I think the soles for this are done by Carbon in Redwood City.
They've done some other cool stuff with their 3d printed lattice like the midsole of an Adidas shoe that's already out and some bike saddles (I think for Specialized and another brand I can't remember).
Carbon3d is an under-appreciated tech. Their materials are incredible. A few years ago, we successfully used their elastomeric polyurethane to build a soft robotics gripper ([1], [2]) and we've got an impressive durability rating of 500K cycles (basically, months of continuous grasping).
While our prototype never got into a production, it was very clear that Carbon3d has a uniquely strong proposition because of their materials.
It's not as unique as it used to be. Neither is their DLS printing technique. They also have an insane valuation [0], so I'm not sure if they are really that under appreciated.
For example Origin [1], which was just bought by Stratasys, is very similar in a lot of ways.
Formlabs printers can do a lot of this stuff but at a fraction of the cost (base model costs around 2k, also has an expansive list of materials with interesting properties: https://formlabs.com/material-selector/ ). Also, you don't own Carbon3d, you lease it, for like 50grand a year or something crazy like that.
We tried Formlabs, of course. We could print fingers like these, but none of the materials we tried (about 10 different options) survived more than 1000 cycles.
As I said, Carbon3d underappreciated, because everyone think "it's just a more expensive Formlabs", while in fact Carbon3d is all about production-grade materials.
If it was possible to take materials from Carbon3d and then print on Formlabs, that would be a very good combo.
I believe the big difference is that the Carbon printers use a special bottom plate that lets air molecules passing through. That's why they can use chemical reactions to create better material properties.
This also means the print doesn't stick to the bottom plate and they can print faster.
Agreed - not just materials either, Carbon printers are fast and for mass production. Making shoes via form printers at scale is not viable (form is nice for prototyping though and small scale stuff).
You might be thinking of Fizik: "The Adaptive saddle padding is crafted by Carbon using its revolutionary Digital Light Synthesis technology" https://www.fizik.com/eu_en/drops/adaptive.html
Adidas is really leaning into a lot of new tech, having just recently announced Stan Smith Mylo too, shoes made with mushroom based leather: https://www.adidas.com/us/blog/663481
like the boosts, i would never consider training in these due to the lack of support. if i were racing i would wear spikes. they look sick though and i would absolutely consider wearing a pair just as regular sneakers. i'm surprised that shoe companies still have to pretend that runners are their target demographic for cool shoes like these, when in reality it's obviously for sneakerheads / the general public who would like a neat and comfortable pair of sneakers.
If they're sold, they will be collectibles (there's a big sneaker market out there where shoes are traded up until some rich enough guy buys them to wear once).
For day to day use, there's $30 Nikes that are fine.
These 3D printed soles are already in mass produced shoes sold in adidas’ website. They are not collectibles anymore, just shoes with a different kind of sole.
Overall I believe the industry of sewing clothes is gonna be the last to be fully automated. The 3D manipulation of 2D fabric is just too finicky and screams for the human touch, experienced, versatile fingers. But starting off 2D, mass produced fabric is just so much more efficient and versatile.
Sadly, this is bad news for the future of human freedom, as it will mean the associated exploitation and environmental impact isn't gonna end soon either.
However, I have this hypothesis this problem can be solved with software otherwise, by removing the complexity and hurdle of custom pattern creation and distribution to the typical end consumer.
You can't make your own fridge, smartphone, and even wood working is out of reach for most people in the city (space, logistics). However, everyone physically capable could in principle sew their own clothes, if it's just down to the sewing part. Traditional pattern creation and/or application is the hard part. Space needed, investment in the sewing machine and fabric, time expenditure... It's all relatively feasible to most.
I think having a FOSS variant of Clo3d, UI optimized to match different experience levels and a github like collaboration network would offer the potential to kill a good chunk of fast fashion, complex supply chains, and consequently ecological and humanitarian exploitation.
Imagine using a camera or lidar scanner to measure your body, load that to the software, have the design specification applied to your personal body model, the pattern projected onto the fabric for cutting, or even laser cut, until all is left is sewing the pieces together according to the community improved instructions.
No more compromises in fitting. Things will last and be cared for, because you made them yourself. Fewer better clothing.
Or delegate this to your local sewing service, which operates under the same opportunity of reduced overhead.
Things just need to be FOSS to have this change the world. Clo3d costs 500€/month minimum... But really it's just blender with extra steps. It can be done.
> The 3D manipulation of 2D fabric is just too finicky and screams for the human touch, experienced, versatile fingers.
Maybe asking robots to do the human thing is not the right approach. Just like dishwashers don't do things resembling how humans approach the task. Which implies the clothing would have to change somehow. Is that even possible? I don't know.
