Just-Style: Sewbots™ Pave the Way for Factory of the Future
While the use of robotics in the apparel production process remains a frustratingly elusive panacea, fully automated worklines for T-shirts and work cells for jeans operations are finally on the horizon – and with them the potential to disrupt and redesign the traditional clothing supply chain.
Listening to Pete Santora, vice president of sales and marketing at Georgia Tech spin-off SoftWear Automation, it’s clear he believes the apparel industry is ripe for change.
It’s not just that today’s long, linear apparel supply chain is “irrelevant” when faced with the need for shorter production production runs and increased flexibility to meet rapidly changing consumer demands, but: “The apparel supply chain has always been broken.”
Speaking in a panel discussion on ‘Embrace your digital reality’, organised by Gerber Technology at last month’s Texprocess apparel technology trade show in Frankfurt, Germany, he explains: “We built this great global supply chain based on cheap labour. When we moved all manufacturing out of the US and went overseas, that was really the start of a completely broken process…and the reason why it’s broken is that it’s not sustainable.
“We go to China, we consume cheap labour; we go to Vietnam, we consume cheap labour; we go to Bangladesh, we consume cheap labour; we go to Ethiopia, we consume cheap labour. And guess what? We’re running out of cheap labour to consume. So we’ve built a business model on top of a broken process.
“This is the place we live in. We buy clothes because they’re cheap, we wait for the discounted sale, and that drives this need for cheap labour.
“So the input of automation, of robotics into apparel manufacturing allows us to change this model to come closer to the customer, to come closer to the supply chain, be more vertically integrated and then find cost reduction there.”
‘SewLocal’ among benefits
SoftWear Automation believes that in addition to enabling manufacturers to ‘SewLocal’ by moving their supply chains closer to the customer, and creating higher quality products at a lower cost, its automated ‘Sewbot’ driven worklines and workcells also have the potential to tackle a wider range of garment manufacturing challenges.
These include the shortage of skilled labour in the US, the financial and environmental cost of transporting textiles and apparel over long-distances, and the increasing need for smaller production runs, more customisation and flexibility for design changes, fasterturnaround times and reduced inventory requirements.
“Our stated intent is ‘Helping customers redesign their textile supply chain to be closer to the customer, and be very efficient’, chairman and CEO Palaniswamy ‘Raj’ Rajan, tells just-style.
“Right now people move things around because labour is expensive. But with our line, that equation now changes.
“The vision I paint is if you have a cotton field, a yarn factory, and a fabric mill next door, and you make your T-shirts next to it, that’s the most efficient supply chain.”
“So if we can enable that, we can reduce the inventory cycle and the turnover cycle; we can ship on a weekly basis instead of
having to ship three or four or six weeks across from a distant supply chain. So we’re bringing efficiencies – and sustainability –into the market.
“The other way of bringing efficiency is to have the finished fabric, but not make the final product until I get orders from my customer. So you can now put this close to where logistics hubs are worldwide, where you can ship very quickly and a customer gets it in a week.
“Those are all possibilities that we enable with our technology that are not doable today.”
Automated robotic worklines
The Atlanta-based business has spent the past eight years developing its fully automated robotic worklines for complete garment production, first at Georgia Tech and subsequently on projects with the US military’s Defense Advanced Research Projects Agency (DARPA) and the Walmart Foundation, securing a US$2m grant to look at how robotics could enable clothing to be made competitively in the US. CTW Venture Partners, a venture capital firm founded by Rajan, has also invested in the business.
One of its first projects was on constructing jeans. “We did the first operation for jeans completely autonomously, we didn’t touch anything,” Santora says. “The material comes in one side as cut pieces, we get all the sew data out of Gerber’s Accumark files, so that when the designer designs the good in 2D or 3D, the sew data is embedded in the file that goes right to the robot; the robot then uses cameras to see the piece as it comes by, routes the good to the appropriate station, and then we use machine vision and robotics to map the weave of the fabric, and then as the fabric distorts we use the robotics to continually manipulate the fabric through the sewing head, just like a seamstress would.”
Commercially, the company currently has “in the tens” of its Lowry automated workline in operation, producing simple home goods like bathmats, towels and pillows. The four-axis gantry style robot automates up to ten operations including fabric handling, pick and place operations and direct sewing, and each makes more than 1m pieces a year.
Complete T-shirt production
The first fully automated workline for complete T-shirts “will go from cut pieces into a sewing
line and out the other end will come a finished T-shirt”
Building on this, the next commercial launch will be a fully automated workline for complete T-shirts – which should be available by the end of next next year – and “will go from cut pieces into a sewing line and out the other end will come a finished T-shirt,”Santora says.
“We picked T-shirts because everyone wears T-shirts, but 97% of all T-shirts made are exported…that means they are made somewhere they are not consumed. The US in particular is the third-largest producer of cotton, it has the cheapest cotton in the world, it is the highest quality cotton in the world…and yet we make no T-shirts.
“So for us T-shirts are this linchpin product that shows how unsustainable the supply chain is and how, when you have all these goods in an area, why aren’t they being consumed and then made into a product?”
