prize for reprap development

I'm President of the Foresight Institute, and we've had a gentleman approach us with the idea of sponsoring a prize that would accelerate technology (and ameliorate poverty) in developing countries. One possibility I suggested was a prize to accelerate Reprap development. We have a draft of a statement but it needs a lot of work. Can anyone working on the reprap project help? Write me with your suggestions for improving the prize definition. It needs to be possible in the time available, but also hard enough that it isn't likely to happen without the prize.

Here's the draft:

RepRap Prize :

An industrial infrastructure to provide the products and employment that elevates illiterate and semi-literate people in emerging economies to an intermediate level of human development can take decades to build. With the success of China in assimilating so much of the global economy’s low-cost manufacturing output, many of the world’s poorest nations have no opportunity to construct and secure their own manufacturing sector. Hence, this stage of human upliftment has become a chasm that many nations are finding difficult to cross.

But if manufacturing itself can be brought to the scale that cottage industries operate in, then the scale of Chinese mass-manufacturing is no longer a requirement to be cost competitive. A technology that removes the fixed costs and volume necessities associated with heavy manufacturing can reduce the barriers to entry for the manufacturing of many commodity goods, and drive costs to unprecedented lows. The RepRap project is a self-replicating machine that could provide a disruptive influence in democratizing access to the manufacturing of commodities. An incentive-driven approach to the invention of such a technology at suitable cost targets would yield the maximum benefit.

The key resistance points are presently the percentage of the self-replicated machine that can be replicated by a parent machine, and the availability of a suitable material that is sufficiently low in cost. Until the overwhelming majority of the machine’s parts are self-replicated, the true benefits are not yielded, and until the material used is inexpensive enough to surpass the cost barriers met by high-volume mass production, self-replicating machines are not fully competitive.

The winner of the RepRap prize of up to $100,000 will be the inventor who can make a 3-D printer that makes use of a material that costs less than 10 cents per pound yet maintains a long-term durability of (TBD on parameters), and can create a self-replicating machine where at least 80% of the mass of the replicated machine is made from this material.

Subsequently, each machine should also be able to produce items of daily utility for individual use or for export, using the same material. The deadline for this prize will be December 31, 2015.

---

(Edit: I've locked this thread to further posts. This thread is historical discussion leading up to the initial prize announcement. -Sebastien)

Edited 1 time(s). Last edit at 02/17/2010 02:10PM by SebastienBailard.

Great idea for a prize. I'm a bit worried about your $0.10/lb material, though. I'm trying to think of anything aside from sand and gravel that you can buy at that price, never mind something that you could extrude. Of the plastics out there out of which you can make things you can eat out of, you have a hard time getting much below about $2.00/lb for the raw resin.

Now recycling plastics may be the way forward there. Polyolefins like HDPE and polypropylene can be recycled with a modicum of care 8-10 times if the studies I've read are real. Factoring that into your cost structure gets you down to within shouting distance of your $0.10/lb number.

The recycling thing is something that has been discussed to death by reprappers since the project began. Basically, you need to grind the recycled plastic and extrude new filament. Both of those are non-trivial tasks.

You might want to think about whether a individual or small scale sized grinder and filament extruder have to be largely replicable with parts made by the reprap machine as well.

As for your 80% replication rate. You need to think about whether that is 80% by weight, by cost, by volume. All those numbers are going to be different.

Anyhow, I am a little worried that as you have the draft looking right now it's more of a material science project rather than a reprap development project as such.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

... this material-question is one of the reasons, i decided to build a paste-fabber instead of FDM.

My experiments with waterglass and glasspheres (later with 'rounded' sand ) and silicone-moulding shows that i can fab any complex 3D-form which is heatresistant enough for metal casting or heavy duty wear.

And the question of self-replicating ... any lowcost-CNC-mill can nowadays replicate more of its own parts, than the reprap

I think for the seeding of tooling/fabbing infrastructure in poor communities you have to combine some of the other technologies and have to spread some serious know-how of methods, materials and simplified 'workarounds' ...

I hoped twenty years ago that this sort of 'homebrew nanotechnological industrial revolution' will start in ten years or so, but now i think it's maybe some 50 years more away

Viktor

Josh - what a wonderful and generous idea!

