The First Industrial Revolution sparked upheaval in all sorts of industries, of course, but few experienced greater change during those smoke-filled decades than textile manufacturing. The spinning frame, the spinning mill, the Spinning Mule, the spinning jenny and numerous other then-novel contraptions changed the way our predecessors worked. They remain recognized, if not all that relevant, centuries later.
Which is why, during this Fourth Industrial Revolution, the work that SoftWear Automation and CEO K.P. Reddy are doing — automating the sewing process — feels so appropriate.
“If you look at CNC machines and other devices, they’re cutting or at least working with some rigid material,” said Reddy, an entrepreneur by choice and an engineer by trade. “The challenge with clothing is that it moves. It flexes, it stretches, it’s hard to rigidize.”
In a growing market, the automation of sewing fabric allows for smaller, more specific orders with more precise details (the same principles of custom additive manufacturing), more quickly. Just like the jenny revolutionized textiles 252 years ago on its way to inspiring change in all industries, this automation process today could lead to upheaval in other, unexpected areas.
IndustryWeek:Where did this idea come from? How did the whole process start?
K.P. Reddy:The IT and technology was incubated in Atlanta, at Georgia Tech. Apparel and the industry is a huge market, and it’s very backwards technologically, a lot of the manufacturing is normally done in third-world countries. We turned to computer vision technology to actually track the fabric at a thread level. All of our IP is based around tracking these threads so that we always know where the needle is relative to a unique thread intersection. The system knows when the fabric wrinkles, and it straightens it out and makes adjustments. It doesn’t require clamping and manual automation, so to speak. It’s all very fluid.
Automation has gotten to the point where a lot of the off-the-shelf technology prices have come down, the processing power has come down, and you can use $300 cameras. Solutions are more in line with the economics of the industry.
IW: What was the process of programming the robots to deal with fabric? Engineers writing code, figuring out a flexible process, all that.
KPR: A lot of that core IP was thanks to a DARPA grant. In the U.S., all military uniforms have to be made in the U.S. While I was at Georgia Tech, the company received a grant to figure out a better way to make and alter uniforms here in the U.S. Cost was a factor, as was the talent pipeline drying up, with a lot of top seamstresses being in their 60s. That’s where the idea came from, and in solving that problem, it became clear that a seamstress is not just about placing the buttons. There’s a touch, an understanding of the fabric.
We needed to understand how that touch related to manipulation of the fabric by automated systems. The best way to do that was through computer vision.
IW: Is that proprietary? Or are you working with another company and using their software?
KPR: It’s all proprietary, yes.
IW: What was the timeline to develop the software?
KPR: The first DARPA funding came in about four years ago, and then we invested about a year and a half ago. That took it from the commercialization process to building a company.
Our biggest market in the near-term is people who want to place orders for 50 to 100 robots. In the world of mass customization it might be three to five robots. That’s a market that’s very interesting because of the complexities around e-commerce and the supply chain. Our vision is for online ordering and when you hit buy, the robot starts working on it. The world of mass customization is a new market for us, but it’s very interesting, the chance to work with and teach smaller designers.
The big markets we’re working in right now are contract manufacturing and brands that have to do their own manufacturing, which is mostly mass production. Some of that is in the U.S., some of it is pilot projects. There’s been a surge of foreign interest — China, India, Bangladesh, Sri Lanka — from companies that want to automate because they believe that if they don’t automate, they’ll be left behind.
Robots for Any Task, Robots for Every Task
IW: What sort of fleet are you working with now?
KPR: We delivered our first product in October (2015), and we’re currently around 20 robots. They’re going out there. One of our customers tried it out for a couple months, then expanded their order.
IW: Is there any ceiling right now on how big you think it can get?
KPR: Being CEO and investor, I have feelings about it, right? But it’s a $100 billion a year marketplace, there’s very little competition in our space — partly because of the technical problems we’re solving — and we do think it’s a big market. As bigger brands have to keep more SKUs, we’re seeing more demand. The robots are also allowing designers to design things that previously would have been too complicated, too specialized. We like making things that only a robot can make in mass quantities.
IW: What are some of those items?
KPR: Anything custom-fit falls in that category, anything with lots of curves or complex patterns. It’s difficult for companies to contract that out. We have one company we’re working with right now that’s actually assembling 30 different pieces and parts to make the product — I can’t say what it is — that would have been a lot of manual labor.
IW: The software is in-house and proprietary. What about the hardware?
KPR: Initially, we made some of the robots ourselves. Now we’re working with other providers; we’ve looked at Festo. What’s proprietary to us, we have a few hardware products — our automatic sewing machine, our ASM — but we focus more on designing software, so we can buy a lot of products off the shelf in terms of robots.
IW: Why did you stop making your own robots?
KPR: That was early in the company’s development, when it was still being highly funded by government grants. Culturally, coming from academic research, there was a lot of thinking of We could build this better, but it was a matter of being cost-effective and scalable. By partnering with our supply chain, we’re able to have a much higher capacity. Maybe we give up a little functionality, but we offer more serviceability.
IW: You have about 20 years in tech, in various areas and interests. If we broaden the topic to robots in manufacturing over next two to five years — big, big window — what else do you see or, rather, what do you not see?
KPR: Right now, everybody seems to be building robots for very specific functions or tasks, and every time a robot comes around that’s more broad, that can do anything, like Baxter, there are issues. I think you’re going to see happen in robotics and hardware exactly what’s happened in software, which is this idea of interoperability. How do I mix and match from multiple manufacturers, do multiple tasks and deliver a unique product?
I have some cloud investors. The beautiful thing about cloud is that it’s your own device and you have the ability now to interoperate between different platforms without having to hire a consulting firm. It just happens. The next wave in manufacturing is going to be all about interoperability. … All the machines need to talk with each other all the time. They all need to work together.
IW: Will tailors still be relevant in a decade or two?
KPR: We think today’s tailors are far enough along in their careers that their jobs will be secure. None of this stuff happens overnight. The people who are earlier in their careers and want to stay in this industry, they will all be repurposed on the technology side. If you look at what happened with ERP and accounting systems, there used to be droves of data in the payables department. In private industry, most of those people got repurposed as consultants or learned the tech. People will find jobs working with these robots, as well as doing other things.
IW: The irony of this, at least to me, is that looking at photos of you, you’re wearing button downs, slacks, Adidas. You don’t seem like the kind of guy who really needs something custom-made. How did you get into this?
KPR: Pretty much everything I wear is custom-tailored. I do have a fundamental philosophy that what we wear is the single biggest expression of our creativity. We can all go to Brooks Brothers and buy the same shirt. I’m a proponent of finding things that are unique to my needs compared to everybody else’s needs.