Hacker News new | past | comments | ask | show | jobs | submit login
Perspective on 3D printers from a mechanical designer (graehamdouglas.com)
50 points by graeham on Nov 6, 2012 | hide | past | favorite | 46 comments



Imagine this article, but in 1977 and replacing 3D printer with computer. It would have been wrong, because computers improved rapidly and got cheaper, smaller, faster, lower power, etc.

Almost all the reasons in the article are problems with the current price, size, and reliability of 3D printers.

Someday, there will be a machine that will whip out objects so easily that you'd use it to make a doorstop.


Nope the article does not focus so much on price. It says 3d printers are getting cheaper and cheaper. But his points are valid: just like it does not make sense to carve your own furniture for everything, it does not make sense to do 3D printing for everything in your house either, because you need different kind of materials for different kind of purpose and no 3d printer is going to be able to reproduce every kind of plastic used around you. Plastic is not just plastic, there are thousands of different polymers used in different industries. Not all of them could be used in a 3D printer setting (some can only be molded under high pressure and high temperature).

I may be wrong, but I think you are falling into the typical false technological assumption that a new technology will revolutionize everything just because it becomes available. I mean, look at computers. Even though they are becoming ubiquitous there are not computers in every object we use every day. A lot of our home equipment is still based on "stupid" electronics because it's good enough in most cases.

3D printers will become more common, more reliable and cheaper as the market develops, but when you can order something on Amazon for 3 dollars and have it delivered at your door in one day, there is a strong competition as to what is the best/most effortless way to get the thing in question. If your goal is to make unique objects that you cannot find anywhere, then, sure, 3d printers are the future, but most of us are living in environments where many things have been standardized, and products fitting these standards are readily available.


>> If your goal is to make unique objects that you cannot find anywhere, then, sure, 3d printers are the future

In reality, uniqueness is overrated. Given a really large pool of objects to choose from(which are manufactured in mid quantities) , there's not a lot of consumer need for unique objects.

The largest market consumers do need unique objects is the healthcare market. But the economics there are totally different. for example 3d printing of dental crowns existed for a long time , but it doesn't(yet) captured the market with lower cost solutions(although it certainly has the power to do so, technically).


That simply isn't true. 90% of the things you can 3D print you can machine with a CNC mill. 10% of the things you can make with a CNC mill you can make with a 3D printer.

If residential manufacturing was coming, it would have come a long time ago via CNC technology. Additive vs subtractive is really just a methodology. There are probably more people out there that need custom metal parts for their car, boat, motorcycle hobby than people that only need plastic parts for their hobby. Although I know several with CNC mills and lathes in their garage, it is simply not going main stream.

I think 3D printing has a lot of value (especially 3D printing powered aluminum to make low-volume plastic injection molds), but I don't think you will see one in every home.

(Mechanical Engineer here)


I don't think you're getting the point he's made. Speed, price and quality will improve dramatically over time (we bought a £90k 3d printer a year ago that now has two subsequent successors). In fact the rate of development is almost scary in turns of depreciation of 3d printers. This will enable the widespread use of printers. Maybe not in every home, just like cars are not being owned by everyone. But it will come as your issues are only temporary. Though you have a point in that traditional manufacturing techniques such as cnc milling, moulding etc will still be widely used..


Computers had no competition for their use case. There was nothing that came close to storing, processing, and transmitting information like that. The only thing stopping them from being used everywhere was size and price. If you were dealing with information, you would benefit from a computer as soon as you could afford one.

3D printing is not like that. There is already an entire set of industries built to produce widgets. 3D printing doesn't offer any benefit except for customized items. So how often do you need something customized? No matter how cheap 3D printers are, it's always going to be cheaper to get somebody else to print a one-off for you using their machine. You don't use it all day multiple times a day. You pay them for that one-off, and then you don't need a 3D printer again for a long time.


Hmm, it's interesting to think about. My initial reaction to it is that I don't have a day-to-day need to manufacture stuff, because when I moved into my first solo apartment, I gradually bought the stuff I wanted to have (thinking things like furniture, various electronics, dishes, storage, etc.). I occasionally add or replace something, but that's something that happens at the scale of once or twice per month at most. I certainly don't need a machine sitting around that I have to buy and maintain and supply to fulfill this need.

With computers, there are a lot of day-to-day tasks that become much easier. For example, before computer, everyone had to balance their checkbooks in order to know how much money they had. Now you can just log into your bank website or mint.com. People used to correspond by phone and physical letter to keep in touch with relatives (also on a daily or at least weekly basis), and now that's much easier.