Looking at what Forrest and Viktor have said, I think the following:

CNC-mill replication: true, though milling machines have been around for two hundred years, RepRap for four. I think we can catch up... Also subtractive technologies are inherently less versatile than additive ones, especially when it comes to multi-material products and complicated geometries; they are also more labour-intensive. 3D printing is essentially fire-and-forget. Finally the software needed to drive CNC mills is also much more complicated (because of the need for 5-axis path planning); this is still not a fully-solved problem.

Material costs: We are probably going to have to be working with polymers, not just minerals. To be useful to people in developing countries, those polymers must be biological in origin so they can grow their own. The price of pine pulpwood is about $10 per tonne (corn is about $180 per tonne), so I think 10 cents/pound should be doable.

80% replication: we find the most technically useful measure is a percentage part-count (on that basis RepRap currently comes in around 50%). But socio-economically, surely the most useful measure is replicated value? The aim should be to have a RepRap that will make 80% of the parts of its children by cost. I realize that this leads to an infinite regress: as the parts become reprapable, their cost goes down, so you're chasing a receding target. So I think we should make the criterion that the machine should be able to replicate 80% of its value, where that value is set by any and every conventional production technology at the time of the creators of the machine's claiming their prize.

So, the only changes that I would make to your draft are to replace "material" with "materials" throughout, to specify the cost of those materials as being 10 cents/lb averaged over the machine, and to say that the goal should be the replication of 80% of the value of the machine, where value is set by all competing conventional technologies.

As I say, I think this is a wonderful idea. How else can we help?

... maybe we should take a closer look on the parts the common reprap can't print: motors/bearings, electronics/wiring and all other parts actually made from solid metal.

Sometimes it's more effective use cheap and/orcommon parts e.g. bamboo-sticks instead of steel-rods or such.

We had some talk in the past about replacements for the motors (reprappable linear drives) or major changes in the setup drastically reducing the count of parts (e.g. my "Tripod-demonstrator" essentially made out of 12 fabbed parts and 3-fold symmetry or a "squid" with fluid muscles).

Many of the actual hurdles could be overcome in the next two years, so until 2015 there's the chance the reprap-design will change drastically until irrecognizability ...

Any of these 80% requirements is going to be problematic. 80% of weight can be overcome by simply printing a massive base or something that is completely unnecessary. 80% of value runs into the problem Adrian raises. What about specifying that the winner simply has to have a vitamin requirement <$X where the prizes increase for decreasing values of X. Deciding what is a substrate and what is a vitamin will probably become an issue at some point as well. I would like to see a prize for a replicating setup that included mechanisms for the production of its own power and major substrate - i.e. a bioreactor or something.

Hi,

maybe we should include all ways of measuring the percantage.?!

Like:
Part count
weight
price
volume
value

add all and divide by five.
(just an example)

Maybe that way we will can say for sure what is possible for us to achieve.

'sid

sidenote:
if we decrease partcount because of more efficient designs, we will loose partcount percentage, price percentage and maybe even volume percentage and weight,
but does that mean we have done a bad thing?
So "value" is a thing to consider, but how ??
What's the more valuable part? the rod or the cornerblock that it's fixed in?
We need both.
Any ideas?

I too was thinking you could reach the 80% just by building an unnecessarily thick base.

It also occurs to me that the rules as described doesn't stop someone from building a $200,000 machine that wins, but isn't really what you want.

How about forgetting about the 80%, and instead have something like:
1) A list of parts that the machine must be able to make, and what quality those parts must be. A complete set must take less than Z hours to print.
2) The machine must be able to make X number of these parts w/o failure. The operator is only allowed to add more filament, remove a finished part, download the next part to be constructed, and hit the start button.
3) The costs to build a second one must be $400 or less. You can use parts printed by the first one (counting only the cost of the filament) or commercially available parts. It must be able to be assembled in less than 16 hours using only the following tools: drill, screwdriver, etc (ie, no special or hard to get tools)

These are all good points. But the danger we now approach is one of over-specification: in (reasonably) attempting to set a bunch of rules that make it hard to 'cheat', we may over-constrain creativity. For hypothetical example: suppose someone made a RepRap that completely replicated all but one of its parts for $20, and the one thing you had to add was a multiply-sourced worldwide-available part that cost $100? That would be a very worthwhile achievement, and it might be a shame to disqualify it.

