If the Shoe Fits: 3D Printing and the Future of Manufacturing Footwear

Adidas has announced that it plans to 3D print 100,000 shoe soles of its Futurecraft 4 shoe by 2018. The German shoe and apparel company is doubling down on its Future Factories initiative where it is reshoring the production of shoes to Germany in highly automated factories. Adidas’ strategy here fits squarely in Industry 4.0 and other initiatives to mass produce goods using ‘lights out factories’ using digital fabrication and automation in Europe and other developed economies. Is this the first sign that 3D printing is the future of footwear manufacturing?

If Adidas and other companies are successful in implementing these initiatives at scale it would be a sea change in global commerce. It could mean that many goods will now not be made in Asia but instead be made in Europe and other places. Or perhaps Adidas and other companies could only do this local manufacturing for some of their goods which would make the impact of this much more limited. To give you an idea of the impact that reshoring large scale production could have: there are approximately four million people employed in Bangladesh’s garment industry and this industry is responsible for 81% of that economy’s exports. We have yet to see if digital manufacturing of garments and shoes will occur at scale but Adidas is rushing ahead to be the first shoe company to implement large-scale production using 3D printing. Meanwhile other companies such as Nike, Under Armour and New Balance are undertaking their own 3D printing experiments.

New Balance and Nervous System’s 3D printed shoes.

Adidas is turning to Carbon, a US-based vat polymerization 3D printing company, to be their production partner for the Futurecraft sole. Vat polymerization refers to a number of 3D printing technologies that all build up an object in a vat of photopolymer using a laser or other light source. Traditionally these technologies (also called stereolithography, SLA and DLP) have been used for molds, casting in jewelry and in prototypes. Vendors such as 3D Systems and EnvisionTEC have for a number of years supplied the hearing aid industry with systems that make millions of In The Ear hearing aids. These custom hearing aids have supplanted traditional manufacturing methods and now are responsible for the vast majority of ITE hearing aids being produced. Instead of working with the established vendors, however, Adidas has chosen to work with Carbon.

Nike Vapor HyperAgility Cleat with Powder Bed Fusion Parts.

Carbon is an oft-hyped technology startup that solved the platform adhesion issue with vat polymerization. Typically in a vat polymerization system light from a laser travels through a screen from below and polymerizes a photopolymer against a build platform. This platform is then detached from the screen and raised. A sweeper then cleans the screen, the build platform is lowered against the screen and the laser makes the next layer. This raising of the build platform and sweeping of the screen is time consuming. In fact this step usually takes much longer than the laser does. Carbon solved this through replacing the screen with an oxygen permeable window. This is innovative and if you would like to learn more about how Carbon’s technology works you can read their paper on it here or you could read the Carbon CLIP patent here and another CLIP patent is here.  As well as this step Carbon has a number of experienced materials people on board and they have improved Carbon’s materials to become more functional. A subsequent heating step in an oven further cures the part (all vat polymerization parts have to be UV cured as well) and improves its mechanical properties.

Carbon’s CLIP Technology

The question on everyone’s mind now is: Does Carbon’s new technology have sufficient reliability and repeatability to produce things at scale? Will we 3D print all of the shoes in the future? Yes, Carbon’s CLIP technology is faster. But, how good are Carbon’s machines at producing tens of thousands of objects at scale? And, has Carbon managed to make a flexible material which is strong enough to function at length as a shoe sole through vat polymerization?

Under Armour’s 3D Printed Shoes.

The team behind Carbon has been working on polyurethane/silicone and rubber elastic parts using CLIP for a while now. They have a number of patents for 3D printing polyurethane using their two part process with cyanate ester, directed assembly and strengthening that material with nanoparticles including silica. In this case their work seems to build on an earlier 2008 paper by a Chinese team at Northwestern Polytechnical University Xi’an and other work in the field.

If these papers are correct and Carbon indeed uses a cyanate ester, liquid polyurethane and silica blend through a dual cure vat polymerization process with an oven curing step then the mechanical properties of the part may be excellent. It would mean they’ve in fact made a kind of new “thermoset elastomer” material using 3D printing. The team is however not just aiming to 3D print an ornamental part but they want to make shoe soles. This is a part that is put under continual mechanical stress. It is not a unidirectional stress as in a lab test however but under complex multidirectional stresses. If the 3D printed shoe sole part is under repeated strain will the residual stress of the previous step cause the sole to eventually shear? What if the stress is not only caused by a simple up and down movement but a continual side to side sway coupled with a vertical movement, pressure and friction?  If any shearing or holes or notches occur in these soles will they come apart? How resistant is this new material to chemical aging and how does it degrade due to UV? Does it only initially perform as well as polyurethane and degrade over time? When we walk on the street our soles are continually exposed to all manner of solvents, oils and other chemicals, how well does this material stand up to them over time? How well will they stand up to temperature shocks and will certain temperatures and stresses cause the material to elongate improperly? These are unanswered questions as of now. I’m at present skeptical that the resin material would perform for long periods as a functional shoe sole.

The Carbon M2 has a build volume of 19.04 by 26.24 by 11.68 CM (L, W, H) that means that with the current machine they may be able to print a US size 9 (European 42) shoe, but they won’t be able to make a 9.5 (43) 26.7CM shoe? Or if all the shoes are printed vertically as in the Carbon Adidas video does it mean they can only 3D print a size 4 (European 19) at the moment? Or are Carbon making a bigger machine just for the Adidas application? If so will that be available on time for their implementation? And how big exactly is that shoe we all see in the video and on the images? The one in the video on the Carbon printer must be smaller than the one everyone is holding up? Does Carbon already have a bigger machine, is it positioned to print at another angle in the build chamber or is it a prototype made of several parts? Or does Carbon only employ really tiny people?

The Adidas Futurecraft sole.

It is difficult to estimate just how fast a Carbon machine would take to 3D print a midsole. The company has been a bit vague on its build speeds just saying that they are 25 to 100 times faster but we don’t know for what shaped object and what sized object in what orientation with how much material. So let’s assume that they can print a midsole part in one hour (please note that this assumption of course changes everything but so would others). That’s 24 a day, if they 3D print for 100 days nonstop that would require 42 machines to do 100,000 soles. A Carbon M2 is $50,000 a year (there are additional training costs and you’d have to get the cleaning station and oven and a UV flash device as well as training and installation costs as well so the actual cost is far higher). If you’d run it nonstop (and print one per run) it could make 8,760 soles per year, meaning that in machine costs alone it would be $5.70 per sole. The Smart Part Washer costs $10,000 per year. The window cassette subscription adds $5,000 a year and the build platform subscription another $1,000. The UV lamp adds $2,700 and the oven $5,400. So that’s an additional $24,100 a year and these would add a minimum of $2.80 per sole. Assuming the material costs $349 per 750 mL as per their site and a shoe sole would be 50 grammes including support, then the materials cost of a shoe would be $23 (not including support or waste). So that’s $23.20 in material cost, $2.80 in leasing costs for ancillaries and $5.70 per sole in machine leasing. Comes out to a cost of $31.70 per shoe.

These supports won’t cut themselves.

In the press release they talk about designing the midsole so it doesn’t need support material but in the video the soles clearly do have supports. In fact they’ll need them to stick to the build platform. So let’s assume that there is no setup time but a worker has to remove the sole from the build platform and then cut off the supports and file them. In the video it looks like the part has a lot of supports to cut down and file down. How long would that take? Then a batch of soles would go in the washing station, then out of it into the UV curing device then out of that and into the oven. Average hourly labor cost in Germany is around $33 so let’s estimate that it takes him total 7 minutes per sole to do all of this, that would add $12.60 per sole. This is why automated support removal, automated post processing and automated handling are so important if we really want to reduce costs in 3D printing. Yes, your oven does all the work still, there’s a bit more to baking a cake.

$44.30 per midsole, as a cost per midsole. If we look at the cost breakdown we can also see that equipment leasing is an inordinate amount of the total especially since this is a low estimate and would be compared to buying equipment outright from other vendors. The material cost is also extremely high per part as is labor. If they would purchase their shoes in Asia they would cost around $26 and on a $100 shoe the company makes around $2 in profit. People may very well pay the premium because it is Made In Germany or a new cool thing. Interestingly if these shoes were all individually sized and optimized for the individual wearer then a premium would probably melt away in the face of an optimized and higher performing 3D printed shoe.

Adidas wants to position the shoe as a premium product. The manufacturing costs are at the moment higher than a shoe made in Asia with traditional methods. Also rather than buying the resin directly from Carbon, could Adidas perhaps get a nice discount if they got it from Covestro, DSM or BASF? Even if the costs were significantly higher and could not be reduced however Adidas may do well on this shoe. If in the traditional retail mass manufactured model they make $2 in profit and forgo $50 by selling to retail then selling these shoes directly online or via their stores make them more profitable for them even given the higher costs to manufacture.

