Open Source, self-replication, and general project orientation (WAS Re: RepRap or conventional manufacturing cheaper?)

Hi all,

Quick introductory note:
I'm new here. This is a cool project and I'd love to get up and running with one of these soon. I'm on a tight budget, so I identify with some of Alan69's concerns. However, I'm also a linux/open source programmer (for a living) and I can identify with both the philosophy and practicality of the ideology involved. I wanted to give my 1 cent regarding why I thought this project was worth getting involved with, and perhaps address some of the other viewpoints I saw. And of course, I am biased, confused, forgetful, arrogant, tunnel-visioned, and a woefully ineffective apologist, so I apologize in advance for whatever wrong or offensive I will say.

The previous thread, in talking about which was cheaper of mass production or raprep custom, discussed materials costs, qualitative value of custom, raprep creating mass production machines or parts, raprep being able to coexist (symbiotically or juxtapositionally) with a mass production methodology, predictions, and so on. In essence, the thread was about, "Hey! It's exciting thinking about the possibility of this scaling up. What could be the implications?" But it got mired in questions of project viability concerning market costs and cost of acceptance, and the equivalent to the question, "Which is better ROI: Linux or Windows?" (which has been argued adequately elsewhere).

The project's worth, however, is not solely economical as was hinted at. I'd like to bring that aspect into full light. This project is both open source, and has self-replication as a strong goal. Why does that matter?

Why do I use open source? Because it's free as in beer? Yes, though sometimes I would have been more productive buying a commercial product when something similar in open source was less mature. I still went with open source in those cases, though. There's something more to it than free-as-in-beer.

Free from corporate control? A little. It's a selling point to clients not to be vendor-locked. I definitely enjoy not feeling the stamp of marketing and advertising on every product I use, having project pages as (or more) willing to tell me what is lacking in a project as what is working, and so on. I also enjoy the communities open source fosters more than most corporate PR.

Putting the power in my hands? Some. I rarely compile my own kernel or use custom compilation flags. I rarely dig the customization depths of linux. But it is nice to know it's there, and that I have access to it all, no matter how deep I'd like to go. It's like having an old car where all the parts are accessible under the hood -- ready to be worked on, but more importantly, learned on should I wish to.

That leads me into the larger reason why I go with open source, and why this project is so encouraging to me: I am sick of the (probably unintentional, side-) kind of quality that a company puts out. Let me illustrate. I get discouraged that my hand-painted wombles aren't as uniform as a factory's or that my cuts aren't straight. I get frustrated that I couldn't make one of these myself -- that it would require a die, a press, a foo. I'm disappointed that I can never come up with the right thing to say like the engineered situations in movies. Perhaps it's a personal flaw of mine that I feel this barrier to my creativity and expression.

Open source makes it something I can do and contribute to. I can learn from it -- all the parts are there for my review. The insides aren't closed to me, or even made inaccessible by dipping them in epoxy. I'm not assumed to be a luser. I can contribute back, even if just by finding bugs and reporting them. I can be part of the process and not just complain if something doesn't work. This last part always surprises me -- so many people get so frustrated with their computers because they're stuck with what they have (buggy, company won't fix or in a costly upgrade). The reflex to be frustrated with my situation isn't needed anymore and this positively benefits the rest of my life.

So, in my mind, there is some benefit to the whole whatever that open source brings to it. Even if it is less economical and even if I don't customize (which I usually don't), I still do open source when I can because I'm not cut off from the product, can be part of the solution, and can learn.

And sure you can do those things by hacking yourself a custom, non-open-source solution, but not only am I not currently at that point of being able to (just like how I'd have to learn cars by having one first and working on it rather than building one from scratch), but also that's similar to writing a program in assembly language. Entirely custom. Can't be run on other platforms. Probably faster (or in this case, less expensive) sure. But far less scalable. And yes, there were many companies using PDP-11's (there are many Lowes stores) but nowadays the worth of compilers and programming languages is undisputed.

