RepRap

Ik was zojuist op "De wereld draait door" met onze Ultimaker!

Photo: A christmas tree designed by Paulloricca and printed on the Ultimaker 3D printer in biodegradable PLA.

I took my Ultimaker to BOTACON 0 (the first BOTACON of, I’m sure, many BOTACON’s to come!). There, I was giving a talk on the collaborative ecosystem that is emerging around digital fabrication. Bre Pettis and his Makerbot colleagues have given me ample opportunity to demonstrate the Ultimaker, which many people were seriously impressed by.

Read more here... (sorry, the Ultimaker blog is not in the blog-of-blogs yet!)

Hi everyone,

Some people have been asking for a copy of the PDF of my RepRap research. Perhaps you didn't find a link to it while you were in fact interested in reading it (if not, feel free to ignore this ). If you had a look at it, I'd love to hear your feedback.

On the viability of the open source development model for the design of physical objects: Lessons learned f...
Full-screen (scribd) or PDF.

So much is happening, that I hardly have time to blog it all! In the meantime I've finished my thesis, presented at FSCONS and graduated! And if that wasn't enough, last week I demo'ed for two days promoting the technological-studies at the "Career Day" (or rather, promoting having your own DIY-tech project where you learn even faster!), had a few newspaper interviews while spending virtually all my spare time on the Ultimaker (mostly working on the 5D interpreter support for ReplicatorG). We will have an Ultimaker announcement coming up, real soon now^TM!

FSCONS 2010

Below are the slides of my presentation at FSCONS in Gotherburg, Sweden (description of the talk, here). It was one of those special events where there was a lot of variety in the people attending, but still they all had a special thing for free and open source software. In all, it was really well arranged and for me it was a great opportunity to talk about open source hardware and the role of personal fabricators while at the same time allowing me to practice for my thesis' defense the monday afterwards. Moreover, I had a really nice time with Johan Söderberg and his friends, staying at his place, for too few days, really! (Gotherburg in 20 minutes!)

Graduation!

On the 9th of November I graduated!


These were the slides of the defense presentation, it refers to the notion of public goods which open source communities often create, and how they provide an alternative to for-profit dominant ecosystems.




The thesis was awarded a 9 out of 10, which is very rarely given. Moreover, Prof. O'Callaghan told me it was the best written thesis he had seen in the 13 years as a professor at the University! That's probably the best kind of recognition I could hope for for this work! It wasn't an easy job, at all, but this makes me look back at it with satisfaction. Of course I have all of the RepRap & Makerbot community to thank for, because the survey allowed me to substantiate my claims (statistical significance)! I won't forget this! Adrian told me it was the first serious research project done on the RepRap project. It will definitely not be the last!

The PDF of the thesis can be found here.

Now, on to the next adventure:

From Ultimaker Protobox

Update: Phooky merged this improvement into the official branch, thanks!!

There was an issue with ReplicatorG posted a while ago:

The stl preview pane seems to be fixed at 10cm x 10cm x 10cm. This may not correspond to the build volume of the machine the user has selected, e.g. reprap mendel using makerbot firmware, or self-built or prototype makerbots with non-standard dimensions. It would make sense to obtain these dimensions from entries in the machines.xml (I am guessing that a new axis attribute would be needed, as the definition of the existing length attribute sounds like it may not correspond to the actual buildable length on that axis).

For RepRaps and Ultimakers a 10x10x10 build volume in the preview isn't the right size (Mendel's is about 20x20x14 and the Ultimaker is closer to 21x21x21). Not knowing Java very well it took me half a day, but I fixed it! It was fairly complicated to get this working, because the machine was initialized after the preview panel. Now the preview panel has a refresh feature which is called upon the loading of a machine and is fed the BuildVolumes objects. The BuildVolumes object currently contains a very simple model of the machine's volume (a 3D block), but there's room for adding cutouts and keep-out zones (e.g. nozzle cleaning brushes) that can also be displayed. With multiple printheads, while your axes can still move a long length, the build volume that both can reach becomes smaller. So in the future the axes lengths are overridden by a BuildVolume element in the XML config.



This is the commit and here is a linux and a windows release. Mind: it's not really ReplicatorG-0021 as the filename suggests, and it doesn't contain a skeinforge profile for 5D printing (yet). If someone can send in a good Vanilla Skeinforge profile for the Mendel (5D), I can include that in subsequent releases.

Note: When trying this, make sure to overwrite or update your machines.xml (in ~/.replicatorg/ or Documents and settings\[username]\.replicatorg\)!

Some things are just no fun, and in a project driven by volunteers, these parts tend to be neglected. Making things user friendly doesn't help yourself, because when you get to the point where you can improve it, the improvement is of no use to you any more. If you're going to sell the machine, you're going to do a bit more in making it user friendly. With a healthy bit of altruism and effects of peer recognition the problem is migitated to some extent. Especially if you look at projects like Ubuntu: lots of people are passionate to help make "Linux for Human Beings". But in the RepRap project, the software isn't as user friendly as it should be!

I've been tempted to work mostly on cutting edge work, experimenting with the Bowden cable, reversing the extruder, combatting backlash, working on the then-beta Mini-Mendel, different extruder designs and most recently, an entirely new machine: the Ultimaker. However, more recently I'm focusing in making things a bit easier to use. In effect I've made a ReplicatorG fork (which will be merge with the official version) that includes a RepRap 5D driver that talks nicely with 5D firmware. Because of the integration of the newer Skeinforge (currently 31) (which I had hacked into ReplicatorG-0019, too) you can have Skeinforge generate the E codes too. This means ReplicatorG is now fully usable for the RepRap. I've done some work on detecting temperature feedback and some other things, such as filtering out unwanted M-codes that will cause the firmware to throw up
The 5D fork is temporarily, Adam Mayer will integrate the driver into the main branch. But I wanted to have some testing by others before it's merged.

Release files are here: http://replicatorg.erikdebruijn.nl/.

Another thing I did: For new Windows installations, ReplicatorG and Skeinforge are going to become easier to install. I've figured out how to make it an EXE that is only dependent on the Python DLL which can be distributed along with it. This makes it stand-alone, so you don't have to first install the Python programming language to be able to just do printing. Not everyone should have to do that. Installing the FTDI driver is enough of a head-ache. This is not included in the main release yet, as I need to automate the process of building the distribution some more, and it needs to be built from windows (read: in need to install Cygwin and SSHd).

My ReplicatorG branch also contains my progress bar patch. You'll be able to see how far it is, also how many layers are filled, etc. The ETA of the ETA isn't know yet, but I'll add that at some point too.

I must be honest with you guys: I'm encouraged to do this now, because the Ultimaker is already so much easier to assemble, software has become the hardest part. I deliberately chose to base everything on the RepRap code and not fork things (merging changes upstream) any more than needed so that the whole community benefits from it and to prevent duplication of effort. It also helps me feel more responsible for making RepRap and Ultimaker easier. The Ultimaker electronics (will) use the RepRap firmware, so you'll be able to use it in RepRap's too.

Given that the community is already very big and doubling rapidly, the value of these incremental usability improvements is HUGE. So I hope this encourages you to do the same!!

[This article is in Dutch, but you can Google Translate it (click)]

De Night of the Nerds was een groot succes! Deze werd gehouden in het Technologiemuseum NEMO in Amsterdam.


Foto: (c) Night of the Nerds

Met name ook de opkomst van veel meiden was opvallend en het stereotype 'nerd' was ook niet van toepassing, of in ieder geval had het zeker geen negatieve connotatie! Heel het "NEMO" liep over van de jongeren die iets met technologie hebben, of die gewoon lekker nieuwsgierig zijn!

Samen met David Lemereis gaf ik een presentatie over 3D printing, mijn deel ging over hoe bereikbaar 3D printing nu al is! Vlak voor ons praatje stroomde de zaal helemaal vol. Verder heb ik de hele avond m'n Ultimaker laten draaien en waren er twee van mijn RepRaps te bewonderen (Huxley en Darwin). Al met al een erg gave avond op een super evenement!



Naar aanleiding van het evenement is er ook een leuk interview beschikbaar.

In the RepRap community, you can observe several interesting applications of domestic manufacturing.

A manufacturer will probably not make their product compatible with competing products. But what if you still want this? How cool would it be to mix anything together!


3D printable adapters allow you to combine two different systems, so what? While this looks like a childish example, the development of the availability of downloadable, adapters can have significant ramifications in industries where creating technological lock-in is common practice. Especially the dominant manufacturers in a market do this often. But since you're legally allowed to make parts that mate with proprietary parts [1] this will require companies to focus on the real-value adding features instead of finding ways to reinforce technological dependence that is not rooted in the value of the product but the lack of interoperability. First movers of creating interoperability will create a benefit perceived by their customers that soon becomes the standard. Those who choose to follow because they didn't see it coming will loose customers to the pioneers. Eventually, the whole industry is better of when competition is centered around real innovation and when customers have more choices. And if companies don't offer the adapters themselves, the users of these products can more and more easily self-manufacture them. The question is, do you want it to be a deliberate choice (and as a bonus: a good long term value enhancing strategy) or not.
Being able to quickly combine different products' modules is also a good way to stimulate user invention, with its beneficial societal effects [1]. Not to mention the environmental benefit of enabling reuse of parts from products with lost, broken or missing parts, which enabling self-repair by individuals. Of course sharing designs is important, but it only needs to take a few minutes, and enables you to get feedback or see other people improve your design to a new level! Thousands of objects are already shared in this way (mostly open source) on Thingiverse.com.

It is not uncommon for suppliers not to want to supply replacement parts: it cannibalizes on their sales of the complete products while it is frequently not a profitable business because the logistics are very unfavorable. The reality is: Suppliers usually don't end up making money of of replacement parts. If they do it at all, it usually is to prevent damage to their reputation and brand value. Supplying small quantities of many different parts to many different destinations is expensive. Plus, you usually end up manufacturing many parts that will not be used, because it was hard to determine which ones will break and which ones won't. Because you have to speculate which parts to produce in which quantities, this creates the problem of under- or overstocking parts that need to be stored and collect dust somewhere until the stocks either run out or become obsolete without being used, ever. Apart from this being a huge waste, these are costs that add to the bottom line price of the products.



Moreover, it's hard to supply a whole variety of parts for a product to fit in many different settings. Tony Buser had a part that wore down, but it never was really perfect for him:

"The handle part of my one screen door kind of wore down and the tip chipped off. Consequently, it wouldn't stay shut. Particularly since I put the screens in for spring and when one of my cats pushed against it to look outside - the door would pop open. So I figured if the latch part was longer, it would catch better. So I designed a replacement, printed it, and now it stays shut securely."

You could imagine that in the future manufacturers will supply digital 3D models of replacement parts as a service, so that you'll be sure to be able to fix it at low cost if needed. It enhances the value of the product and will not burden them (nor the environment) with production of spares in advance in the wrong quantities.
To some extend they won't need to, because especially for those parts that break frequently there tends to be someone that shared a replacement that he/she designed for himself. You just need to download an print it. And use it, of course.




If we don't want, more affordable, interoperable, reliable, better-fitting, sustainable and more innovative products, we don't need 3D printers. But I think we do...

References:
[1] See this paper on the legal implications of widespread 3D printing capabilities.
[2] 1. Henkel J, von Hippel E. Welfare Implications of User Innovation. The Journal of Technology Transfer. 2004;30(1-2):73-87. Available at: http://www.springerlink.com/openurl.asp?id=doi:10.1007/s10961-004-4359-6.