You got 3D woven clothing already in stores. But I think there is a fundamental limit in form and function. Clothing isn't just means to keep you warm and protected. Aesthetics matter.
I think the loom is already the dishwasher here, where sewing is collecting used, unordered dishes.
I think these tasks need a deeper understanding of what's there to handle. Prediction how things will behave and adapt intelligently if they don't. E.g. spontaneously using your pinky to stretch away the fabric getting in the way. You need to understand the meaning of the seam to manifest it, even out tension, redo some part, ... Non petrol based fabric has imperfections and variances.
I think you need general purpose, human-level AI to sew a good jeans.
> The 3D manipulation of 2D fabric is just too finicky and screams for the human touch, experienced, versatile fingers. But starting off 2D, mass produced fabric is just so much more efficient and versatile.
3D robot woven fibre composites were a thing for more than a decade.
Before you said it wasn't possible. Now when someone points out it has been already done for a long time you say "severely limited aethetical (not a word) expression".
Where did I say it's not possible? This very article is about 3D printing of clothes. So someone pointed out 3D woven clothes exists, probably didn't really read my comment either. I too had 3D woven clothing in mind in my original first paragraph, when I emphasized the aesthetic benefits of 2D fabrics, over direct 3D fabrication.
What's your point, really? Is what I said invalidated, because you assumed some inconsistency off topic? Do you think that's constructive and helpful to discourse and exploring new ideas?
Yes. That's not 3D woven, is it? Robots aren't sewing jeans, but they can weave some variation of tube like socks or a simple pullover. Anyway, I think you are just trying to fight and feel superior over some pity. Hope you get better. Good night.
Thats 50/month, but ok it's not for everyone.
Designing cloth is not something everyone can do and it takes practice and some are more talented at it so it makes sense to pay these people for their work.
More customisation is also harder to scale, if everyone start making their own color combination the knitting or weaving machines need to be re-threaded, which takes time.
Printing on fabric requires not just a printer but also an autoclave to cure the paint.
I'm actually in the middle of making a clothing CAD, and I would love to be able to make my own stuff but scaling so it's not 500€ for a tshirt.
I didn't mean to imply to tackle the fabric industry with this. Just the sewing. I think fabric is still better off mass produced by long running machines and then you design around what's on the market. Much like you would do with wood. This is all compromises, but I think fundamentally sewing is the most approachable craft to not offload to a specialized industry. It's the pattern issue and convenience at heart, I think. As I said, much harder to make your own fridge or couch, than a t-shirt or dress.
You are correct about the pricing it seems. Either they changed that, or I misremembered. Anyway, 50€/m is still way too much for hobbyists.
I think you missed the point. It's exactly the answer to "not everyone can design clothing". You download the design specs, the software does the calculation and application, and then you cut the fabric. No designing, no following and fitting lines on large paper patterns... All this can be done in software.
Either "poster print" on multiple A4 papers, overhead projection onto the fabric, or laser cut. That's what I had in mind. I imagine an optional small helper device with lidar scanner and mini projector to measure the body and help with the pattern transfer. Both lidar and mini projector modules are becoming really cheap, as you find them in cheap chinese phones already.
The problem is that there is a lot of of variance in preferences. Clothing etailers have the same problem where some people are used to it being snugger in different areas. Folks who share dimensions may be used to different fits.
It's common to shape the sole of performance cycling shoes, but not by way of scanning. I had a pair done a while back which where made by me stepping on a heated rubbery platform that slowly took the shape of my feet. The thing I stood on was then used to vacuum-shape a carbon fibre insole.
I've had this done with some custom birkenstocks before. They heat molded the cork insoles to fit my feet. They were very comfortable and well worth the money and time saved from not having to break them in.
Hockey shops are 3d scanning feet, but they don't have individualized boots to go with that; it just gets you a shoe size and one of three fit profiles within that size.
Better ski touring (and even just alpine skiing) boots have inner soft parts that are heat-moldable to fit perfectly to everyone's unique feet. Ski shops all have these heat machine specifically for this, but there are tutorials on how to even do it at home with just rice, socks and microwave in case you shop online (with almost no risk of destroying it as usual)
Most custom-made clothes are made with machines in most operations. The solution to clothes that do no fit and are low quality is getting higher quality clothes that are either bespoke or tailored. This is a solved problem.
"we're targeting fast runners, 5:30minute mile runners"
way back in high school, i was a 5:45minute miler, and hit 5:30 once. it was not fast, at least fast enough to win. sure, 5:30 isn't a walk, but maybe my views on the mile speeds are skewed because i was in it.??
Yeah, pretty sure they were talking about longer distance runs. Mile/1500m events are typically run with spikes... this stuff looks cool, but pretty sure they'd still lose out to similarly lightweight spikes.
