3D printer to churn out copies of itself

  • 13:32 18 March 2005
  • NewScientist.com news service
  • Celeste Biever
Printable versionEmail to a friendRSS FeedSyndicate
 
The metal circuitry was printed onto this autonomous robot using a low-melting point alloy (Image: Adrian Bowyer)
The metal circuitry was printed onto this autonomous robot using a low-melting point alloy (Image: Adrian Bowyer)
Related Articles
 

A self-replicating 3D printer that spawns new, improved versions of itself is in development at the University of Bath in the UK.

The "self replicating rapid prototyper" or RepRap could vastly reduce the cost of 3D printers, paving the way for a future where broken objects and spare parts are simply "re-printed" at home. New and unique objects could also be created.

3D printing - also known as "rapid prototyping" - transforms a blueprint on a computer into a real object by building up a succession of layers. The material is bonded by either fusing it with a laser or by using alternating layers of glue. When it first emerged in the mid-1990s, futurists predicted that there would be a 3D printer in every home.

But they currently cost $25,000 (£13,000) and so have not caught on as a household item, says Terry Wohlers, an analyst at Wohlers Associates, a rapid prototyping consulting firm in Fort Collins, Colorado, US. Instead, they are used by industry to develop parts for devices such as aircraft engines, spaceships and hearing aids.

Plummeting prices

Now Adrian Bowyer hopes to change that by making the first 3D printer capable of fabricating copies of itself, as well as a wealth of everyday objects. He reasons that prices would plummet to around $500 if every machine was capable of building hundreds more at no cost beyond that of the raw materials.

Better still, the machines could evolve to be more efficient and develop new capabilities, says Bowyer. Once he has the software to guide the self-replicating process, he plans to make it freely available online, allowing users to contribute improvements, just like the open-source Linux computer operating system, he says.

Bowyer dreamt up the idea of the RepRap in February 2004. But now he has he figured out how to print conducting materials in three dimensions without using a laser, a key step if the machine is ever to make copies of itself.

"We are very constrained in our access to materials," he explains. They must be sturdy enough to make up the body of the machine and yet simple enough to be fabricated entirely by the machine. "We have to avoid any design needing lasers and high precision measuring systems," he explains.

Tepid metal

3D printers normally build circuits by fusing together a powdered metal with a laser. But Bowyer plans instead on using a low-melting point metal alloy of bismuth, lead, tin and cadmium that can be squirted from a heated syringe to form circuits.

Bowyer has already produced an electronic circuit by squirting the alloy inside a plastic autonomous robot, which itself was created using a commercial 3D printer. Because the heated syringe he used is very similar to the nozzle that deposits plastic layers in the printer, he envisions squirting both plastic and metal from the same nozzle in future self-replicating machines.

The machine need not be capable of assembling itself, he says, only producing all the necessary parts, with the exception of the microprocessors and the lubricating grease. These could later be added and the various parts clipped together, Bowyer says. "People are quite capable of assembling things if they want to," he adds. "I am not interested in self-assembly, just self-copying."

Whether such a machine would work has experts sharply divided. "I think Dr Bowyer's idea is very plausible," says Matt Moses, a consultant who has built a small self-replicating robot and advises NASA on research into self-replicating machines for space.

But Wohlers disagrees: "[Bowyer] is referring to something that does not exist and has not been demonstrated. Will it develop in the future? Unlikely."

He adds that even if all its components could be replicated by the machine, the concept does not make economic sense. "Many of the components could be produced much faster and cheaper by other machines," he says.

Comment subject
Comment
No HTML except lower case italic tags or lower case bold tags, please:
<i> or <b>
Your name
Your email
 

We need your email in case we need to contact you about the comment. We will not use it for any other purpose.

 
 
There are 50 comments on 6 pages
1  | 
2  | 
3  | 
4  | 
5  | 
6
 | Next
 | Most Recent
 | See all

Entropy

By Jon Pryne

Wed Jan 09 00:42:07 GMT 2008

"We have to avoid any ... High precision measuring systems," he explains.

Without high precision measuring systems, the newly replicated model won't be as accurate as the original replicator. This means an increase in entropy, which in turn means that the machine will cease to function before too many iterations are produced, if any.

REPORT | REPLY

Entropy

By Demented Chihuahua

Sun Jan 20 05:59:13 GMT 2008

The orignal machine itself is built without the use of high precision instruments and functions just fine. I think the current figure of the parts of itself that it can replicate is 60%--January 19, 2008. To say that entropy will increase suggests that there is no human intervantion to clean things up, put things together and generally help out. Also, total system entropy may be increasing but local order can and does increase. I give you the crystal and DNA as examples. Suggesting otherwise displays a lack of knowledge about thermodynamics. Take a class and then comment.

