[ Disclaimer, I bought a 'Replicator' [1] from Makerbot so its the only one I've got 'first hand' knowledge ]
The 'hobby' ones are still very rough. Think Altair/IMSAI level of personal computer. The lack of a reasonable filler/support material (so that you can print 'voids' and later dissolve out the filler) means that a lot of obviously useful shapes are unprintable at the moment (horizontal tubes for example). But the writing is, as they say, on the wall. My instinct is that plastic printing will overtake injection molding for small run products in as few as 10 years.
The biggest challenge to robotics in the US market has been organized resistance. Ford had the most productive auto plant in the world, in Brazil, because it was mostly robotic. They tried, and failed, to get it built in the US. But the depth and length of this recession has changed the politics on that to the point where 'any' jobs trumps 'no' jobs, even if the 'any' comes with the uncomfortable reality that nobody in the new plant will be able to work without at least a solid high school diploma and 2 years of apprenticing. The 'high paying' jobs will require a four year degree in manufacturing technology.
I don't think 3D printing is going to replace injection molding any time soon, even for relatively small runs. The cycle time is just so much longer - an injection molding machine can make most parts, even quite large ones (think big plastic trashcans) in on the order of less than a minute, sometimes much less depending on the machine. That's also ignoring that the plastics you can injection mold have better mechanical properties and surface finishes, too.
Of course, true small run stuff isn't usually injection molded anyway, due to the amount of time and effort and high grade tool steel it takes to make the mold.
Where 3D printing really IS changing things is prototyping and the occasional one-off part, but we're quite far off production usage.
Ok I gave it 10 years as a minimum, is that soon? That number came from the transition in DIY computers to computers for business which I put between 1978 (all hobby) to 1988 where we've got the Mac, the PC, and some lesser known brands (Amiga and Atari) replacing the 'big iron' in companies for some tasks.
As with computers, if a 3D printer is cheap enough you can run 200 in parallel to get production rates into the 'few thousand widgets per day' Today the surface finish and the materials are not up there (but a 2K byte Altair 8800 wasn't really a contender to do desktop publishing either :-)
The thing you have to be careful of is this: "That's also ignoring that the plastics you can injection mold have better mechanical properties and surface finishes, too." because that thinking will bite you back. It just has to be 'good enough' and cheaper and it puts those characteristics into the 'nice to have' category. I agree that there will probably always be a market for some injection molding but I can see that it will be the 'specialized' stuff and not the 'general' stuff. Just a matter of time.
Commercial injection modeling is going to be at the rate of perhaps a thousand per hour, not per shift. Many parts will be set up so that a single "hit" of the mold can produce dozens of parts.
In manufacturing, processes are used that have a minimum number of steps. Cycle time is critical. Even CNC doesn't have that much penetration outside of a very few select industries, namely aerospace and some automotive. Very little, if anything, on a consumer product will be CNC made, for much the same reasons 3D printing will have a very hard time catching on for production.
One real niche for CNC is actually making molds for injection molding - that's one case where taking dozens of hours to cut a single part isn't a problem, as that one mold can stamp hundreds of thousands of parts.
Source: I'm actually a manufacturing engineer by training.
EDIT: And yes, I think 10 years is much too soon. Industry has a huge amount of money invested in existing factories. Much of that equipment is useable for decades to come. You would have to get quite a bit better than the existing standard for them to even think about replacing the existing lines.
Awesome, I disagree of course but its a great discussion. My experience is with the integration of low cost compute into the flows rather than manufacturing. However I find it striking how similar your argument sounds to the ones that IBM made early on about how PCs would 'never' threaten mainframes because they didn't have the IO capabilities or the printing capability. And while certain markets still use them (credit card processing for example), a lot of things they were used for succumbed to a new way of doing things because it was cheap.
I suspect that much of the challenge with CNC penetration is with the costs involved. But could easily be wrong.
Current injection molding does make thousands per hour and that's certainly achievable with thousands of 3D printers running in parallel, but the place where I think you will first see disruption will be in the 'short run' segments, where people would make 100,000 parts (minimum run) and only sell 10,000 units. Their price per unit was effectively 10x in that case. But they have no other way to get to 10,000 at the moment. In the compute space that is how the market disrupted, lower performance of the unit offset by the savings of not having to over provision for the compute capacity. Will be interesting to watch.
Yes, even for DYI's, 3D printing now (even with the rough ones) is a cheap tool to just try out prototypes. And try out A LOT of prototypes cheaply. Even though they are slow, it is cheap and you just can try out so many things before you go to production with CNC & injection molding. And you can give away limited prototypes so you can do market research with the printed models; even though they are rough, I have seen them work well enough for people to get this 'wow I want this' feeling.
Makerbot charges $45 per KG for their raw material, whereas steel (in bulk) is more like $1/kg, and even in small orders (like say small bar stock) is $10-$15/kg.
That's why I wouldn't use it for anything else but prototyping. But what would you consider for fast prototyping instead of a 3D printer? With the same flexibility; you basically input (not that simple, but you know) a 3D model and get your prototype as output (like said; that's oversimplified but that's how I use it)?
But I am very new to this; it's only a hobby; i'm checking out 2nd hand CNC machines, but in my mind that's for the next step, after the prototyping.
Well, it depends on what you're doing. The problem with 3d printed stuff is that it isn't very strong - you can make crazy curved gears and stuff with it, but it's not going to hold load.
Honestly, for prototyping for small amounts of money, your best bet is probably a bridgeport-type vertical mill. You can do turning/lathe operations on one of those too with the right accessories. Not automated, and not fast, but if you need to make one off parts it's about your only option for hobbyist-type money. You'd be buying used, obviously, and you'd need a garage/workroom (This isn't something you set up in your house)...
There are some fairly economical small CNC machines now (We're talking $2k or so, price-wise), but they are intended primarily for model makers - any part over about 2"-3" and they literally can't cut it, where-as with a good manual milling machine and some time/sweat, you will typically have 24" x 24" x 10" travel available.
>Google recently announced that its Nexus Q streaming media player would be made in the U.S., and this put pressure on Apple to start following suit.
So Tim Cook announces Apple's intention to do this at the end of May, and this is somehow "following suit" when Google hadn't even announced the Nexus Q yet (June 27)?
>We can only guess. Autodesk CEO Carl Bass says that just as we have created new, higher-paying jobs in every other industrial transition, we will create a new set of industries and professions in this one.
What are average humans capable of that robots aren't?
My memory is that Apple announced a "desire" to do more manufacturing in the US. Google announced an actual product that will be built there. Not really the same thing. Someone correct me if Apple is actually sourcing anything other than semiconductors from the US for a planned product.
The whole "where is it built" thing is dumb, though. But still, if you're going to take afront, this doesn't seem like the right spot.
I'm just pointing out an inaccuracy, not making a claim to Apple's greater virtue: the article asserts Apple is being influenced in May by something announced in June (unless one is to assume Apple got an insider tip).
I think you're reading too far. The assertion is pretty straightforward: an Android device built in the US puts pressure on Apple to do the same. And the sentence links to a story about Apple wanting to build in the US but not being able to do so. I don't see an assertion of causality. They just looked for an opportunity to link to another Forbes story, saw a relevant one, and threw it in. The author may not have even written the link in the original.
Manufacturing design jobs are for average humans, unfortunately. Until something replaces the jobs lost that are suitable for average humans, there will likely be a risk of political instability (which is likely one of the reasons why the US has been heavily investing in domestic counter-insurgence infrastructure over the last decade).
True, but I think Qworg was instead suggesting that humans are good where the confidence/accuracy of robots drops off. Machine learning tasks usually get to 70% good enough pretty quickly, but after that each percent gain costs more and more. There's a point where humans can complement robots when the robots are not yet good enough to solve the problem.
For simple tasks. Identifying thousands of different products by sight alone in infinite numbers of configurations at varying light levels/obstruction? Humans are best.
I've been saying this for a while. I think all manufacturing will turn local at some point, though it may take decades still. If I ran FedEx/Kinkos, I'd be researching all the best 3D printers and trying to pick winners and losers.
A while back (seems like it's been about a year) Radio Shack solicited ideas to make their stores more relevant. I suggested they install 3d printers in every store to bring hobbyists back in.
The US already has plenty of manufacturing, we make more stuff in the US right now than we ever have in the past. The thing is that we just use robots instead of people to do most of it - and 3D printing will only tend to continue this trend.
Surely if the robots are replacing people then they're not really creating jobs, just taking jobs away from the Chinese.
Sure there will be some people manning the robots, but isn't the whole point of using robots to cut down on the human workforce?
Even the robots aren't created in USA - some are Kuka robots from Germany.
Many simple products are made completely by machine yet have been coming for a long time from places like China. Sometimes it is due to the Chinese owning better machines (e.g. mills in China are generally more advance than their counterparts in America).
3D printers are probably going to be a boost for Mexico and low cost countries in Europe. That is if they get their act together.
I've seen a lot from the DIY companies and academic demos - what companies are designing this kind of production-quality cutting edge robotics and AI solutions though? I'd like to be a part of this future, but it seems like those jobs are in much shorter supply and harder to find.
I would say the jobs in 'cutting edge' robotics (no pun intended) are in surgical robots. Intuitive Surgical is growing very quickly. Proto-typing is at companies like Zmachines.
You need to realize that the personal computer market was all DIY and academic demos until Lotus 123 on the Apple came out. That filled a niche in business. When the PC came out it was supported by a 'real' company (Apple was not considered much of a company then). I would expect a similar trend in 3D printing. Some niche will open up and cause a lot of 'enterprise' money to be spent there and then one of the printer companies will jump in with the kind of product these things will evolve into. (And you'll get to complain to your grandkids about all the beige 3D printers are so boring, you remember when they were innovative and quirky)
Much of the AI stuff has been in pick and place robots (see the Japanese competition on this) and the military's use.
The DIY aspect is a very good point. I just finished watching a few videos on YouTube on early machine tools and one of the narrators mentioned that a lot of early milling machines didn't even have manufacturer's names because they were built in-house by the same people who needed them.