The other thing that was so obvious that we have all (including me) forgotten is to specify that the winning design must be open-source and free.

Edited 1 time(s). Last edit at 08/03/2009 05:15AM by Adrian Bowyer.

... i've had some thoughts abot the 'optimal' design of a real reprap - for me this could be a single but complex part out from an elastic resin resembling a tower-like structure with some dozens canales and cavities within the body.

This part is fixed to an array of valves, which could open and close the canales for pressure, so the elastic body deforms and moves the tip with the toolhead around.

With some of the canales connected to pressure sensors, you should be able to measure the distortion and actual position, so a feedback for close-loop steering is possible.

(sidenote - this sounds a bit like the nanotec-assemblers, but with elastic morphing bodies this could be made in a macroscopic range too)

So the partcount reduces to the single morphing body, the valve-array and the valves (maybe reprappable too), an oszillating pressure-pump and a fairly complex software on the microcontroller ...

Viktor

It seems to me that this prize works against open source. It will encourage people to keep their ideas to themselves to win the prize rather than publishing them to the benefit of the RepRap project.

---

[www.hydraraptor.blogspot.com]

... there should be a regarding clause, so the winning system will be auotmatically published as open source.

But you're right - people or groups hoping to win the price will start to develop in secrecy as opponents, not in cooperation with others ...

Viktor

I think that it more likely that the prize committee will be in a pickle trying to figure out just who or what team actually created the prize winning system.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Nophead's point about secrecy is a very important one.

I think that the only way to avoid the problem is to say that entries must be developed in public, leaving a dated audit trail on blogs, file repositories and wikis.

We then have a situation analogous to a sporting competition in which teams' training schedules are public knowledge. This doesn't mean that all teams will end in a dead heat.

Indeed, it would probably make the race more interesting

I also agree with rodzite that specifying a percentage of parts that it can make is going to be problematic and can be gamed.

Instead, the mass of vitamins should minimized. This avoids gaming with multifunction parts, or massive parts or lowball parts pricing. Mass can not be gamed in any way.

I also agree with Adrian that entries must be developed in public.

Much as I like and respect Adrian, I think that his proposal that progress be subject to audit is simply unworkable. For example, who decides what constitutes an adequate audit trail?

As well, how likely is someone with a killer concept likely to want to publish it when money is at stake? The moment he publishes others with more time and money than he has can take it and run. I am reminded of an old, old habit the American National Science Foundation used to have. More than a few NSF project managers there would take a nice proposal from somebody with no clout, make a copy and send it to somebody in the research community who was in a position to do the project manager some good.

I had one of these creatures tell me of the practice as being one of "the right proposal but the wrong principal investigator".

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Wow what an awesome idea!

A potential problem I see is that given the open source nature of the project, It is highly likely that the winner may use concepts previously developed by other teams. Therefore I would suggest splitting the prize to say, 60k to the winning entry and a further 4 prizes of 10k each to "significant milestones" e.g. printable steppers.... that were required for the winning entry to function. These milestones could have been achieved by other teams and would therefore hinder secrecy.

... or instead of a single price for the finished concept, maybe funding interesting evolution lines from start works better

Viktor

"who decides what constitutes an adequate audit trail" That's easy - the award committee.

Given that one of the principles of RepRap is to make its development progress public as that development is being done, and given that an obligation to do that would put everyone on a level playing field, I think that a requirement to do that is a sine qua non for any prize. And I don't think it would be difficult to track.

"how likely is someone with a killer concept likely to want to publish it when money is at stake? The moment he publishes others with more time and money than he has can take it and run"

Well - that's simply answered too: if he doesn't publish it, the money is forfeit anyway.

When you run a race, you can see where the other competitors are and so adjust what you do accordingly. I can see no difference here.

I like Enrique's idea that the mass of vitamins should minimized. Just weigh 'em and write down the number of grammes. Simple - and it can't be faked.

If you compare it to a race it would be like a relay race where only the last runner in a team would get the prize for winning.

Also one could develop in the background and then publish it later in a flurry at the end. That would make it hard for somebody to take your ideas and run with them but would be hard to prove you have not blogged in realtime.

---

[www.hydraraptor.blogspot.com]

Adrian Bowyer Wrote:
-------------------------------------------------------
> I think that the only way to avoid the problem is
> to say that entries must be developed in public,
> leaving a dated audit trail on blogs, file
> repositories and wikis.
>
> We then have a situation analogous to a sporting
> competition in which teams' training schedules are
> public knowledge. This doesn't mean that all
> teams will end in a dead heat.