It's hard to say though. With computers, our "need" to perform large-scale automation and computation grew to fill the new capacity. Manufacturing is not as clear cut, but it seems reasonable to theorize that if our abilities to manufacture things grows, our "need" to do so will as well.

And let's not forget that with computers, there was a period where people were saying much the same things that we're saying now.


Also he looks at the future through the lens of today. When you do this you can easily get into a situation where you find a way that it is just not possible for something to work the way it does today using a new technology, and that causes you to discount the value of the technology but it should also prompt you to consider the possibility of change. For example, smartphones are bad for making voice calls on, but people tend to use their smartphones for other things than making voice calls these days.

One thing that I think might happen is that various consumer items may become even more disposable (or perhaps more recyclable). More than that, I suspect that manufacturing itself may become increasingly viewed as a temporary, disposable utility. Right now making most goods requires extensive capital investment in factories and assembly lines, which tend to be large permanent fixtures. But what if manufacturing could be transitory? What if you only needed to make a small run of something? What if you could bootstrap a production line from a small core of machine tools and then dispose of the materials afterward? What if you could efficiently and rapidly break down the production line into its component recyclable parts?

Coupled with completely automated configurable manufacturing I think these innovations will revolutionize our relationship with manufactured goods and with manufacturing. Manufacturing will not be something one consumes it will be something that most people do (either through proxies or through "disposable" production lines or both). The era of cloned consumer goods may give way to an era of boutique goods and to transitory goods. Do you need a radio? Print one off at the corner factory then have it automatically disassembled and 100% recycled when you are done with it. Do you want to give your kids a set of toy soldiers to play with? Set up a quick production line at home and produce a set, then dispose of the machinery. Then recycle the toys when your kids grow bored of them. Don't worry, it'll be just as easy to print more if they want to play with them again.

This is a world that will confound our imaginations because it will disrupt the assumptions that are so deep seated we don't even know we have them and it will create new opportunities and change the balance of feasibility in ways that we lack the creativity to think of, yet.


That's good to hear, because I am constantly running out of doorstops in my home. With current technology I have to go the hardware store 2-3 times a week just to pick up new doorstops. If I had a 3D printer in my home, I could simply print out new doorstops every morning, and never have to worry again about needing a doorstop at 2 AM when the hardware store is closed.

I'm not sure where I'd get the raw plastic, though. Do they sell that at the hardware store?


The issues raised about the materials engineering are not trivial. If you are in the market for custom designs, you are likely looking to spec high-grade parts/materials, for a special use.


Hm ... "Computers are only good for those problems that can be reduced to a bunch of truth table look-ups."


That statement is true though. It just becomes something altogether different when your truth tables have billions of entries.


3D printers are a bit odd. In their current form, I will whole heartedly say no one will use them (short of the tinkerers and creatives). It's not like email where I can go on and connect with someone. It doesn't give me something that I didn't have before. I can't print the expensive things in my life, I can print stupid things like cups, cases, boxes, clips, holders, etc. Light things, cheap things. I can customize them, that is the huge thing 3D printing allows - customization. Short of that, it doesn't have any of the fancy flashery that electronics do.

This isn't to say that it won't happen, but it needs a new spin. It needs the 'killer app' that makes people want it. 1 Kg of ABS plastic (1.75 mm in diameter) is ~$30. PLA is more expensive, but it is renewable (~$45 / kg). It's not free to make stuff once you have one (it will last for a while, but you'll still need raw material).

They are slow, at least for 1D printers (filament printers are parametric printers, we move our nozzle in a line that gets longer with time). SLA printers are faster, way faster. They print planes at a time. They also can have better resolution. The trouble with them is that the liquid material is significantly more expensive as they are UV cured (so they need special handling, and whatnot).

I've been trying to think of something cool for what could be done with the situation (as it would get me a lot of money), but it's a hard problem. We're spoiled by how luxurious electronics are, how precise they are, and how little we have to think about how well they work. Little electro-mechanical things would be cool, and I could probably hand out a lot of personalized presents, but do I really need one in my house? Yes. Do my parents need one in their house? No.


>I can print stupid things like cups, cases, boxes, clips, holders, etc. Light things, cheap things. I can customize them, that is the huge thing 3D printing allows - customization.

Yes, you can print all the cheap low-value add things we currently outsource to China and other low-wage countries, but you can also personalize them.

That's a potential recipe for major disruption.


That's a potential recipe for major disruption.

I can get a personally engraved iPod in a startlingly short period of time.

Lean manufacturing has been pushing towards the idea of pushing different things off the end of a production line for years. They'll carry on doing that. As the technology improves they'll integrate 3D printing type technologies into the supply chain.