This brings me to the other asset to this project: the goal of self-replication. Compilers can compile themselves, or the next generation of compilers. Generally this is thought of as a Good Thing, but not one that your Jane Schmoe would care about, and isn't particularly life changing. Why is that helpful? Well, could you imagine continually having to program your c compiler using python (or vice versa)? Why have to know two languages? It is so much more efficient to have to deal only with c. It's only one language to learn. You can change your language if you need to. It demonstrates viability.

Why is it important that raprep can make rapreps? Because rapreps are only dependent on rapreps (which is important sometimes). Because the developers don't have to think about trying to debug everyone's custom setup. Version numbers can be submitted with bug reports. Devs can talk "one language" instead of having to worry about suppliers, etc... As has been pointed out, it enables community participation. Because I can do it and not be dependent on possibly-not-there companies.

But it goes beyond that, though. When programs started being able to maintain themselves, download their own updates, etc... that was a very Good Thing. It took a while for software to get to that point, but it was wonderful when it happened. When installers became so user friendly as to only ask me the questions that needed to be asked (but allow for further customization if needed) that was a Good Thing, but it took a while for that to happen too. I look forward to when you can build a raprep, with its own assembly robot, perhaps with some custom options for if you have materials you supply from elsewhere (like Lowes). I look forward to having an upgrade bot, that allows me to just click yes to the question, "Do you want to install the update for 5.0.2?" and the thing does it itself (instead of the software equivalent of having to patch the sources, reconfig, recompile, and reinstall by hand (though I regularly track wesnoth in svn)). I look forward to not having to scrounge for all the parts (hassle hassle hassle), but rather having everything available in my parts library. I look forward to "Could you make me one?" and "Why sure, let me rep you one." And (though this doesn't have anything to do with replication), I look forward to having a scanner hooked in with one of these to make a spatiocopier.

But most of all, I like that (in both being open source and self-replicable) it enables and empowers me, I who don't yet have the skills at hacking hardware together. This may be rather hobbyist of me, but that's what open source has made me.

-- Erich

For me, there is a desperate importance that the first domestic 3D printers are OpenSourced. If a commercial company makes one - they are pretty much guaranteed to do one or two of the following:

* Sell the machine at a loss - and make profit on the raw materials.
* Sell the machine at a loss - and make profit by selling heavily locked-in DRM-protected 3D models.

The first is the model by which inkjet printers are sold - the second is the model for computer game consoles.

Neither are very nice for the consumer. However, once a market takes off with one or other of these methods, it'll be very tough indeed for an OpenSource gizmo to break into the business. This seems surprising - but see how hard it is to give Linux away for free when Microsoft are out there monopolising the world of computer software! People will happily pay $200 for a copy of Word and yet be unable to give you a single reason why it's better (or even different) from OpenOffice!

We need to have cheap feedstock - and we need an open/free source of 3D models that work on commercial and open hardware. The only way to do that is to get there first with OpenSource and truly make the machines so cheap that a commercial organisation can't compete.

If we can make it - then we'll have designs for all sorts of interesting products being given away for free on the Internet - people will accept a 'Wikipedia'-like approach where people can improve on designs cooperatively. If we don't make it then it'll be DVD's all over again (You can't buy a DVD in England and play it on a US DVD player) - and they'll play the same tricks of expecting us to buy the same set of designs over and over again as new machines appear (can you say "Blue-Ray"?)

We also can't afford for a big name printer manufacturer to sign a big contract with materials producers and prevent raw materials being sold at a market rate. Imagine if your car only ran on "Ford-brand" gasoline?! It seems unlikely - but that's exactly what happens with ink-jet printers. We pay $50 or more for a few cubic centimeters of ink - and the printer manufacturers use the DMCA to prevent people from reverse-engineering their cartridge designs.

This must not happen this time around - the stakes are too high.

Hey Steve,

Obviously I agree with you on the fact that this stuff needs to be open source, heh.

Unfortunately, #1 is the reality right now. Well, that and they don't sell them at a loss. A Stratasys machine will set you back $25,000 MINIMUM. They also charge about 10x the price of the raw material.

Fortunately, nobody will ever be able to 'sign a contract' on the raw materials. Once we find a way to reliably print with HDPE, we're golden. HDPE is *very* ubiquitous, as are a variety of other promising plastics. There are plenty of suppliers, and plenty of independent shops that are willing to turn that plastic into filament for us. After all, our money is just as good as money from a large manufacturer of RP machines.

The real issue we face is making this technology work. It's pretty tough, and based on skills that you don't really learn in high school. However, we're making good progress and its currently within the reach of a passionate and/or talented person.

As for DRM laden 3D models, thats something we can easily get around. There are tons of open source 3D modeling programs, with large communities behind them. As soon as we get many machines up and printing, I imagine they will want to collaborate quite a bit.

Yay!

HDPE ubiquity: yeah it is. Anything recycling number 2 is HDPE. I presume anything used for food contact (e.g. milk jugs) could be melted and used again in that capacity.

Hello guys, a small introduction: I am a RepRap fan and I support you 100%. I am in favor of a post-scarcity society and plan to contribute my time/talent/money in order to achieve our common goal. I am a computer science student, but not advanced enough to start making a RepRap myself. I am an open-source supporter and regularly put links to CreativeCommons on my sites, blogs, social networking space etc. I plan to print cards with links to post-scarcity ideas including this site. Therefore I hope to attract some smart hardware engineers to this project.

One suggestion to all those supporting this project: don't be a passive supporter! Go out and do something! Spread the meme! If you're talented enough try and build a RepRap yourself and improve it.

An advice to the developers: create student RepRap workshops- you definitely need to increase the number of developers.

Anyway, cheers to y'all and I hope to contribute to our cause of creating a moneyfree world as soon as I can.

There is no such thing as a post-scarcity society (rich people want stuff they can't afford too). There is no such thing as a money-free world as long as you have trade (money is just that commodity which everybody takes in trade -- in the future HDPE maybe?). I appreciate enthusiasm for the reprap, but ignorance of economics is just a shame.

(1) There IS such thing as a post-scarcity society (at least in theory). You might want to read up a bit. Please reply only after you do so.

(2) Actually once the fundamental resources are practically free (energy, matter and intelligence) and I don't see why they can't be, a gift economy will take over. See [en.wikipedia.org] for more information.

(3) If you people disagree with these ideas then please remove the quotes from the main wiki and change your slogan.

Hi Damien,

... actually we live in a market-driven society, so money rulez ...

If through technical, ideological an social evolution a broader self-reprapping basis would be seeded, then the market-structures would be massively changed:
- all material parts and usefull hardware, which the people build at home or in local community-fabbing-centres goes away from the normal building-selling-reselling-buying-paths, so the market would collapse in this regions ...

In exchange there would be an evolution of raw-material-reselling companies, but this is another branch ...

So some companies are aware of the comming 'wind of changes', but actually there is to much turbulence out in the world, to realize the real consequenzes of such an process ...

Viktor

Cory Doctorow's first book "Down and Out in the Magic Kingdom" has an interesting take on what economics is about in a 'post-scarcity' world. The idea is that you have your reputation and thats what you value so you do things that will benefit others and they will increase your reputation for you and make you 'richer' becuase of it.

All economics at the most basic come down to trading the one true scarce resource we all have, our time. This is why you're willing to take your car to the mechanic and pay him for one hours labor because you know it would take you 3 hours to do the same thing.

You build up your skills and specialize and you can do quite a lot in one hour in that narrow field of your speciality compared to everyone else around you.

In the context of RepRap, the post scarcity world is one where most material things are made at the push of the button with little cost in the way of materials. So people will instead charge for the time it takes them to design the product, troubleshoot issues, advise others on what products will be needed in their unique situation etc. which could give rise to a gift economy but probably only after a time of transition to a more explicit time/skills based economy.


Essentially you take the software industry model of making money to the hardware world (hopefully using the open source software industry as a model rather than the closed source+software patent model).

Once you have the open source ethos in the culture I would like to think you would naturally transition to a gift economy but I don't think todays 'modern man' has enough sense to see beyond the whole 'I've got to have a measure which says that I'm better than the next guy' to 'If I do something that benefits me and him, we're both better off'

I agree with what you wrote. And yes, when I said that a PSS was developed in theory I had Cory Doctrow, Neal Stephenson, Richard Stallman, Ray Kurzweil, Eric Drexler, Marvin Minsky and others in mind.

There's no doubt that the transition to a gift economy will be a painstaking process. Hopefully we'll have organizations conscious enough to spread the technology throughout the planet. After all we have enough food in the world to feed 2 times the current population and yet there are many people in developing countries dying of malnutrition. Proper planning and organizing is everything!

Hello folks,

It would seem to me that the current manufacturing model is efficient in optimising production efficiency but is lacking in allocative efficiency. In other words, you can have any colour Ford model "T" you want so long as its black.
I do not see 3D printers competing against the equipment required to produce plasma screens, nor do I see what is essentially a very unspecialised machine vying with specialised production machines.
From an engineering perspective, the validity of the RepRap concept stems from the ability to locally produce artifacts whose design is driven/shaped by the end user to fit a specified function. This is something which industry can not yet compete economically against.
To coin a phrase, I see an evolution of "pull" design filling the niches which "push" design cannot economically accommodate. My interpretation of "pull" design is an attempt by the end user to create something to fill their specifications. Conversely, "push" design is the fruits of market researchers, lead user analysists and fashion predictors to anticipate and create a product to fill a perceived niche.

Toby Borland, mechanical & manufacturing engineer

"Unsure of what my employers would make of my opinions"

Toby,

I may be wrong--have been a lot!--but it seems to me that no one else is able to do room temperature printing of HDPE. Now, I may just be blowing smoke but isn't this true? In that respect, the RepRap is quite sophisticated all on its own regardless of what other machines are capable of doing.

Also, consider what it will be able to do with further investment of time and money. Sure, right now it prints plastic in a half-ass manner. Five years from now it will be printing plastic admirably, circuit boards well, and individual components in a half-ass manner. Ten years from now...you get the idea.

Great innovations don't have to happen all at once. That's like nocking a kid for not knowing everything when they are three. There is great potential there if we let--and help--it develop freely.

Yes, it fits the personal customization paradigm well, but that isn't the only thing it will fit well given time.

Demented

Demented Chihuahua Wrote:
-------------------------------------------------------
> Toby,
>
> Now, I may just be blowing smoke
> but isn't this true? In that respect, the RepRap
> is quite sophisticated all on its own regardless
> of what other machines are capable of doing.
>
It isn't clear that WE can print HDPE at room temperature just yet. I've had some success so far, but I haven't got past a lot of the problems by any means.

Mind, to the best of my knowledge none of the commercial firms making 3D printers are trying to do FDM at room temperature.

... most of you are focussed on extruding melted plastics, as some comercial 3D-printers do, but there's much more possible, if you let them some time and seed helpful hints

In the image you can see some of my 3D-manufactured parts over the last years.

- In the foreground there are milled mini- and micro-parts, i realized on 'low-cost' 3D-mills from ISEL (the first one at 1986 for 2500 Deutschmarks, the actual for 6000 Euros) ...

- In the back are some 3D-printed parts from different materials ...

Only the red part is an FDM, the others are from powder-printing or STL-epoxy.

The difference for me, between commercial and home-brewed manufacturing, is not so big now, as twenty years ago - i'm on the way, to make my own 3D-printer/mill with micro-accuracy, and here i want to use every material and technology, i want.

Actually i'm on 'room-temp'-printing with laser-sintered pastes from different materials (gold, glass, ceramics, plastic, ...), the next task is stacking of thin sheets and laserdiode-cutting the contures of the slices (=>3D-LOM).

In some project-groups i'm on evaluationg of micro-fabrication with nano-parts, so it's not so far away (maybe 2 years or so), when i'm capable of micro-fabbing at home - and with the ideas of a 'mega-reprap' with a working-area of 2 cubic-metres or more, the greater dimensions are on the way too ...

So it's NOW, we are on the edge to self-manufacturing of sophisticated parts at home!

And this is also my main interest in the reprap-forum - here i can exchange ideas, possibilities and other stuff with people, that are not so money-driven, as the 'professional' folks

Viktor

Hi Forrest,

Forrest Higgs Wrote:
-------------------------------------------------------
> Mind, to the best of my knowledge none of the
> commercial firms making 3D printers are trying to
> do FDM at room temperature.

