World-wide reprap printing webstore

I've been thinking (which is my way of avoiding doing) and thought I'd share my thoughts...

It looks like in the next few weeks/months we'll have maybe 4 people able to print/supply reprap parts (Zach, Ian, Vik, and nophead) so it looks like now would be a good time to get together a reprap print ordering store with the option to choose who prints out the parts for you. I know Zach was looking at putting this into his store but I'm not sure how far thats come.

What I'd like to see is a combination of what we have now with the BitsFromBytes and RRRF store as well as what Ponoko are doing (http://www.ponoko.com/).

The user goes to the store and can upload their own .stl file(s) and set their own profit, or download others .stl files, or (the primary purpose of the site at this point in time) get printed out one of the uploaded pieces/projects by somebody with a working reprap.

Once you've chosen the project you can choose the reprapper of your choice - the closest to you probably (using a mashup with Google Earth and some shipping cost calculator) who have each set their own profile with price per cc and any additional info needed such as materials used, printing speed, any delays in printing/shipping etc.

Then you pay the price agreed and go on your way (possibly choosing to download the .stl file and any instructions before you go). Behind the scenes the website will divide up that total payment between the various players and everyone will be happy.

Eventually, as repraps get more common the additional cost that each reprapper can charge for printing things out will go down to the cost of the materials but at this point I don't think anyone would complain about the cost being the cost of materials + 10-20%. This gives everyone with a reprap/mold an incentive to participate and this cost also undercuts the current 3D prototype places by a considerable margin.

Ultimely as everyone has a reprap of their own the usage of the store as a place to find a reprap machine will decrease and the store will become more of a place to get and share designs.

The issue I see is that the uploader of the .stl file will have to trust that the individual reprappers who are given the task of printing out the file don't just upload it under their name and undercut them.

If the act of uploading required the creator to agree to a GPL type license then this would take away the legal implications, and if the profit is more a voluntary "suggested donation" and there is some sort of reputation system then this takes away most of the other concerns.

It would also be nice if there were a nice hierarchy of donation schemes as well. If you incorporate somebody else's design into your scheme you could for example donate 10% of your donations to them.

It kind of hard to bring in capitalism into this project but I think its about time we had some way of rewarding the people involved; especially as this will mean a sacrifice of time for them.

In fact I'd encourage people to go to BitsFromBytes and RRRF now and donate to these guys who have made this project so much easier for the rest of us.

Most of us aren't in this for the money but we'd appreciate it :-)

Also, I think the reputation building part of this is quite important. If we have transparent feedback we can build up trust in each other as individuals and people who trust each other always work better. The prospect of ego-stroking on a world stage probably also makes people more productive :-)

Ok, so what do you think? I have absolutely no idea how hard this would be to do but I think its the next step forward. Do we add to one of the stores we already have or do we create a new one? I'll leave that to the web-devs to figure out.

Edited 1 time(s). Last edit at 04/12/2008 12:55AM by reece.arnott.

I think you are missing the point here. I don't think anybody is going to spend their time churning out large numbers of parts. Each person only has to make two extra sets and pass them on and within a few weeks everybody who wants one can have one due to the exponential growth curve.

If we need to grow faster, perhaps the first people make three or four sets but very soon supply will outstrip demand.

We don't need a store for parts, we just need a waiting list for people wanting them, split by country to minimize postage.

---

[www.hydraraptor.blogspot.com]

Quote
nophead
within a few weeks everybody who wants one can have one due to the exponential growth curve.

More like a few months. A "generation" seems to be taking at least a month to reach reproductive maturity.