Demented Chihuahua (RepRapper for life!)

REPORT | REPLY

Entropy

By Jon Pryne

Wed Jan 30 22:04:55 GMT 2008

Wow, DC, pretty snarky, there! Okay, good luck! I think you'll have troubles without some kind of precise measuring component, but that's just me. (Me and lots of others, that is.)

REPORT | REPLY

Entropy

By Kelek Serodores

Thu Apr 10 23:18:06 BST 2008

I think you're both disputing over the notion of 'highly precise' vs. Precise. CPU wafers produced are not perfect, and sometimes do have defects. Thus those fabrication methods aren't 'highly precise', yet still the current method chipmakers use. Sure, they try to produce more accurate yields, but they are never 100% accurate. Theoretically, the measuring has to just be 'accurate enough'. Even some mechanism which can determine which parts may be corrupt and may need to be reprinted would allow for precise, but not 'highly precise' measurements to be used. ...

1 more reply »

REPORT | REPLY

Entropy

By Whipster

Mon Apr 14 12:01:15 BST 2008

Hey DC makes a lot of sense. And anyways... We're not saying this thing will let you print out a new computer or mobile phone (maybe some day down the line), but allow you to quickly prototype simple ideas. Lets say you were starting up a door chime company, and you had all of your designs, but lacked the know how to get it produced. You could prototype it with this... Until it was perfect at it's job - with a case and everything that would make it an actual product.

Then you can self replicate your 3d printer until you have say 10 or so, and pump out 10 of these door chimes a day. Then, at the end of the week setup a stall and sell them or something!

I think this thing is perfect, and if i could buy one, i would - just so that i could build some simple circuits and the cases for them...And the best thing about it would be that you could go on the net, see what other people have made with them, and download the latest schematics for some really cool, self assembly electronics!

Honestly, some of you guys ask the earth - We have to start somewhere, and if the first person to try his hand at it keeps getting shot down then we'll never progress!

REPORT | REPLY

Entropy

By Blues

Wed Feb 20 18:53:23 GMT 2008

This machine is total fail, give up before your mom goes broke

REPORT | REPLY

Entropy

By Anonymous

Wed Apr 09 20:42:11 BST 2008

Entropy can be lost as heat through the surrounding environment. The machine basically squirts hot material to make things. Thus there can be a decrease in entropy.

And "high precision" is relative... All the machine needs to do is make parts more precise than it is. And that has basically been the history of human development; we make crude tools, to make better tools, to make even more precise tools, etc.

REPORT | REPLY

Entropy

By Rob K

Wed Apr 09 21:49:49 BST 2008

Depends on how much precision you need, wouldn't it? If it can create a version of itself within the functional tolerance for it's own specifications, it seems it would only have to be "precise enough". To use a photocopier analogy, more like making multiple copies from an original (the specifications), than making a new copy using the last copy that was just made. Just quibbling, really.

REPORT | REPLY

Partnership

By Aniko Rubes

Tue Feb 26 18:33:59 GMT 2008

This comment has been found to be in breach of our terms of use and has been removed.

REPORT | REPLY

Doesn't Make Economic Sense?

By Michael Rubini

Thu Mar 06 16:43:49 GMT 2008

I don't understand why it wouldn't make economic sense. If you had "things" that can replicate themselves, then you can end up with exponential amount of things in a short time. That is the point, isn't it?

REPORT | REPLY

Doesn't Make Economic Sense?

By Dan

Wed Apr 09 19:30:58 BST 2008

Having this machine make simple parts would cost more than having a factory make millions of those simple parts.

REPORT | REPLY

Doesn't Make Economic Sense?

By Travis

Wed Apr 09 19:41:03 BST 2008

Sort of... But you're missing some major knowledge of manufacturing. It's easy to create a bunch of parts, but it has to be done as quickly as possible as well. I have a 3D printer in my shop (tech school) and it's very slow. Now, ours isn't the newest thing, but I imagine they haven't increased in speed THAT much. (Rapid prototyping doesn't mean something is made in the blink of an eye)

The problem is that the parts and components can easily be manufactured by autonomous machines at a much quicker rate that by the printer itself. If it takes the printer a day to make the parts for one more but you can use an assembly line to make the parts for 50 in a day you are at a loss economically.

Anyway, it could easily amount to something later on, but at the moment it won't do much good other than for proof of concept which can be applied later.

REPORT | REPLY

Doesn't Make Economic Sense?