There is a very large DIY contingent building CNC machines of all flavors, not just the basic lathes and mills, but routers, benders, presses, wireforming tools, etc. Take a look around cnczone.com; you can waste hours on that site.
The DIY part of the industry needs software. The dominant tools are still very basic and there doesn't seem to be much support for 4 and 5-axis machining because most DIYers are only building 3-axis machines at most. So there may not be a huge number of jobs, but there is room to make your mark writing software for the home-built project machines.
I'm awaiting the day someone posts their design for a 5-axis homebuilt mill fully integrated with their own CAM software. I think that person will become very rich.
I seem to remember hearing some argument about why it is hard to bring manufacturing back to the US, and it had to do with the supply chain in China being vastly shorter. I'm not sure how robotics would address that. Thoughts?
the kind of jobs that would be created would be to install and maintain these robots and machines, until someone figures a way to automate that as well
Such is the fate of China also, they're just a bit further removed from it.
Automated manufacturing is vastly more reliable and higher-quality than anything manual labor can pull off. As the standard of living rises in China and cost of labor exceeds the threshold, automation will take over just like it has here.
The death of US manufacturing is a little overblown (though not entirely) - what people are really talking about when they bemoan the death of American manufacturing is the death of American manufacturing jobs.
"The death of US manufacturing is a little overblown (though not entirely) - what people are really talking about when they bemoan the death of American manufacturing is the death of American manufacturing jobs."
Exactly, if you look at manufacturing output, America has remained a leader [1], but the number of people employed by manufacturers has gone down. The recession did put a dent in that increase of course.
"The wars of the future will not be fought on the battlefield or at sea. They will be fought in space, or possibly on top of a very tall mountain. In either case, most of the actual fighting will be done by small robots. And as you go forth today remember always your duty is clear: To build and maintain those robots."
But with any luck, the products of fully automated local manufacturing will be so much cheaper that humans won't need to work long hours in monotonous jobs in order to buy them.
I love the sentiment, but I doubt it will come to pass. That's exactly what people have said at every stage of industrialization. But the gains always end up being consolidated and the long work weeks and boring jobs march on.
The long work weeks continue because apparently we are willing to trade less time at work for more stuff. If you want to live a 1960s lifestyle with a 60s size house, 60s level health care, 1 car per family, only 60s era gadgets, and cooking the vast majority of your meals--you could easily do it on 20 hours a week for most jobs.
As an experiment to illustrate my point, look around the room you're in and count the things that you own that didn't exist in 1960.
I think we have to factor in the VAST increases in quality of life that industrialization brings. Increasingly automated car manufacture undoubtedly still involves boring jobs, but it also brought car ownership within the realm of possibility for millions more people.
The same applies for anything from textiles (we can now buy clothing for a few dollars that would once have only been available to royalty) to electronics (in a world where almost everybody carries a mobile phone).
If you look at the history of the industrial revolution , many of the gains the workers got:reasonable working hours, minimum wages, etc. were gained by a strong political fight(that many times took blood).
There's no reason to assume that this time it will be different.
For those who aren't well read (grin) a very good read looking at a dystopian future where 3d printers have displaced manufacturing is Cory Doctorow's "Makers". The Creative Commons licensed book can be downloaded from his website, craphound.com, or purchased elsewhere.
I suspect it may be one of those near-prophetic works of science fiction that will amaze future generations - "How'd he see that coming?"
Does anyone know of any US based consumer electronics factories set up to take small-medium sized orders from small companies...say for kickstart campaign funded products like the Ouya?
One problem for consumer goods production in europe / US is component sourcing. You will need a PCB-Fab, ICs, injection molding, ... . While you can get most (some things just arent built statesside anymore) of those things in the states, they will be more expensive and maybe not built to capacity (since they specialized on really small batches for example). In china you will have all of the suppliers down the road and pretty cheap.
From someone who worked for a large automotive supplier, which does a lot of manufacturing here in germany, that wanted to release a consumer device I heard that you can get the whole product finished in china for the price of sourcing the components in europe.
It's FAR more expensive to produce in the US, Usually between 2x-4x. I do prototyping in the US and the companies that I use also do small-medium production runs. You can find them in the Manufacturing thread I wrote.
The 'hobby' ones are still very rough. Think Altair/IMSAI level of personal computer. The lack of a reasonable filler/support material (so that you can print 'voids' and later dissolve out the filler) means that a lot of obviously useful shapes are unprintable at the moment (horizontal tubes for example). But the writing is, as they say, on the wall. My instinct is that plastic printing will overtake injection molding for small run products in as few as 10 years.
The biggest challenge to robotics in the US market has been organized resistance. Ford had the most productive auto plant in the world, in Brazil, because it was mostly robotic. They tried, and failed, to get it built in the US. But the depth and length of this recession has changed the politics on that to the point where 'any' jobs trumps 'no' jobs, even if the 'any' comes with the uncomfortable reality that nobody in the new plant will be able to work without at least a solid high school diploma and 2 years of apprenticing. The 'high paying' jobs will require a four year degree in manufacturing technology.
[1] http://www.makerbot.com/docs/replicator/