I think the idea is a good one. It was actually similar to what I wanted to propose. It's not just about making RepRap parts available to those who want a RepRap. Anyone with a RepRap could sell services and earn back some of his expenses if he wants so. It's usually way too much trouble to start a webshop yourself, but for this there could be a central (middleware) platform on which demand and supply can meet. Imagine that the platform contains an ever maturing and expanding base of objects that can be printed, you can order more and more items from a growing number of outlets (RepRappers that sell their services). Objects can be in themes: gadgets, jewelry, machine parts, robots, etc. and can be composites. Objects can also be composite objects that may need to be assembled. A site that hosts just STL files is inherently incomplete: any normal product you buy today except very simple/homogenous stuff (e.g. sugar) has a usage description with it (or assembly guide for composite objects). A product is more than just the object. The quality of how-to's is very important and will help determine the popularity of products.

For this shop system to work, it would be best if an order could be put into a print spool when paid for (and perhaps reviewed), but it would not necessarily require much work, even for many small parts. For printing an entire RepRap a spool system would be very desirable anyway... It can be as simple as a folder being monitored for content and if it has correct .stl files (perhaps with a corresponding .ini file for print options (print orientation, material, etc.)). The spool can be filled by a Shop application that's mostly web-based.

Don't see this as the only option to 'seed' RepRaps initial growth stage. Other options would still work (just one-on one agreements to make parts for friends). I do think that the platform could count the number of RR parts made for another person and that you get 'karma'/credit for it. Many of us would favor recognition of their effort over a financial compensation. This platform will become important after there are some more RepRaps around and a proportion of the RepRappers who are willing to put things in boxes and ship them once in a while.

Regards,

Erik de Bruijn
[blog.erikdebruijn.nl]

There are a number of levels I can see this evolving...

I see it initially as a way of getting people involved who would not otherwise by simple serendipity web surfing. Such as those who are willing to design innovations or custom things for themselves but who don't have access to the facilities to print them.

In the short term it would more be for those who are not as interested as the current crop in actually creating a fully fledged rapid prototyper but I can see those who 'come for the food, stay for the show'.

Its also a combined front for all of us in our scattered homes to say to the world 'we're here'.

Yes, I think the best idea would be the classic give one to a friend idea (particularly as then you'd be on hand to help out) but I think the people of the world currently have a lack of local friends with repraps :-)

There is always someone who wants the benefits without doing the hard work... so offer value added services. Offer a ready made reprap set up and installed for a basic price plus commuting time. A lot of people are sure to take you up on it in the early days.

The thing I see as the most immediate plus is that it would give people with repraps with spare capacity something to print and use the machine to its potential. If you've used up the pool of local friends who are interested in repraps and your reprap is sitting idle while you're helping all these friends build their own, you can print a set of parts for someone else , maybe in the same country, maybe not, who doesn't have any friends like you but wants one anyway.

Obviously at the start we'll all be devoting a large part of the worlds reprap printing power to creating more sets of parts but as the number of repraps gets to the point where ther are a couple in each country (thats less than 500 by my count of countries) I can see a shift from 'getting the word out' to finding things we can use this technology for (including tinkering to make them better). At that point we each have two repraps in the garage: one constantly printing out the next set of parts, one being your hobby business making knick-knacks for the public or experimentation. A few months later you have a group of you doing that and coming together every few weeks with cool new designs for making it so much better/faster/easy to assemble etc.

By the time we get general awareness in the world at large, the same web-infrastructure allows the creating and sharing of the best of these ideas with that selfsame world at large.

I see this as simply the next step in what we're doing. If we do it right, we can set it up one e-commerce website (a specialist task) and each use it as we see fit. If we don't I'm afraid we'll each re-invent the wheel if/when we want a presence on the Internet.

I see it as equivalent to blogs. Most people don't know how to (or want to go through the hassle of) set up and configure a website to act as a host for a blog but a whole lot of people use blogspot etc. to have their blogs.

Why not a reprap oriented e-commerce website? Probably grown out of what we have now.

The limits of this exponential growth we're wanting are the limited number of friends each person has. Make that potentially everyone connected to the internet and things could just explode!

Ever seen the Google 3D warehouse: [sketchup.google.com] ?

Imagine a "get this 3D printed near you" button.

I would certainly be interested in helping with this but think it may be a little premature as the current RepRap is taking many hours to replicate Vik would give a better idea but i expect it is in the region of 300 - 400 hours (a corner block takes 10 hours) I don't think anybody leaves their RepRap unattended during build so these parts require a non-working day / very late night to produce.

I would expect a complete replication to be it the order of a few months minimum when everything is factored in.

Nophead has just started this process i believe so it will be interesting to follow his progress as during Vik's replication he was doing a massive amount of debugging and therefore timescales are not indicative, what we will see very soon from Vik is how good a true RepRap Darwin is and how reliable the parts are when made in PLA (very exciting times )

---

Ian
[www.bitsfrombytes.com]

I agree that it will be some time before the RepRap produces more easily and of better quality than with moulds (if you include the mould making process as an investment).

What I'd really like to see is that it is possible for anyone (if they like) to use his machine's 'idle time' to create parts for (themselves and) others. Essential for this is:
- Safety (fire, anything else?)
- A print spool!
- More reliable design (especially the extruder should have good up-times)
- Early detection of failures/problems (high temp GM3?)
- Dedicated "RepRap" server: E.g. Monitoring through a USB web-cam. The server could be controlled (and reprap monitored remotely through VNC)
- Warning system (SMS message on failure?)

One of my goals will be to make the RepRap run unattended. I've done a few steps to prepare this:
- Bought an optical smoke detector
- Bought more temperature sensors (microchip sample) for checking motor and board temps. I hope to run a small I2C network to read motor + board temperatures (if they get hot, I've just started constructing it)
- Planning on mounting a Fan to the PIC based PCBs and perhaps the GM3 motor
- Add some heat sinks to the stepper motors

You should be able to manage (and prioritize) a list with print jobs. A rudimentary version should be very easy to implement (just add .ini files with orientation, (x,y)-coordinate, materials (usually just extruder 0), fill rate, speed settings). This print spool and unattended printing are an important part of the bigger picture of world-wide 'generic' fabrication capacity!

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

If you have every Darwin generate and redistribute two sets of parts you take two months to make the parts and probably another two to ship them and have them made into working Darwins. That's a pretty generous estimate, mind.

Given that you'd see...

month 0 - 1 Darwin generating parts

month 4 - 2 Darwins generating parts

month 8 - 4 Darwins generating parts

month 12 - 8 Darwins generating parts

month 16 - 16 Darwins generating parts

month 20 - 32 Darwins generating parts

month 24 - 64 Darwins generating parts

month 28 - 128 Darwins generating parts

month 36 - 256 Darwins generating parts

That's not a really wonderful growth rate, if you think about it for a moment. Being serious, I suspect that the growth rate would be considerably less than that, if Vik's experience with how much trouble it is to print out a full set of parts is any guide.

We need to have a printer that requires either a much smaller volume of parts to be printed or one that runs a whole lot faster. Otherwise, replication is going to be a curiosity instead of the way that business is done.

Well, I'm not quite as pessimistic as you Forrest.

I think that the average replication time could probably be shortened to 3 months instead of four, considering the first kit to get printed will be in assembly while the second is being generated. Also, there are at least seven or eight RepStraps being made from cast parts right now as well as more castings being made - we'll say those become available at a rate of 5 a month, just as an estimate.

Month 0 - 8 printers (not necessarily darwins) generating parts - 8 in existence

Month 3 - 31 printers generating parts - 39 in existence

Month 9 - 77 printers generating parts - 116 in existence

Month 12 - 169 printers generating parts - 285 in existence

Month 15 - 353 printers generating parts - 638 in existence

Month 18 - 721 printers generating parts - 1359 in existence

Month 21 - 1457 printers generating parts - 2816 in existence

Month 24 - 2929 printers generating parts - 5745 in existence

As you can see, while printers-from-replication (exponential increase) quickly outstrip the production of printers-from-casting (linear increase), their ability to seed the system has a huge effect on the rate at which replication is able to increase in the first few months.

This also assumes that on average each new Darwin prints two - I think that's a fair estimate, because some will print more, and some won't print any at all.

Once you get out past two years, I think that there will be enough Darwins to get one to most anybody who wants one. Of course, by then, maybe it won't be Darwin anymore at all (yay Mendel)!

I thought the parts were about 1Kg. In which case I should be able to print a set in 88 hours at pi mm3/s. No reason why Darwin could not go that fast with Python driving it.

If the parts were made in ABS or PLA then I expect most of them could stand being made at 50% density so we are only talking a couple of days. Say a week for doing it in evenings and weekends. So one month to build, 2 weeks printing before the next generation

---

[www.hydraraptor.blogspot.com]

Best estimate I've seen is about 1.4 kg for a parts set from Ian. Doing 50% density should indeed half the print time. I wonder, however, if you can do those nasty corner brackets at 50%, though. They have so many holes going through them at odd angles that it could be a bit tricky to get the strength right.

You're forgetting the part where the inter-generation shortens as documentation gets more intuitive and better (more robust, simpler and easier to construct) designs are made (less trial and error needed before they can also print). Also a printing speed-up with upgrades will help. This makes the growth rate increase. Growing complexity might be an inhibitor to the growth rate *increase*.

So can we be even more optimistic? Maybe, but know there's a bigger picture. If you want to model this growth well, you have to take into account not only 'replication potential' but also 'diffusion of the innovation' in its market. Open source software has the same potential to be on everyone's computer since it can duplicate unlimited. Yet, the general majority doesn't have open source software (except libraries and some exceptions). This has everything to do with 'usefulness' and 'awareness' and other adoption factors. One of these factors will not do: Even a superior product need not be on everyone's computer... Luckily we're getting considerable attention. We still need: usefulness (things to print) and better quality in general.

There are models for adoption of innovations. I've worked with them and done a lot of simulations constructing variations. Actually, I'm making a RepRap adoption model as we speak and I'm thinking of collecting data to tune it (hence the "Xth registered RepRapper" counter).

A more important aspect of RepRap is probably market saturation. If the RepRap is in a state where it is not particularly useful, its adoption will be limited to "innovators", this is its current market. When it can print pretty stuff (depends on CAD accessibility/intuitiveness and also depends on what pretty stuff 3D warehouses there are by then), then perhaps an "early majority" may wish to try building RepRaps. The adoption will only increase up to a saturation, according to an sigmoid curve (S-shaped) or "diffusion rate". As RepRap matures, it becomes ready for new (broader) markets.

RepRap can go beyond the limitations of physical products in terms of distribution (like software and services), but the 'laws' of marketing will still apply (like with unrestricted software).

So, there's work to do:
- Making sure people who want it CAN HAVE it (replication)
- Unattended printing (print spool)
- Making sure people WANT to use it
- There's an object database to fill (stimulating usefulness)
- Making RepRap's products higher quality (resolution, materials, etc.), faster printing
- Keep cost low (parts + materials)
- Making sure people CAN USE it
- Availability of parts and for example finshed sub-assemblies (e.g. extruders) that are hard to get right. (on topic: World-wide printing webstore would make this possible )
- The product should evolve to be more end user friendly.

The replication potential gives nice numbers but isn't the most fascinating thing to model. The marketing picture introduces more uncertainties and interesting dynamics. There's also the risk of a substitute. What if HP / Xerox / OCE / IBM or any other big (printer or Fab) company decides THEY want the end user fabrication market? I guess we just need to be ahead of them... or do we need patent protection ourselves (under the RRRF)?

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

I'd say who cares?
Provided they don't corner the market, and control the price. If we make it, and everyone gets it for next to nothing, we win.
If HP, IBM, and Xerox all make it, and everyone gets it for not quite next to nothing, we still win.
If they try to make the things for $50, but sell the filament for $500, all we have to do is release aftermarket kits that allow you to void that warranty, strip those parts out, drop them into a new, non-proprietary chassis, and print with $5 filament.
We still win.