3d printing technology, while wonderful, does not scale well. It can do remarkable things, but the hype-level surrounding it is so extreme that it’s worth pointing out what 3d printing does NOT do well, and likely never will.
Why do we bring this up? Because 3d printing has become a brand, a concept – shorthand for ‘new ways to make things’, or ‘the future of manufacturing‘. It’s similar to the way that in the public mind the word solar means “solar electricity”, totally forgetting about solar thermal. (Yes, we’re biased).
What lies behind the excitement is a dream that 3d printing can transform manufacturing in a Moore’s Law-like way. Investors: “Stodgy old manufacturing will be displaced by magic new 3d printers! Look, they’re powered by COMPUTERS! We did well with our computer investments… Let’s jump on this and wait for 100x returns”.
Again, there’s a parallel with solar – the early ’00s dream was that solar electric was going to see exponential growth – but instead the business saw strong linear growth. Investors: “LINEAR growth? SO boring, why even get out of bed for this”?
So below I’d like to take a second and crush a few dreams, so we can all wake up and get back to work. Disclaimer: 3d printing IS awesome, has lots of value, does transform / disrupt / broaden manufacturing. But it’s not magic.
1) Energy Needs Don’t Change
If you mold a plastic part, the energy required to melt the plastic and form it to your desired shape changes NOT AT ALL when a print-head melts the plastic one tiny squirt at a time vs the plastic melting in a ‘bulk’ old-school process like rotomolding. To the extent that energy cost contributes a measurable chunk of the total cost to make something, 3d printing costs the same as older methods. It takes the same number of Kj to melt 30 Kg of polyethylene no matter if you 3d print, roto mold, blow mold, etc. (Yes, efficiency matters, and some processes are more or less effective at getting heat TO the plastic target – but the fact remains that 3d printing still has to melt plastic like every other process).
2) Mass Customization is Overrated
Mass customization is coming! Yay! Throw off your chains and design your very own special (insert your favorite product here). With 3d printing the sky is the limit!
True – and a terrible idea for many products. Mass production implicitly promises that the product’s design has been deeply thought about, and that the product interfaces with the rest of the world. Sure, for some fields the ability to do short runs is critical – fashion, out-of-production spare parts, parts impossible to make any other way, etc. – but do we really need or want mass-customized water tanks or other industrial commodities?
Remember MySpace V1.0 and the horror of (horribly) mass-customized user profiles? Many people don’t have the skill, interest, or time to design anything. Sorry Make. It’s great to put powerful prototyping tools into more designers hands, and to let more ‘civilians’ enter the design world and scratch their preferred itch. For a designer, even an amateur one, 3d printing is a godsend. For the actual mass-market it’s irrelevant except in edge cases.
We think everyone should have tools or a workshop, should be willing to take apart things, void warranties, learn about the products they own, mod or build their own stuff, and 3d printers are a amazing tool.
But don’t be fooled into thinking that increasing the size of the Maker-world means that Maker tools are going to eat mainstream production methods at scale. They’re not. Billions of customers don’t want to design anything, and prefer standardized stuff advertised and sold and warrantied to them in mass quantities, so they can get on with their lives.
3) Tooling Costs Are Irrelevant at Scale
It’s true – 3D printing avoids the crazy tooling costs of injection molding. You can make a fork, or toy or what have you without dropping $100K+ on a massive mold (and God help you if you’ve made a mistake in designing that mold). So in the short-run market 3d printing blows up the economic model and that’s great. BUT tooling costs are asymptotic to zero in all at-scale plastic processes. If I mold four million forks, my tooling costs per fork disappear. And that’s for injection molding, the molding process with the highest tooling costs. Rotational molds are a order of magnitude cheaper, as little as $5-$10,000, and quick to make. It takes just a few hundred parts to render tooling costs irrelevant in rotomolding. Similar math applies to compression molding, vacuum forming and other methods. So the excitement about 3d printing’s zero tooling costs is warranted in the pre-production phase. But as soon as you want to do real volume, tooling costs don’t matter. What does? Cycle time.
4) Long Cycle Time is a Disaster
3d printing has gotten faster, but it is still slooooooooow. You have to programmatically place every bit of plastic where you want it, for the whole object, every time. How can we improve this? Multiple print heads? Faster lasers sintering the goop? It’s non-trivial. Know what IS trivial? Making the 3001st part using a big simple mold via rotomolding or vacuum forming, after you’ve previously made 3000 parts. The intelligence is built into the tool (mold) itself, and each production cycle takes zero computational power, relatively few operator interactions, etc. You can mold a massive water tank in about an hour with rotational molding (either our Solar Rotomolding Process or traditional molding). If you’ve ponied up the cash for injection tooling, you can bang out a plastic briefcase in less than a minute. 3d printing is nowhere near this kind of performance.
The Good Stuff
3d printing is ‘machining plus’ – an infinite-axis process by which one can slowly make awesome things, sometimes objects that can’t be made any other way. It will get more and more powerful, and eat market share of some other processes. It’s solving problems for aerospace, medicine, and desktop hobbyists.
…and the Limits.
But world, please get this: 3d printing will never produce exponential improvements in manufacturing speed for large production runs, and it has no hope of displacing ‘boring’, mature technologies like rotational molding, vacuum forming, compression molding etc, especially for making large objects.
A 3d printer is not the physical analogue of Steve Jobs’ “Bicycle for the Mind”. It’s not a Star Trek replicator. But for a lot of investors and trend-spotters it is definitely a Bright Shiny Object.
