Sewing robots

In the July 13, 2006 print edition of the Economist, I found an article called Closing the circle: Automated tailoring comes closer. The article mentioned a collaborative effort known as Project Leapfrog. The goal of Project Leapfrog can be described as an effort to automate sewing, replacing human operators entirely. According to the article, the first goal has been to figure out ways to hold and manipulate fabric without damaging it (one would hope a second goal would be handling it so it is sewn correctly). Toward that end…

Dr Molfino is trying three approaches. The first is straightforward: vacuum suction pads. The other two are more unusual. One employs a thermoelectric phenomenon, called the Peltier effect, to cool the graspers handling the cloth (and thus the cloth itself) well below zero by passing current through them. Cooling cloth this way makes it more rigid and easier to control…[the] second unusual approach involves flat surfaces that grasp the fabric they are manipulating much as geckos stick to ceilings. Geckos’ feet are covered with microscopic hair-like projections that stick electrostatically to anything they touch. Dr Molfino plans to employ the technology used to etch computer chips to create similar surfaces on her graspers.

The article goes on to say that once the fabric handling problem has been solved, the next problem to solve is how “to sew it into the right shape”. Apparently, Dr Molfino thinks the fabric pieces can be vacuum attached to a mannequin -that will change size and shape according to the garment being made- and garments will be sewn with “robotic sewing arms buzz around stitching the pieces together, much like the spot-welding machines found in car factories”.

Robotic sewing arms buzzing around a vacuum expandable mannequin like spot-welding machines in car factories? This is one of the most ridiculous things I’ve ever heard of. Assuming it were ever possible to make something as simple as a modified tee shirt -something that didn’t need stitching, turning and then stitching again- on a vacuum powered dress form, just how would this represent sufficient cost savings to justify its development? I don’t see it.

It seems like a real big waste of money to me. I went and looked at who all was involved in Project Leapfrog and it’s a very intimidating list of people who are a whole heckuva lot better educated than me, to say nothing of having more money but good grief, I think this whole thing is a waste of money. I can’t figure out why they’d spend the money. Even if it were possible, I don’t see how they’d ever get their money back. And assuming it really were possible, that would mean we’d have to start designing specific to the equipment. You know, teaching to the test? Clothes would look more templated than ever. You wouldn’t be able to do any kind of variation in design; it’d be out of tolerance for the machine; you’d have to match the parameters. Can you imagine one of these things even sewing something simple like a skirt gathered into a waistband with a set zipper? A pre-gathered skirt plastered to a vacuum packed dummy? Just how will those robotic arms apply the peltiered piece in even proportions to gather it up?

The emperor is wearing no clothes. It really irritates me when money is wasted on useless projects like this. They could be feeding starving children instead. Or am I the one who’s nuts?

Get New Posts by Email


  1. Adil Sardar says:

    These inventors and investors probably believe that mechanizing clothing production is going to make it cheaper to sew clothing with robots than humans. Thats there bet. Perhaps eventually sew better quality clothing than humans. The end result is greater production and the resulting higher profits when selling to the big box retailer (walmarts etc…), where huge number of cheap units count.

    I don’t see this is a waste of money. There is definately a risk involved but that just part of investing in a new technology venture.

  2. Jess says:

    It seems like that kind of technology would be cost prohibitive for most companies and I agree with you that it seems like a waste of time and money. I would imagine that every new style would need some kind of software that would need to be programmed. We think clothing is turning into basic boring sacks now I can only imagine how boring clothes will become if robots take over, lol.

  3. jean says:

    R & D is risky. In the early 70’s I worked for a company involved in R&D for CAD Systems. The company I worked for was not successful but eventually, there were successes, as we know. We also know that the successful approaches have been “computer-aided” as opposed to completely “computerized” but that determining the successful balance of automation and human intervention took many iterations of trial and error.

    This attempt to further automate clothing production will be no different. (I say further because we have already automated spinning, weaving and knitting. Knitted sock and glove production is pretty well automated while undershirts have only the sized tubes of fabric automated.)

    I tend to agree with you that the all encompassing automation envisaged by Project Leapfrog seems to be a bit much. However, the lessons learned will likely lead towards some kind of increased automation in clothing production. At least they are searching for approaches that might reduce waste by making reducing the cost of custom-sized clothing.

    I believe that the automated clothing production would involve the acceptance of different kinds of clothing. The re-styling and automation of uniforms for police, military etc. seems to be a better first application as I see it. I also think that there is much more potential for automation in production of knitted garments as they can be shaped by stitch increase/decrease and changes in tension.

    Just my 2 cents worth.


  4. Carol Kimball says:

    Nit-picking trivia:

    “Geckos’ feet are covered with microscopic hair-like projections that stick electrostatically to anything they touch.”