Rajan adds that the company’s first full apparel workline will be able to carry out between ten and 20 operations, and while it would have been possible to focus instead on products like sundresses and women’s tops with the same technology, “the volume of products is higher per T-shirt. With 11bn T-shirts sold worldwide every year, this means it’s not just for the US market; there’s potential to apply the technology in multiple geographies in the world.”
Combining machine vision, robotics and advanced computing, the T-shirt line is based on different components of the Sewbot family. High speed vision technology ‘sees’ the product and material “just like a seamstress,” counting the individual threads in a fabric and their intersections to control and guide the sewing operation at the needle. Budgers are individual components of a 360° conveyor system that work in unison to move fabric panels in any direction across a surface, while Budger Balls move in different directions to manipulate or stretch the fabric in different ways.
There are also Pick & Place capabilities to mimic a seamstress moving the fabric with her arms and fingers, enabling a number of different operations to work in real-time.
According to SoftWear Automation’s T-shirt tech pack, the 13 operations that make up a single T-shirt – from quality inspection to heat transfer, collar, label attach, seaming and hemming – currently take 10 operators 348.6 seconds or 5.8 minutes to complete.
However, it calculates that one operator overseeing its T-shirt workline can carry out the tasks in less than half this time: 161.2 seconds or 2.69 minutes.
Furthermore, if the current output per eight-hour shift is 669 T-shirts, SoftWear Automation’s output is almost double this at 1142 T-shirts. “Using our T-shirt work line, a single operator can produce 1.25m shirts per year,” it claims.
With still another 18 months until the workline’s commercial launch, Rajan says the time is being used to ensure that when the system reaches the factory floor it will be reliable enough to run 24/7/365, “and could be running for the next 100 years or more.”
He adds: “We want to give ourselves enough time to experiment, so that when the product is shipped to our customers we want them to be delighted. We want it to be proven to work. Because it is going into factories, we are very mindful that production volumes are being maintained.”
He also explains that while all the individual components work, “we’re not yet full apparel end-to-end.”
There are “a lot of variations and unknowns,” especially when it comes to fabric, that have to be addressed. In addition, creating the most efficient production system may also require re-engineering or re-imagining how an operation is done.
While the challenge of getting robots to handle soft, flexible fabrics with the dexterity of the human hand is one of the most enduring reasons why automation and robotics have been slow to take hold in the sewing room, Rajan says SoftWear Automation is not following rivals who have gone down the route of stiffening fabrics prior to sewing.
“We don’t do anything to the fabric, because our view is that fabrics have coatings and finishes to give them a specific look, so if you stiffen it and remove that it changes the property, it changes the finish. So we have taken great care to make sure we can work with fabrics in their natural state.
“We’ve extensively worked on keeping a gentle touch on the fabric. That’s another reason why our technologies have taken so long to develop, because there’s a lot of thought that’s gone in on how to re-imagine completely out-of-the box.”
Interest in the T-shirt workline is already high, with three orders secured from US-based companies. The system is also on the radar of some overseas manufacturers who want to diversify their production capabilities in the US, closer to both product designers and large centres of demand; as well as some overseas producers who are looking at ways to offset rising labour costs and labour scarcity.
While the company won’t reveal the cost of the T-shirt workline, it will say the return on investment “for pre-orders that we have received for customers in the US is around two years.”
It is also taking pre-orders for automated workcells for jeans hemming and binding.
While fully automated jeans lines are in its sights, these garments involve some 30-40 operations. “We can do it in a lab, but we haven’t tested it 24/7/365 in a factory.”
At some point, Rajan says, “we will get to a dress shirt, which has 78 steps to make with pockets.”
In a bid to accelerate the uptake of its Sewbot worklines in the sewn products industry, the company last month named Dr MikeFralix, the president and CEO of Textile Clothing Technology Corporation [TC]2, as its ‘technology evangelist’. Among his tasks willbe to help identify key partners to support and enable more local manufacturing.
Impact of automation?
While the new automated technologies still have some way to go before they impact reshoring apparel production back to countries like the US, there is already debate about the potential implications for jobs.
A report published last year by the International Labour Organization (ILO) claimed the use of sewing robots in destination markets such as China, Europe and the US posed a significant threat of job displacement.
“We do create jobs, but it’s a different kind of kind of jobs,” Rajan responds, pointing to the potential to boost employment in distribution, ecommerce or retail “because there’s more product being manufactured. Or if I’m sourcing more materials locally that means I’m creating more jobs in fabric construction or in farming, so you create a multiplier effect. Then you create a service economy because you’re creating goods and service industries around to support it.
“Technology has to move forward; there’s anxiety about job losses and transition, but anxiety is short-sighted. We have to think forward, and then we have to figure out how to transform the economy in other ways, it’s re-education, re-training.
“One of the things that keeps cropping up in our industry is exploitative labour, so how do you solve it? Exploitative labour gets diminished with automation.”
To view the article in its orginal format, head to Just-Style.com.