Personalisation is certainly interesting - game changing in some instances almost certainly.

That doesn't mean that it's going to happen via 3D printers in your home. I find it hard to think about any kind of personalised object that I'm going to need urgently enough to print now gosh darn it, rather than wait for next day delivery. Especially at the ongoing cost of another box taking up space in my house.

I might have one coz I'm a sad techie who loves fiddling with stuff... but that's not a mass market. That's in the order of magnitude of the home-CNC machine/lathe market.

I can imagine technologies that might hit a mass market. For example if printing and disposing of kitchen mugs / saucepans / cutlery / etc. becomes cheaper/easier/smaller than buying / storing / washing them up then I'd buy a 3D printer like a shot. That kills a major pain point for me. But nothing like that is even vaguely close.

I've yet to see any problem that will persuade 'normal' people to get a 3D printer in the home that's anything close to an implementable reality. I'd love to get educated if I'm wrong ;-)


I'm thinking more ~20 year horizon than next five years. I don't know if 3d printers will advance at the rate that transistors and microchips have, but I expect in ~20 years they will be able to do some interesting things.


Sure, but once you're printing them at home, they're no longer cheap. The material that a manufacturer uses to print a million items will be purchased at wholesale prices. Your onesy-twosy items will be priced retail costing orders of magnitude more.


I expect people will buy 3d printing material in bulk at wholesale prices from Costco's the same way we buy toilet paper and bottled water, use it as needed, and restock when low.


There's a substantial price difference between buying 50 lbs of something you need 5 lbs of and buying 5 tons a month of the same.

There's Costco's bulk lots and then there's manufacturers's bulk lots :-)


Yes, but when 50 million people buy 50lbs bulk lots from hundreds of Costco's (and other wholesalers) across the country, that approaches, maybe exceeds the manufacturer's.


It doesn't matter if, in aggregate, 50M people buy more than a manufacturer does. Each of those people will still get worse pricing than the single large-volume manufacturer would.

The economies of scale in producing the raw material will likely mean that its price will drop a bit, but the guy buying 5 tons a month will always get a much better price than the guy buying 5 lbs a month. Even if there are 50 million guys doing the same thing.


There's a little markup at the wholesaler, but if a wholesaler like Costco buys in bulk they get the same deal as a manufacturer buying in bulk.

The price each individual pays approaches that of the manufacturer, enough to make this a viable alternative. This is nothing new.


For the present time, 3D Printing is NOT revolutionizing the actual production of end products. However the decreased pricing and increased availability (3D printing has been available for a very long time at a high cost) of 3D Printing machines IS revolutionizing the very important part of the product life cycle....Design For Manufacture (DFM)[1].

Rapid Prototyping, Kanban, lean, agile, and other concepts that are revolutionizing software development were all concepts that developed in the Manufacturing world many years ago. 3D Printing is a very valuable tool for rapidly validating and testing designs....release early and often, get feedback, validate, etc.

Most products for the time being(once through the DFM process) will still get produced using traditional methods (Injection Mouldings, Castings, etc).

[1]http://en.wikipedia.org/wiki/Design_for_manufacturability


I've paid for some prototyping runs on a 3D printer of very high resolution (16 micrometers x,y,z if I recall correctly) but I can't imagine wanting to own one. Unless something comes along that drastically changes their usefulness I think they will remain in the realm of the hacker's tool box.

It's not a matter of price coming down, or size, or speed. There's a lack of use cases. There's just no large scale need for everyone to be printing out their own custom widgets.


Not given our current way of life, indeed. Then again, our current way of life is based on the premise that things take time, effort, money, transport, and so on to get into our homes. But why would I need clothes, extra chairs, tableware, tools, and the like if I could just print them when I need them? What if it would become more economical to print common objects instead of having durable items? I would argue that there currently isn't a need because we organised around the idea of durability.

An interesting example in this discussion: Why do a lot of fast-food restaurants use disposable tableware and coffee shops disposable cups? Disposable cups don't make sense if it is impossible to make disposable cups that cost nigh to nothing. If you want to take coffee on the road, just stop at a restaurant, or take a thermos can from home. There is just no need for anyone to have disposable cups, that's just wasteful.

Once these cups became a possibility, however, the use-case of a coffee to go became reality. And with some interesting changes bubbling through society to boot, I think. Compare train stations of a century ago with train stations of today, for example. Or canteens in many a factory with those of 50 years ago. Or your office: where has the coffee lady gone?