In essence if we do a quick back of the envelope calculation the Carbon shoe printing implementation seems to be higher cost. But, what if having an excuse to sell directly is enough? What if by manufacturing and selling directly the company skips shipping cost and does not give any margin to retailers?

There are six things happening here at the same time.

1. Adidas is using automation to bring manufacturing back to Europe. This could be a good thing for Adidas, Europe and Germany but we must realize that there will be knock on effects should we do this with a wide array of goods. If lights out factories make European manufacturing of low-cost goods viable at scale again then this will have huge geopolitical impacts.
2. Adidas shoe implementation seems risky because the vendor and the material has not been tested and used extensively. Apart from this there are of course other manufacturing risks in adopting such a technology for the first time. At the moment it would seem that 3D printing a midsole would be much more expensive than making a complete shoe in Asia. This however does not necessarily make this a bad idea. By controlling manufacturing and selling directly to consumers Adidas may yet obtain more profit in the long run. In addition to the 3D printed shoes idea the significant thing to realize here is that 3D printing could also fundamentally disrupt markets by taking out the middlemen. This is an often underappreciated effect that our industry can bring about.
3. The Carbon material is a completely new one and has not been rolled out at scale or used in an end-use product before. In the shoe application it would be used in an uncontrolled environment (as opposed to in the ear or in an assembly) and subject to many chemicals, exposed to many different environmental conditions and subject to considerable wear as well as complex stresses. The material may not function as advertised in this application. The cost of the material and the leasing of the machines seems prohibitively high even with low-end estimates. If a shoe company truly wants to industrialize 3D printing then they would be likely to need redundancy in vendors. Procurement departments would also look very critically at Carbon’s material cost in this application if it would be rolled out in mass manufacturing.
4. If Adidas gets this right then they could make truly individualized shoes that are the perfect size and are adapted to not only fit you perfectly but to fit the way you walk (or should walk) better. To go away from mass-made shoes in few sizes and to go to an individualized higher-performing shoe, that is the big long-term opportunity here. The shoe company that gets this right will actually be able to outproduce and outperform other shoe companies as the market would shift to higher-performing individualized gear. A 3D printed variable density insole that fit perfectly would give the wearer increased comfort, health and performance. This would mean significant costs and investments on the part of the shoe companies but they can’t afford not to do this because the other one may get it right. The sports apparel industry is a bit smoke and mirrors. Run harder, run faster, super ultra super dooper performance midsole, HeatSweat 2.5 Titanium 2000 Extra, UltraMega As Seen On TV Boot, Triple Shark Skin T-shirt, Lion Heart Quadruple Core Socks etc. It is high-performance marketing slapped on mass-manufactured things made by eleven-year-olds. 3D printing has the potential to truly bring individualized performance to apparel and shoes. The key thing is to individualize the shape, fit and performance of the sole to fit one person. This is a huge opportunity and no one shoe company wants to do this because it would require so much additional investment by them. What we’re seeing at the moment are the opening shots of a 3D printing arms race by shoe companies. It is a bit like the outbreak of the First World War. No one wants it to happen, but it might just because of continued provocation by all sides.
5. I personally don’t really see the usefulness of printing the entire shoe, since you could use other digital manufacturing technologies to make flat parts in less time and at less cost. I am however certain that 3D printed unique insoles will be a mass market opportunity for 3D printing. I simply can’t see how one would get the same level of performance from an insole made through other methods. By varying the density of the sole by variations in infill or structure one could 3D print an insole that has higher performance than one made with subtractive manufacturing or through other processes. Other technologies could let you get a unique insole or one that could fit perfectly. But, by varying the internal structures to make the performance of the sole uniquely suited for one foot 3D printing could do an essential trick that no other technology could match.
6. The idea of without any labor making a shoe with one manufacturing step is economically very appealing also. In the long run this could become a viable low-cost 3D printing application if the shoe companies and OEMs invest in reducing labor and optimizing processes. In developments such as Boost and Flyknit we can see that shoe companies are increasingly relying on fewer parts and increased levels of polymers in shoes. Underneath the marketing hoopla there seems to be a development of design for manufacturing going on in the shoe industry. Rather than just the sole being a plastic part, most of the shoe is now plastic. This would reduce costs significantly. If 3D printing is just seen as a way to achieve this in the future then our technology may just be supplanted by another that is higher performing for that one application. In that case the shoe industry’s flirtation with 3D printing will remain just that.

Do you think 3D printing will be a significant part of the footwear industry in the future? Share your thoughts in the 3D Printed Footwear forum at 3DPB.com.