    Electrostatically, hell. Their little “projections” stick into and grab crevices in the surfaces (often stucco) where they hang out. Or down from. Mechanical. Something I’ve always admired about geckos.

    About as competently researched as the rest of their concepts.

    I am sure robotically-engineered clothing construction is something we will see come to pass. Probably not these guys, and probably not for a while yet.

    When it does happen, is this type of product anything any of us are going to want to buy and wear? They will surely find a market. It (mostly) isn’t us, and it certainly isn’t our client base. There will be enough people willing to pay for compentently engineered garments to keep us afloat. As always, the challenge is to find them and satisfy them.

    I do admire the mind(s) that came up with, “chill the cloth to stabilize it.”

  5. Georgina says:

    The research is very important. You never know where it will lead us.
    I disagree completely that automation will make inferior products. On the contrary, automation usually increases quality and uniformity. Increased technology makes better products available and more affordable to more people. Think of knitting machines. Imagine if we didn’t have them, and we knitted every sweater by hand?

  6. Marguerite says:

    I’m surprized by the approach, I would think it would be far easier to design clothing woven all in one piece with a computer dobby loom.

  7. Jane says:

    Visionaries have traditionally been viewed as crazy. But hindsight is alway 20/20! Look at people like Eli Whitney, Elias Howe,(invented the sewing machine) Thomas Edison, the Wright Brothers, Henry Ford etc. I think not only will the way clothing is produced will become more automated, I also think what we wear will change in line with that. The comment made by Jean that much of knitted apparel is already automated will drive more and more of that style garment into ready to wear. Look at the space age looking active wear Nike is producing; Color blocking without traditional seams, raised embelishments that are woven in with the fabric, etc.

    We may end up wearing stuff that looks like the uniforms on the Star Ship Enterprise after all. Geez, my hips already look huge in cotton/lycra blends…..

  8. jinjer says:

    I agree with Jean about the value in the research being the lessons learned on the way rather than the acheivment of the stated goal. A rather major assumption behind the research is that consumers will contunue to want clothes that are produced on an extremely massive scale. This is a gamble.

    As a former biotech geek, I can’t help but lump this in the same category as genetically engineered food. When the research began, conformity & yield were the most valued properties in food, and the greatest advances in agricultural productivity came from mechanized farming (aka technology). It seemed like a smart idea to explore the next technology that could increase productivity & uniformity in the same vein. By the time the research was bearing fruit (tee hee), the tide of consumer opinion had started to turn, and the resulting backlash against “frankenfoods” destroyed some honestly valiant and noble research efforts (like golden rice…sigh). Now, the hottest things in agriculture are organic foods, heirloom varietals and farmer’s markets/buying local. Even Walmart has gotten into the act. yay agricultural diversity!

    In other words, the pace of research like that is so slow that by the time it pans out, people may have returned to home sewing, or one-of-a-kinds may be the hot thing….leaving fancy, over-specialized machines to rot in a basement. Still, they may learn some really great techniques along the way to help human operators make quality clothes more safely and efficiently–who knows?

  9. Dave says:

    Can sewing be automated? I’m pretty sure almost any process can be developed and automated. But affordably with sufficient quality and flexibility to recover cost is a big question mark , for now. I currently work in aerospace (but my heart is working towards a DE). CNC (Computer Numerically Controlled) machines have their advantages, especially in lean manufacturing environments. These machines have the ability to hold over 200 different tools and can change working from one part to different type part (and multiple operations on each part feature) in seconds.

    Some garments may lend themselves more likely to be processed. Materials that are more rigid are more likely – maybe not all steps would be automated. As it comes closer to reality it will be interesting to see to how fabrics will be developed to take advantage of the automated processes. This is just speculation based on how I’ve seen processes and condition of materials change … they usually change in tandem. Materials as we know them now may change.

    For example: During the manufacturing the fabric may arrive much more heavily sized/starched so that when heat/pressure is applied to it and shaped over a form to make it semi-rigid so the pieces can be fitted together (e.g. seams could be at some angle so they are brought together in such a way that are more readily machine accessible). “Invisible” dots/lines that the machine can see (placed on the original material or be the cutting forming process) will allow to align fabric bias, design alignment, etc. A new type of sizing may be such that if cut with a laser there is a small heat affected zone say 1/64th inch that prevents edge fraying and reduces the need for serging (you‘d be surprised how small the heat affected zone is on metals cut with a laser and cloth lacks the conductivity metal does). The garment is then washed and whatever else for final packaging. Maybe only ortions of the process will be automated.

    Will the consumer like it? I don’t know.