They could but in this case “spikes” doesn’t just mean the pointy things it refers to a specific design of shoe
spikes (the shoe) eliminate pretty much all the padding and weight. And trade it for a rigid plate designed to put you on the ball of your foot for speed. It’s a trade off optimized for speed
I was going to comment about that as well. If you're doing long distance like steeple chases or cross country where it is on turf, spikes are okay. I personally didn't like them, and preferred the cushioned distance shoes.
like i said, i was into it, so that was my world and definitely skews my reference. so when my 5:30s couldn't compete, my reference says it's slow. interesting to see that data graphed like that. my personl best for the 5k was just over 18mins. of course, now, it takes me that long just to lace up the shoes properly.
Was working on a small scale startup trying to do this a couple of years ago using more DIY-style 3d printers and flexible fillament:
https://dialectic.design/project/solemaker
Very impressive to see Adidas committing to this concept with real industrial robots and printers for a couple of years already with the Speedfactory and this etc.
you wear knitted jeans? of all of the clothing items to refer to as unraveling, jeans would not have come up in my mind at all. sweaters on the other hand...
The human foot is a strong and complex mechanical structure containing 26 bones, 33 joints (20 of which are actively articulated), and more than a hundred muscles, tendons, and ligaments.
I suppose you are thinking "Why should I care, those Walmart sneakers fit me just fine" but if you had to pay podiatrist $400 for custom orthotics every year this article would get your attention.
TFA says the first applications are for elite athletes.
Pravda says this is potentially a way to free the poorest paid workers of the world from some of the most difficult repititious labour.
Isaac Asimov says in the future lots of things will be made this way.
What will all the people be doing in the future when all the work is done by machines? Some of them will be writing novels, and editing documentaries, about how the world used to be before robots.
You can use that line of reasoning for anything, like prostitution, hitmen, drug-dealers.. Work is not necessary for surviving really, we are just used to it for some centuries
Of itself, automated production is not a recipe. It is just a factor of production. Society could be organised a number of different ways, post-automation, just as it has been organised in various ways since the end of slavery (in some geographies and forms).
> The human foot is a strong and complex mechanical structure containing 26 bones, 33 joints (20 of which are actively articulated), and more than a hundred muscles, tendons, and ligaments.
And evolved to walk to walk barefoot.
> I suppose you are thinking "Why should I care, those Walmart sneakers fit me just fine"
No, I use comfortable quality shoes. But most feet aren't that unique that a massproduced shoe that conforms to them after a break in period wont fit them.
> Pravda says this is potentially a way to free the poorest paid workers of the world from some of the most difficult repititious labour.
Do you expect millions of low price shoes to be produced by specialized machines after taking a 3D scan of your foot?
> What will all the people be doing in the future when all the work is done by machines?
I don't have anything against machines or robots, just as I don't have anything against regular juice presses. You can automate those jobs with more conventional robots. Having custom shoes printed to order is not a realistic production method for the vast majority of shoes sold. When it comes to bespoke equipment for a handful of elite athletes or rich hobbyists anything goes of course, you can already buy 3D printed bike saddles and handle bars, but this is not the future of shoe manufacturing for the masses.
That really depends on your perspective. Sure, no one needs a shoe where the design and manufacturing is broken down all the way to each individual thread. And the vast majority of people wouldn't notice a difference wearing footwear made 20+ years ago. But that doesn't mean all innovation is pointless either.
A new way of manufacturing won't have the same tradeoffs and limitations, the setup costs, tolerances, material properties. Direct injection is how Adidas makes something like 350 million pairs of shoes a year now. And they have been selling 3d printed soles with knit uppers for a couple years now. The innovation in this story is more about the design and the manufacturing process than the materials. You can't buy size 13.5, or wide or narrow(except a few models), or a pair where one is a half size smaller or larger traditionally. And beyond just the sizing, there is a lot more aesthetically when you are customizing at the individual level.
Sure, this shoe is a marketing piece, but this kind of tech could be the way forward to end forced/child labor in china used to produce the average shoe.
If Adidas can solidify the tech and pave a way forward, this could be huge.
This article is about automatic weaving of the upper[0], combined with a separate 3D printed sole. That's not done by forced or child labor today either[1], the supposed added value is a bespoke upper tailored to your feet by 3D scanning them first. That's not something that's really needed, regular uppers are flexible enough to work with different feet already, and it certainly isn't a way to produce the average shoe.
They've done some other cool stuff with their 3d printed lattice like the midsole of an Adidas shoe that's already out and some bike saddles (I think for Specialized and another brand I can't remember).
https://www.carbon3d.com