Disclaimer: I have only hobbyists knowledge of CAD and 3d printing

The author may be a mechanical engineer, but he certainly isn't an entrepreneur. The interesting question is not "Does it make sense for everyone to have a 3D printer now?" but "What kinds of things become possible once everyone has a 3D printer in their home?"


You are putting the cart before the horses. The first question is "why would anyone buy it?", and if you don't give an excellent reason for that, you will not see a 3D printer in everyone's home. And this makes your question about possible things as useless as a designer's contest (where most of the prizes are awarded for physically impossible designs).


Printing hasn't replaced book-publishing, so there's little reason to think 3D-printing will replace traditional manufacturing.

However, I think he underestimates 3D printing's potential when he says its only for people with design skill. Some people may want the ability to print 3D models they find online, just like people print some PDFs nowadays. In fact, this could be much more useful.


This was exactly what I was thinking as well. To take the analogy further, computers are still used to a ridiculously large extent to prepare paper documents, and the software has become relatively easy to use with a little training. Similarly, I believe, there could be 3d-shape editors for regular people that make the design process for certain kinds of objects easy (such as, connectors, lego-like blocks etc.)


Every point he makes is absolutely correct, and I think comparing them to computers as a rising technology is apples to oranges. For one computers made most of its improvements exponentially because the number of transistors doubled every 18 months. 3D printers don't have an equivalent part, like the transistor, that just simply scaling up makes things perform better. The things holding back are outlined in his article: limited choice of materials, lower quality of the part, slow print times, high cost for mass producing, inability to mass produce, etc. All of this will improve with time, but it won't be on the same time scale as computers.

Now there are somethings I don't think he discussed that change manufacturing because it requires imagining a different way of working. These are the more transformative changes that might have no previous equivalent to compare them to.

For example, the ability to have a de-centralized, agile work force. My wife worked for a company where the M.E. were in the states, but software teams were in Europe. They were constantly sending manufactured parts with people as they physically traveled between offices. It was far cheaper to do this than ship the boards and cases of manufactured products to the team. However, with a 3D printer they could email the CAD drawing to the other office and in a day they could print it out. This is a big change in how fast things could be turned around, and save $1000's of dollars, allowing them to iterate faster.

Another change is 3D printing molds for doing injection molding. Some rapid prototyping shops already do this today. Creating the molds is time consuming, but 3D printing gives you faster turn around. And also more complex molds. This plays to 3D printings strengths low cost for low volume one offs. The quality of the part could be a problem, but hopefully that will improve with time.

I agree with what the author said, but I see the real limit right now is our own imagination for how 3D printers enable us to do new things we can't do now. This is where I see 3D printers are like computers as making something that was impossible before possible. These are much harder to predict. In a way it wasn't just computers that transformed us as much as the internet. Maybe there a companion technology that changes much of these assumptions made.

If we find a problem that 3D printers do better than anything else some of his assumptions could really change. For example, his assessment of 3D printers in homes right now is 100% accurate, but there might be some application we didn't see today for which 3D printers push into the home without significant changes to technology.


The problem with 'doing things that you can only do on a 3D printer' is that you have locked yourself into a 3D printer.

My plan as a hardware entrepreneur would be to 3D print prototypes and MVPs, maybe even the first production run. Then transition to prototype molds or CNC machining with a lot of setup/teardown. Finally you go with full blown steel molds, dedicated machines, fixtures, etc.

If you design for for only 3D printing, you can't scale past 3D printing.


But that can be said for any basis of any tool or technology. Your use of CNC matching creates a dependency on that tool, and it's advantages and disadvantages. You've locked yourself into that tool if you don't want to change the way you work. A shift to any other tool will come with modifications to your work, and I don't think that is unique to 3D printers.


Thanks for your thoughts, you brought up a few points I hadn't thought of. I haven't worked in a international corporate environment but the collaboration ability is an interesting advantage. Printed injection molds are also interesting. I had a project that involved this but it ended up getting shelved before this step so I don't have much original or personal commentary on it.

I am hesitant on the argument of getting excited about 3D printing for as-of-yet unthought of uses. Its a fine argument for science exploring but if we are imagining the future or if I were an investor in 3D printers, a more concrete vision of how A becomes B is needed.


Very fast turn around, low quantity molds (~25 parts) is one of the few extremely promising possibilities for rapid prototyping. For speed you just can't beat the speed that you can get with this method, you can have fully functional parts in your hand a week after sending the model. You can turn around different hardnesses and colours in 2 days. I've used it, and it is a fantastic resource for experimentation with a part.