... as i know, work most other 3D-printers at room-temp, but not with FDM:
- Z-Print prints thinned cyanacrylate from an inkjet on a powder-surface
- Objet prints small droplets of epoxy and harden them with an UV-lamp
- STL-epoxy-printing is at room-temp too, the UV-laser for hardening the structure is 'cold'
- the EZ-LOM-printers stacks the sheets with a solvent and cut the contours with a knife, so no heating is necessary
- paper-laser-LOM is cooled down, so the paper wont burn, when the CO-2-laser cuts wider areas
- SLS-printers (Selective Laser Sintering) melts the powder just in the spot, but the overall temperature is under 30

Edited 1 time(s). Last edit at 10/11/2007 02:02PM by Viktor Dirks.

The capabilities of the machine are evolving, as are the capabilities of manufacturing. My thread response looks at the niche of a home 3D printer/fabricator within a wider socio-economic context.
The potential of a machine which can drive its own refinement is also interesting; not to mention the interaction at the level of the forum. This is, in a sense, one of the first distributed hardware design communities not wholly based in academia or industry.
As with most open source developments, it is probably impossible to isolate one overarching motive common to all developers (excepting possibly the sheer hell of it).
I have wondered how critical the easy transmission of code/information/software is to the success of the open source movement. Now that *.stl files can be transmitted and executed in a similar fashion will this aid the open hardware cause?
My views in this thread are purely philosophical, I certainly do not wish to circumscribe the potential of the RepRap hardware here.

Toby

Toby,

Nor did I think you were attempting to circumscribe the RepRap's potential. However, I just don't think its accurate to say it doesn't "measure up" somehow in light of other things or within another context.

The computer control system on something like the F16 is incredibly complex and fault tolerant even though serious work hasn't been done on its development since 1996 or so. Does that mean it isn't sophisticated because we have Palm computers with more processing power? Certainly not. They are two different contextual frameworks and must be evaluated on their own merits and in their own contexts.

I think what I wanted to avoid is comparing apples to oranges with the RepRap. It doesn't make sense to use or build a RepRap if you have a huge machine shop at your disposal (excepting, as you said, for the hell of it).

I, for one, am most definitely in this for the hell of it. It just sounds like fun, philosophy aside.

Demented

There's a lot of theorizing about post-scarcity societies. There's also a lot of theorizing about zero-point energy and perpetual motion machines. It's all fantasy. Scarcity exists because humans dream beyond their abilities. The only way to eliminate scarcity is to stop humans from dreaming (I think they call that "hell") or to give humans unlimited abilities. And yet we live in a society where our abilities are limited by our need to interact with other people. Thus, when you talk about a post-scarcity society, you are not talking about a human society.

But ... about the reprap. Compare it to a manufactured 3D printer, such as the one that Singer will come out with eventually. (You know Singer will do it first because they already have a 2D stage that uses thread to create things; not far from that to extruding threads of plastic.) Who wins?

Cost: a plastic part manufactured using an injection mould will always be cheaper than a printed part. Motors are motors; it will be a long time before they can be printed. Cost to build will always be more expensive for a reprap.

Price: Parts have to be assembled. So, the price of buying a reprap depends on how much the builder values their time. The overhead of a reprap business is pretty low, so a person with a reprap never need to be completely unemployed. They could sell repraps without having to invest (risk) much money.

Time to market of a new design for a 3D modeller: The reprap wins this one every time.

Standards compliance: A reprap is likely to win this one as well. Look at how many filesystems and protocols Linux supports.

Accuracy and precision: These are a function of the cost. The big win goes to the reprap if a Generation N reprap can build a Generation N+1 reprap which can build parts with a better tolerance.

What other aspects am I missing?

Hi Russ,

... actually the reprap is more like a toy for a hobyist, but in near future it will outrun most other manufacturing systems, as it's evolving in an exponential growth-curve!