    I agree I don’t see how it will be able to sew on an adjustable form described AT THIS TIME. But then again Boeing’s new 787 Dreamliner is mainly carbon fiber and I’ve seen machines weave complex shapes that become the final aircraft parts. They have a machine now that can nearly duplicate how a person can rotate their wrist when polishing a part in tight areas. I was amazed the first time I saw it doing one. The process on a per part basis takes about 50% longer – but the process is way way more consistent and scrap is now unheard of. Of course the machine doesn’t take a break and works nearly 24/7 so it’s daily through put is much closer so that overall it produces a higher quality part at a slightly higher cost – and contrary to popular belief/propaganda sometimes quality costs more, it isn’t always free (as in process improvement will recover all the costs)

    I don’t think sewing will ever be fully automated especially at most of the DE levels that have smaller productions runs, the cost of automating is too high. But you may see develop are companies that do short production runs for numerous people. Job shops.

    Hope this didn’t bore y’all but I work as a Supplier Quality Assurance Rep for commerical and military jets engines. Myjob is to drive process control, process certification, and lean manufacturing, I help manufacturers look at how things are done and how improvements can be made. (Particularly for dependability … I, my family and friends fly a lot so I have a very vested interest in making flying safe).

    Who knows, maybe in the future people will be looking for “work shop” label so they know their item was made by hand.


  10. joni says:

    For all the right reasons, we should be rid of the sweat shops that house “live” robots already. Real people need and thrive on real work…is mechanical automation always the answer?

  11. graham says:

    I’m willing to bet CNC sewing of high value items will happen much sooner than you think.

    It’s pure economics. If you can reduce wastage from the production process, you will be profitable. When working on mass produced jackets $35 a meter fabrics, it behooves you not to have somebody screw up your fabric.

  12. The major garment manufactures all support an organization called TC2. TC2 was originally created 30 years ago and funded by the U.S. government for the purpose of creating complete robotic production lines in the garment manufacturing processes in the U.S. Up until about ten years ago when more production was moved oversees there was no longer a need for this study because labor became so cheap. They then shifted focus to 3 dimension body scanners for size and fit studies. The first generation can be seen at the Brooks Brothers store on Madison Ave in New York City. They are one of the leaders along with Gerber in taking their sizing scans directly to electronic patterns that then feed straight to the robotic cutting table. TC2 is also on the cutting edge of digital print technology, and provide extensive studies on lean manufacturing and TPS. You can find more information on them at

    The study and use of robotics is prevalent in undergarment manufacturing and is progressing into more complex patterns with reasonable results and quality.


  13. Chaitanya says:

    i am a fashion technology student of National Institute of Fashion Technology – India.I have read about the research carried out for placing robotics on garment production floors. I felt happy , on regard of technology advancement.

    But one thing I request every one to think up on is”Where will all those Sewing and other operators working in this field go? ”

    Do we need this technological advancement? Don’t we feel that “Only big game players will be in a position to use this technology”. And this even increases the cost of production!!!

    From consumer point of view, if a consumer is ready to pay that price (the price of a garment that is being sewed by robots) , he/she would as well go for designer wear.What do u say?

  14. Jean2 says:

    It seems to me that the main idea behind an automated process is mass customisation, while reducing cost.
    I’m a garment wholeseller, and our biggest problem is over stock.
    We sell in small quanitites as low as 100pcs with mixed sizes, but we need to buy in advances different size of shirts like S,M,L,XL,XXL and we much of the time end up with over stock.If we could manufacture automatically small series on the go, while keeping cost down, we would be able to offer mass customisation of our garments.
    I also see teh point where manufacturing in China or indenosia is becoming more and more expensive, with little alternatives.
    With automation, manufacturer could go to EU or North America, and produce shirts there for example, it will also bring jobs back, as you still need someone to supervise the machines and operate them, even is not about intensive it is still a very good prospect.
    As for making skirts with waist bands, there’s will alwys be needs for skilled low cost labour to make operation that robots would be able to do, so high street fashion and women fashion isn’t in danger I think.

  15. Dennis says:

    Soon I can visit my local taylor, have my body scanned, have a manekin made, and then have some fabric material sprayed on to it. I could have the shirt or pants within an hour. Yipee!

  16. David S says:

    Nit-picking trivia:

    “Geckos’ feet are covered with microscopic hair-like projections that stick electrostatically to anything they touch.”

    Electrostatically, hell. Their little “projections” stick into and grab crevices in the surfaces (often stucco) where they hang out. Or down from. Mechanical. Something I’ve always admired about geckos.”

    (yeah, I realize I’m years late…)

    Nope. Geckos’s seta (the hairs on their feet) allow them to climb amazingly smooth surfaces, using van der Waals force, which is indeed an electro-static force. people are hard at work trying to make glue and tapes taking advantage of the effect, but nothing is yet available commercially.
    There are interesting garment possiblities, though: imagine a better velcro, that’s invisible or nearly so, doesn’t add an extra set of layers to the closed area, and is silent to open or close. I’m sure there’s more, that’s the only thing that occurs to me without thinking about it.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.