Everyone's comfort level of accepting new technology is graded on a spectrum. 3D printing is probably at the beginning of the early adopters/innovator stage. The people in this stage have to invision what jobs 3D printing can be applied to and test those ideas in the market place. I think we see that taking shape with companies going public, and other rapid prototyping companies adopting 3D printing and combing it with injection molds, etc. The economics of these ideas and the ideas themselves are being tested with the manufacturing community. These ideas have to cross the chasm between innovators and majority adopters/pragmatists before real adoption will start, and it has to do it on several verticals like: consumer, manufacturers, etc. Each crossing of the chasm will be different in each vertical. I think right now 3D printing has much more promise in crossing the chasm with manufacturing over say consumer, but that could changed depending on the innovators finding applications of it.


This fits with things other people who have worked with 3D printers have told me. I would have liked to see longer timescales described though. How many of these points are going to change soon or not-so-soon? I've heard that prices aren't going down quickly enough to make it economical for everyone to have a 3D printer in their homes in the near future, but it would be nice to see what other people think and to know about other issues blocking widespread adoption.


This article isn't saying that 3D printers aren't important, it actually acknowledges that they are already revolutionizing and democratizing the way new products are developed.

More striking arguments are unfortunately point in time statements, about the current capabilities of technology and the market maturity.

I am not saying that the CNC manufacturing will progress by leaps and bounds towards bringing $100K prototyping machines down to $2K in the next few years, but with increased focus on improving the lower end of the spectrum, and pushing its limits, we will actually find new uses for them as well.

I also don't believe that the ordinary person will go about designing their own parts. However, there will be a good market for people to hire designers, or for small time designers to collaborate to improve open source products.

e.g. One of the biggest ways that I see 3D printing and CNC in general as being important in producing parts replacements. How many times have you needed a broken or missing part for the affordable furniture that we buy from IKEA, Walmart, or Target?

As we use 3D printed pieces, we'll have to use materials which will perhaps be weaker, or in some cases unnecessarily strong but expensive. But, it will be fine, because it will do such amazing things we will not care.


As an aside on Kickstarter since I brought it up, it is interesting they recently banned virtual renderings of design projects. Rapid prototyping allows for moving from virtual models to prototyped models easily, quickly, and cheaply. The problem is that the prototypes in no way prove the company is ready to handle the demands of transitioning into production, or that the prototype has had any reliability testing.

-- This was interesting aside


There are lots of things in the toy, art, and spare part category that can use the 3D printing technology. Just like many other technologies, 3D printing will be adopted by designers, hobbists first. If there is enough demand, it will show up in shared workshop such as Kinko. It may or may not make into every home. Who knows. It really depends on if there is any killer app that drives the demand.


There are products that can use the customization and one-off nature of 3D printing. You just have to think back to before mass production made everything the same.

I still like the example of clothing. Clothing may be good enough now, but people like it customized to them. Put a 3D photopolymer printer, 3D scanner, and some brandname designer in a shop in SF and watch a new trend appear.


I've been thinking long and hard about getting a 3D printer recently. After doing some research on the kind of things that can be constructed with the low-end devices, I've decided against doing it. The things which these devices can make aren't very useful, because the material properties are just not that good with regards to overall strength, temperature range, etc..

I'm now looking at my options for 2D printing, such as cutting parts out of plywood, acrylic, and such. However, the lower cost CNC machines have a very small work area.

I'd really like to get one of those position-correcting routers as in the recent MIT video, or something else which would allow me to easily work with a 4x8 foot plywood sheet.


Second the "build your own" CNC. A good friend recently made a 2x2 CNC that does a great job on wood and Halloween pumpkins for about $600. He did most of the structural pieces out of carbon fiber so it looks awesome but took some time. If you're building it yourself, it really doesn't cost that much to make a bigger one as structure and belts are not the expensive parts.


Build one!

There are many people on CncZone building DIY routers that will work with a 4x8 sheet, or at least will route an entire door.


One of the greatest, long-term advantages of 3D printing I see is the ability to localize manufacturing. Figure out how to print materials we can make domestically, and suddenly stop importing thousands of products.

Imagine small local factories, filled with 3D printers and robots, run by skilled workers. Such factories would be capable of manufacturing thousands of different products for the people nearby on an ad-hoc basis. This could eliminate shipping costs, reduce waste, would make recycling easier, create jobs, increase GDP, etc.


Don't forget reduced inventory cost. No need for an auto parts store to a litany of rarely-purchased plastic parts if they can be printed at will.

Customization is the early killer app, but a world opens up as the material quality and speed improve.


I'm so very glad, in a selfish way,that there is so much pessimism toward 3D printing. More money for the cunning.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: