alexander McQueen’s imaginative Spring/Summer 1999 show, two robotic arms spray-painted a white dress worn by Shalom Harlow. Today, the industry’s relationship with automation is much more practical.
In the distribution centres of e-commerce giants like the Yoox Net-a-Porter Group and Amazon (which, in 2012, paid $775 million to acquire Kiva Systems, a manufacturer of robotic fulfilment systems used by Gap, Gilt Groupe and Saks 5th Avenue) software-controlled robots routinely navigate giant warehouses, picking and transporting inventory faster and more accurately than humans, enabling services like same-day delivery.
“Automated storage and retrieval systems provide high storage density as well as inventory accuracy and management, yet require a smaller footprint,” explains Steve Crease, director of operations at Yoox Net-a-Porter Group, which uses ASRS to deliver its “key service level” of same-day delivery. The robots retrieve items from giant storage systems and bring them to human-manned packing stations, where “light displays instruct the operator where and how to allocate stock to orders,” he says.
But beyond e-commerce fulfilment, fashion has yet to seriously embrace robotics, even as robots have become cheaper and smarter, driven in part by advances in artificial intelligence. In 2014, global sales of industrial robots reached 229,261 units, according to the International Federation of Robotics. The same year, shipments of industrial robots in the textiles, leather and apparel sectors numbered just 289 (up from 25 in 2009).
The use of robots in garment manufacturing, for instance, remains low. This is partly due to a key technical challenge: while automobiles or electronics products — often assembled with the assistance of robots — are largely comprised of hard parts, garment manufacturing involves pliable, elastic fabrics that have, traditionally, made the process difficult to automate.
SoftWear Automation, an Atlanta-based robotics firm focused on sewn product manufacturing, believes it has the solution. SoftWear’s automated system uses high-speed photography to take pictures of garments as robots work on them. The images are interpreted in realtime by software, which, in turn, directs the robot’s movements. “Our robots know when the fabric is being stretched and by exactly how much, and it can make adjustments accordingly,” explains KP Reddy, chief executive officer of the company, which was born out of a university research project three years ago.
In the last few years, the cost of this kind of system has also dropped rapidly. “To buy a $5 million robot to do the work of a $50,000 a year employee, you could never get the economics to work,” says Reddy. Today, SoftWear’s systems cost $50,000 to $100,000. “If your seamstress is costing you $50,000 a year, you’re getting a sub-three-year payback,” he says (in the developing world, however, garment workers earn considerably less — in 2014, Bangladesh’s government raised its minimum wage for garment workers to the equivalent of $68 a month). It takes one person to manage four or five Softwear robots, enabling companies to limit their labour costs or invest in staff in other functions of the business.
At the moment, SoftWear’s systems can make a limited range of clothing such as jeans, basic dresses and skirts, as well as home goods like towels and curtains. Softwear declined to reveal its current apparel clients, but Reddy says most are focussed on fast fashion or athletic apparel.
Indeed, some apparel companies are ahead of the curve on automation. Later this year, Germany sportswear giant Adidas will open Speedfactory, its first shoe manufacturing plant controlled largely by robots. The company, which employs over a million workers in contract factories, and sources most of its production in Asia, is tapping automated production as a way to quickly produce and deliver goods in response to consumer demand in major markets closer to home. The first Speedfactory will be in Ansbach in Southern Germany.
The factory is also a bid to deliver personalised products through mass-customised manufacturing. “One of the most important trends will be toward customisation,” says Martin Ford, author of books including “Rise of the Robots: Technology and the Threat of a Jobless Future,” of the use of robotics in apparel manufacturing. “Customers will be able to design or customise their own clothing and then have in produced in automated factories and delivered within days.”
Robotics are also beginning to revolutionise fashion manufacturing in the developing world, according to KP Reddy of SoftWear, which has shipped machines to major garment manufacturing hubs including Turkey, Bangladesh, India, China and Sri Lanka. “In those markets, most of these factories are in rural areas. The young people are not staying in these rural markets, they’re moving to city centres,” he says. “The [factory] CEOs are telling us that in 10 years it’s not going to be a function of cost of labour — it’s going to be a function of no labour. There will not be people in these towns and villages to do this kind of work.”
Robots could also improve efficiency at retail. In a 2015 report on The Future Of Jobs, J.P. Gownder, vice president and principal analyst at Forrester, forecast that by 2022, 76 percent of sales tasks — such as cashier work, restocking shelves and inventory control — will be done by robots (which he defines as both hardware that automates physical tasks and software that performs intellectual tasks). “Customers will increasingly use mobile technologies to access information and advice while in stores,” adds Martin Ford, thereby lowering the need for human sales assistants.
As the pace of automation continues — the IFR predicts double-digit growth in global sales of industrial robots between 2015 and 2018 — Ford predicts that, in the fashion industry as well as the broader economy, “robotic technologies and machine learning technology are likely to eventually automate nearly any job that is on some level routine and predictable.”
This will challenge companies to develop plans for human-robot collaboration among employees and implement these technologies gradually to give their businesses and workers time to adapt to the change. According to KP Reddy, fashion is now in a position to follow the lead of industries like automotive and electronics, which have tapped robotics to implement just-in-time delivery and just-in-time manufacturing. “If we can minimise the amount of clearance racks, the amount of product that just gets kind of wasted and can move more on demand, that’s just a better thing for everybody,” he says.