By Eric

Wed Apr 09 20:56:03 BST 2008

I don't agree simply because for every assembly line you need a worker there is no such thing as the perfect autonomous assembly line. I cringe as I say that however. This works no differently. What you would save in terms of Cost would make up for time. IMO; I believe that what you all are referring to is you would be at a loss in efficiency not economically. Due to the fact that it takes longer for the machine than a factory, but I see some savings there monetarily.

Just my .02

REPORT | REPLY

Doesn't Make Economic Sense?

By James

Thu Apr 10 03:22:39 BST 2008

"If it takes the printer a day to make the parts for one more but you can use an assembly line to make the parts for 50 in a day you are at a loss economically."

You are kidding right? Surely. If it can copy itself once a day then on day 2 you will have 2 copiers to make copies. Day 3 you have 4 copiers, then 8, 16, 32.

By day seven you are building more than the factory, the next day 2x as as much as the factory, then 4x, 8x, 16x etc.

REPORT | REPLY

Doesn't Make Economic Sense?

By Kirby

Thu Apr 10 18:57:54 BST 2008

Lol! That is assuming that you have infinite space to setup your newly fabricated printers.

I think its just a proof of concept. To lure people into looking into the product.

"Wow if this thing can actually make a copy of its self, it could make all sorts of stuff"

I don't think the creators of the product would be putting this info into the public sphere, if they were worried that family RepRap shops would print them out of house and home.

I could see a friend and I splitting the cost, and printing a 2nd copy for one of us. BUT We are nerds, normal people will have that intent and then windup printing puzzle kits or something....

Good text book econ though :p

REPORT | REPLY

Doesn't Make Economic Sense?

By Anonymous

Wed Apr 09 20:46:57 BST 2008

The model for self-producing 3d printers isn't the manufacturing line, its the cottage industry. Cottage industry was abandoned for a reason; its less efficient than the manufacturing line.

REPORT | REPLY

Doesn't Make Economic Sense?

By James

Thu Apr 10 00:31:08 BST 2008

It's not about efficiency, its about locality and independence. Say you want to make a camera. You've got to have 10 different kinds of screws, plastic molding, lenses, servomotor mounts. In the current economic system, all these parts would have to be made in separate factories and shipped to some central assembly point where they would be put together. However, a reprap is versatile, so that it would make one of all the parts and then you would put it together. It's less efficient, yes - but If you just want one camera you don't care about efficiency as much.

REPORT | REPLY

Doesn't Make Economic Sense?

By Jb

Thu Apr 10 02:41:55 BST 2008

This sort of production doesn't really fit into any current manufacturing model. If I have one of these machines, I could make one for you and a couple of other friends without really entering into a manufacturing relationship. Then each of you could do the same. It's more like reproduction than manufacturing, and population growth would happen with no traditional manufacturing taking place. If I am willing to push the button and load the machine with materials because you are my friend, the cost is negligible, and the time required for the machine to make the parts is of no importance since there is no direct human labor.

Also, under this model, the role of capital investment is entirely different from traditional manufacturing, in fact - hardly existing at all.

REPORT | REPLY

Doesn't Make Economic Sense?

By Wangenstein

Tue Apr 15 20:03:54 BST 2008

What about someplace where it's impractical to set up a full manufacturing facility, like the moon, or the bottom of the ocean? Perhaps this technology could allow us to send one small device that would then build everything else we need (including additional copies of itself to speed production). Think outside the (self-replicating) box... :)

REPORT | REPLY

Doesn't Make Economic Sense?

By John Crofford

Wed Apr 16 21:33:53 BST 2008

The cottage industry was abandoned because it could not compete with factories in terms of worker productivity (or quality, in some cases). Those were the concerns at the time of the Industrial Revolution. My concern is the energy cost of moving the raw materials to the factories and moving the finished goods to the stores (China really is a LONG way away). The efficiency of an individual printer is not really a concern because, as long as materials waste is not horrid, you can just set up more of them.

REPORT | REPLY

There are 50 comments on 6 pages
1  | 
2  | 
3  | 
4  | 
5  | 
6
 | Next
 | Most Recent
 | See all

All comments should respect the New Scientist House Rules. If you think a particular comment breaks these rules then please use the "Report" link in that comment to report it to us.

If you are having a technical problem posting a comment, please contact technical support.

Printable versionEmail to a friendRSS FeedSyndicate
Cover of latest issue of New Scientist magazine
  • For exclusive news and expert analysis every week subscribe to New Scientist Print Edition
  • For what's in New Scientist magazine this week see contents
  • Search all stories
  • Contact us about this story
  • Sign up for our free newsletter
 
Password Login
Subscriptions