– KvK 3d Printing is Not the Future of Manufacturing
3d printing technology, while wonderful, does not scale well. It can do remarkable things, but the hype-level surrounding it is so extreme that it’s worth pointing out what 3d printing does NOT do well, and likely never will.
Why do we bring this up? Because 3d printing has become a brand, a concept – shorthand for ‘new ways to make things’, or ‘the future of manufacturing‘. It’s similar to the way that in the public mind the word solar means “solar electricity”, totally forgetting about solar thermal. (Yes, we’re biased).
What lies behind the excitement is a dream that 3d printing can transform manufacturing in a Moore’s Law-like way. Investors: “Stodgy old manufacturing will be displaced by magic new 3d printers! Look, they’re powered by COMPUTERS! We did well with our computer investments… Let’s jump on this and wait for 100x returns”.
Again, there’s a parallel with solar – the early ’00s dream was that solar electric was going to see exponential growth – but instead the business saw strong linear growth. Investors: “LINEAR growth? SO boring, why even get out of bed for this”?
So below I’d like to take a second and crush a few dreams, so we can all wake up and get back to work. Disclaimer: 3d printing IS awesome, has lots of value, does transform / disrupt / broaden manufacturing. But it’s not magic.
1) Energy Needs Don’t Change
If you mold a plastic part, the energy required to melt the plastic and form it to your desired shape changes NOT AT ALL when a print-head melts the plastic one tiny squirt at a time vs the plastic melting in a ‘bulk’ old-school process like rotomolding. To the extent that energy cost contributes a measurable chunk of the total cost to make something, 3d printing costs the same as older methods. It takes the same number of Kj to melt 30 Kg of polyethylene no matter if you 3d print, roto mold, blow mold, etc. (Yes, efficiency matters, and some processes are more or less effective at getting heat TO the plastic target – but the fact remains that 3d printing still has to melt plastic like every other process).
2) Mass Customization is Overrated
Mass customization is coming! Yay! Throw off your chains and design your very own special (insert your favorite product here). With 3d printing the sky is the limit!
True – and a terrible idea for many products. Mass production implicitly promises that the product’s design has been deeply thought about, and that the product interfaces with the rest of the world. Sure, for some fields the ability to do short runs is critical – fashion, out-of-production spare parts, parts impossible to make any other way, etc. – but do we really need or want mass-customized water tanks or other industrial commodities?
Remember MySpace V1.0 and the horror of (horribly) mass-customized user profiles? Many people don’t have the skill, interest, or time to design anything. Sorry Make. It’s great to put powerful prototyping tools into more designers hands, and to let more ‘civilians’ enter the design world and scratch their preferred itch. For a designer, even an amateur one, 3d printing is a godsend. For the actual mass-market it’s irrelevant except in edge cases.
We think everyone should have tools or a workshop, should be willing to take apart things, void warranties, learn about the products they own, mod or build their own stuff, and 3d printers are a amazing tool.
But don’t be fooled into thinking that increasing the size of the Maker-world means that Maker tools are going to eat mainstream production methods at scale. They’re not. Billions of customers don’t want to design anything, and prefer standardized stuff advertised and sold and warrantied to them in mass quantities, so they can get on with their lives.
3) Tooling Costs Are Irrelevant at Scale
It’s true – 3D printing avoids the crazy tooling costs of injection molding. You can make a fork, or toy or what have you without dropping $100K+ on a massive mold (and God help you if you’ve made a mistake in designing that mold). So in the short-run market 3d printing blows up the economic model and that’s great. BUT tooling costs are asymptotic to zero in all at-scale plastic processes. If I mold four million forks, my tooling costs per fork disappear. And that’s for injection molding, the molding process with the highest tooling costs. Rotational molds are a order of magnitude cheaper, as little as $5-$10,000, and quick to make. It takes just a few hundred parts to render tooling costs irrelevant in rotomolding. Similar math applies to compression molding, vacuum forming and other methods. So the excitement about 3d printing’s zero tooling costs is warranted in the pre-production phase. But as soon as you want to do real volume, tooling costs don’t matter. What does? Cycle time.
4) Long Cycle Time is a Disaster
3d printing has gotten faster, but it is still slooooooooow. You have to programmatically place every bit of plastic where you want it, for the whole object, every time. How can we improve this? Multiple print heads? Faster lasers sintering the goop? It’s non-trivial. Know what IS trivial? Making the 3001st part using a big simple mold via rotomolding or vacuum forming, after you’ve previously made 3000 parts. The intelligence is built into the tool (mold) itself, and each production cycle takes zero computational power, relatively few operator interactions, etc. You can mold a massive water tank in about an hour with rotational molding (either our Solar Rotomolding Process or traditional molding). If you’ve ponied up the cash for injection tooling, you can bang out a plastic briefcase in less than a minute. 3d printing is nowhere near this kind of performance.
The Good Stuff
3d printing is ‘machining plus’ – an infinite-axis process by which one can slowly make awesome things, sometimes objects that can’t be made any other way. It will get more and more powerful, and eat market share of some other processes. It’s solving problems for aerospace, medicine, and desktop hobbyists.
…and the Limits.
But world, please get this: 3d printing will never produce exponential improvements in manufacturing speed for large production runs, and it has no hope of displacing ‘boring’, mature technologies like rotational molding, vacuum forming, compression molding etc, especially for making large objects.
A 3d printer is not the physical analogue of Steve Jobs’ “Bicycle for the Mind”. It’s not a Star Trek replicator. But for a lot of investors and trend-spotters it is definitely a Bright Shiny Object.
– KvK