In 1986 i bought me an 3D-mill and because of lack for appropriate software for my AtariST i had to write my own 2,5D-CAD and the drivers for the CNC too (i have it running until now on an old ST, so here is a possible crosslink too )

Now i'm on discussion with some guys, which could develop for me a neural-network-simulating-software for a kind of 'self-learning' controller, which will then be able to run every hardware with every kind of motor (stepper, servo, hydraulic, ...) and any kind of mechanics too (cartesian, scara, 5D-robot-arms, tripods, hexapods, ...) without further software-development.

With playing around with different 3D-methods (extruding, lasersintering, laminating, milling, ...) -- possibly on the same machine -- there would be no limit in capabilities, and the size and accuracy can easily be defined through the scaling of the mechanics.

As with my tripod- or a more sophisticated hexapod-rod-mechanics there would be the possibility to assemble the mechanic in a half day, so i can build 3 or more sytems for every method, material-branch or geometry, so this limit is practically eliminated too ...

The 'self-learning' software should be open source, so the costs for a new reprap will define mostly through the tools, motors and electronics (until the reprap can build them on his own too), so in developing of the software is for me the most powerfull path to rapid progress ...

I think, with the right controller i should be able in a half year to build systems with every accuracy and capability i want, in respect to the outputted material - and here i can brake the actual limits with laser-sintering of pastes ...

Viktor

Mr. Dirks,

Those buckles in your picture, they're ubiquitous. Were those your design?

Hi Sean,

... call me Viktor, it's more comfortable for me

If you meant my avatar-image, then yes, i created some 3D-objects and scenes for fair-presentations and articles.

It's no real problem for me, only some time-consumption, to create virtual scenes with available or selfmade 3D-objects and output them in photorealistic iamges or animate them and make small clips - there are some open-source programs like PovRay or Blender, but i use mostly Lightwave or some older 3D-programs, which weren't for free.

If there would be interest, i can make 3D-objects and scenes for your repraps and different designs too ...

Viktor

Hi Sean,

... sorry, i had to look in my dictionary, to sort it out ...

The 'buckles' are my designs and prototypes of micro-grippers and tool-tips for microassembly, but you can find similar designs from some german universities too - it's a generic 'parallel-kinematic'-structure and i was in cooperation with some groups, but the basics are much older, so it's not my personal invention

The clear parts in 'STL-Greifer2_kl.jpg' are from epoxy, made in STL on a professional 3D-system (as technology-test for free!).

The other parts, as in 'Greifer-Spitzen2_kl.jpg' i milled on a ISEL-3D-mill, which i had optimized for higher accuracy ...

Viktor

Ah. I thought they were belt buckles.

Sorry, I should have known better. I can think of at least one example of what I mistook them for, that was manufactured on a large scale back in the 80's. I thought they were the type of plastic buckle sometimes used for backpack chest straps and child seats.

I see, from your close-ups, they are more along the lines of robotic grippers, which are slid in and out to articulate. A nice design.

On your single example, it's not clear to me, outside of the elasticity of the joint, what makes the claw open back up when you press the plate forward, though. It looks as if, once fatigue set in, the claw would extend, but not open.

Hi Sean,

... it seems, i misinterpreted a second time

Then a bit more explaining:

- The white big 'buckles' at the very left is a 'normal' child-seat-design, but made from stark with cyanacrylate for elasticity-tests.

- The red 'buckles' (or clamp?) beside is a generic sample/giveaway from DimensionPrinting made in red ABS, i received as test for accuracy and stability, it was made on an Alphacam-3D-printer without any support.

- The gripper-designs with the rod-pairs (clear= 3D-printed in STL-epoxy, white= milled in DELRIN/POM) are relative strong and elastic enough, to re-open properly, when released - but the 3D-printed are not so smooth at the outer side, so they would 'ripple' a bit, when the closer move back- and forward.

- The plastics- and aluminium-toolheads with the flexible hinges (nearly 50 micron thick) are very soft-elastic and i force them with a slider to close and open them appropriate ...

- The other parts were 'proof of concept'-designs or prototypes for different applications, as peristaltic pumps or magnetic driven and ferrofluid-feeded micropositioners ...

Viktor