Unfortunately that wouldn't help either...

Say I "publish" all my inventions on a blog a wiki on my own website.
I just have to change some variables to make sure its invisible to anyone browsing my site until I send in my finished design.
All dates are okay, noone can prove I didn't publish.
"it's not my fault" that noone reads my site actually

We simply cannot make sure that a price like this will not lead to competition an secrecy.

As long as the result will be open source and we all carry on with at least trying to improve further more in public, I think that's okay.

And what about third party people,
the guys from fab@home for example... what if THEY build a machine that matches the official rules? without making the result open source, without developing in public....
We cannot control anybody but ourself,
not even the award commitee

So either we carry on with developement like there was no pricemoney involved and see how far we get and contribute the price to what ever we choose to be worth it (wwf, unicef, childrens hospitals all over the world, fountains for africa etc.pp)
or we care about the price and start making a machine without telling anybody.

In fact there are four to five people here in the forums that would actually have a chance to win the price (I'm NOT one of those unfortunately )
Loosing they're knowledge would slow down process until the end of 2015.

Personally I think the price is a bad thing for us at this time.
Can't help it.

'sid

Edited 1 time(s). Last edit at 08/04/2009 09:16AM by sid.

Ah why make it a single shot, do it like the Nobel prizes with periodic awards?

Then it wouldn't be one big prize, but milestones of contribution to the cause. Or heck make the award as premium developed hardware that the winner can award on to deserving people or intuitions?

You know, that sort of makes sense. In fact, it makes a lot of sense. Individual contributions are a LOT easier to identify and assign credit for. A one shot award for a system which really revolutionises the field is going to be much harder, especially when it is highly likely to have a very muddled ancestry.

In fact, you could look at medium sized awards to people who've seriously advanced the state of the art with a series of innovations and sometimes just developing know how using other people's innovations. I think there would be a lot less squawking if things were structured that way.

The awards committee could then be fairly flexible about what constituted a significant contribution to the field.

That, I like. It wouldn't have the problem of tending to suppress publication because the award would be largely dependent on the work individuals had done as reflected in their blogging.

Yes!

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Given the Open Source proviso, I think it's a good idea. I am wary of it resulting in RepRap factories made using the traditional approach. This would result in the design being optimised for conventional mass-production rather than ad hoc manufacture in the field. As an example, there is a tendency to move to microelectronics boards which cannot be built using a RepRap due to their fine track pitch.

I'm also wary of the use of components that do not have world-wide availability. For example, I'd like the residents of New York, New Zealand, Senegal, Cuba and Iran to be able to assemble these things.

Finally, I am not sure that one design fits worldwide needs. I could use bamboo as structural material to help me hit the 10c/lb figure but someone in Siberia might prefer to use spruce. What we're looking at is a design system rather than a specific machine, and I think that flexibility needs recognising in the prize.

Vik :v)

Good points all.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Greetings all,

I like the idea of a prize (or series of prizes) to help motivate the development and wider adoption of making useful things (e.g. via reprapping.) As far as prize criteria go, it is unfortunately quite hard to design explicit prize criteria that do not motivate behaviors we don't want. (Assuming, for the moment, that we even agree that secrecy or adding a massive, otherwise useless base to a reprap are behaviors "we" don't want. FWIW, I don't want to motivate those....)

If explicit prize criteria cannot be formulated (without incentives for bad behavior), I think the best that could be done is to have a prize committee with good judgment and minimal bias. Sometimes a little "strategic ambiguity" can help avoid people gaming a system. Properly selected, a committee could recognize significant progress and reward it.

I was motivated before this discussion, and I've seen that many others are as well. However, a prize -- and its attendant publicity -- could motivate other clever folks to think about and work on improving reprapping technology. Increasing the number of people experimenting would be a win, no matter who won the prizes.

---

Larry Pfeffer,

My blog about building repstrap Cerberus:
[repstrap-cerberus.blogspot.com]

The Netflix prize was won by a team which was actually a collection of teams which pooled resources. I have a feeling this prize would be won the same way.

[reprap.org]

Can Mendel print out 80% of its own parts? How close are we to winning this?