Spot: FESTO's iFab

I'm interested in your thoughts on this.

A friend pointed me to the following URL:
[www.festo.com]

I can't help but think that there is quite some inspiration from various RepRap parts. Especially the screw/filament holder. Interestingly a tripod platform was chosen, which was also evaluated for the RepRap.
[forums.reprap.org]
By Viktor: [forums.reprap.org]

I must say that it is used very elegantly. The non-orthogonal axes allows you to make use identical subassemblies, but completely identical. And it would allow, depending on the implementation to have more degrees of freedom of the toolhead. For instance you could extrude on the side of an object that you're making...

Obviously there are many differences, those are of course the most interesting.

One of the designers is dr. Evan Malone, who is also core developer for the Fab@Home. I wonder where FESTO stands on intellectual property. I guess that a dialogue with communities like ours and the Fab@Home community are very valuable for future innovations in this product line.

It also tells me that a RepRap is very easily productized. Of course Bits From Bytes have made the Darwin design more accessible to a wider audience. I think their variant is excellent. By the way, the BfB filament holder looks similar too

---

Regards,

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

Hi Eric,

... go here for a complete documentation of my tripod-assembly: [builders.reprap.org]

And yes, it's very interesting, how ideas spread out

Ciao, Viktor

ErikDeBruijn Wrote:
-------------------------------------------------------
>
> I can't help but think that there is quite some
> inspiration from various RepRap parts.
>
Actually, if you read their PDF you'll see that they credit Evan Malone and Hod Lipson of fab@home credit for the concept.

It appears that Hod Lipson and Festo have been dealing with each other for some time.

His lab also came up with this thing...

[www.youtube.com]

that Festo has taken to their hearts.

Edited 1 time(s). Last edit at 04/20/2009 01:14PM by Forrest Higgs.

---

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

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

Intriguing, I love their tools because of the innovative solutions put in there. I wonder which direction this goes. The molecules look interesting too.

... it's the way into the 'technological singularity' - a situation where anything will be possible and it would be impossible to predict any straight line in social and/or technological behaviour ...

Look into the descriptions of utilityfog ( [www.wildirisdesign.com] ) or foglets ( [www.kurzweilai.net] ) - and then think some steps further

Viktor

Is it just me, or do the 4 springs on the extruder mechanism look eerily similar to the Mk. 2 Extruder design?

Viktor; I love The Singularity is Near. It's one of my favorite books of all time. I credit it with allowing me to me able to see the greatness in reprap a year ago. Now if only I could get my friends to read it.

... look at the development-timescale from the first transistor to the PC and the present situation - this lasted 60 years or so ...

Now compare with Feynmans first ideas, Drexlers writings and the present situation - i think this could be a similar progress on a timescale of 70 to 100 years, so maybe around 2040 this could be a real challenge

On my private line i mastered the submicron area with CNC-milling 15 years ago, with 3D-fabbing i have reached accuracies of 10 microns, maybe more precise next two years, so this could last twenty years more until i'll reach molecular precision at home.

With some specific development and possible crossover with day-job or funding it could be faster, but it's a really big chunk to master this on my own

Viktor

Forrest Higgs Wrote:

> ErikDeBruijn Wrote:
> > I can't help but think that there is quite some
> > inspiration from various RepRap parts.
> >
> Actually, if you read their PDF you'll see that
> they credit Evan Malone and Hod Lipson of fab@home
> credit for the concept.

I did, that's why I wrote this:
One of the designers is dr. Evan Malone, who is also core developer for the Fab@Home.

Anyway, the molecubes are certainly cool (also posted in Things to print: [forums.reprap.org] ), not the least because they're open source and we could be printing the STL files right now: [www.molecubes.org]
Files: [128.253.249.235]

Too bad that you need 200 dollars in parts besides the RP parts. I can't help but think that it should be possible to build a cheaper version of them... The slip ring might be fab-able (100 dollars down in price).

Also, I'd expect that a molecube's system boundary would end at the pivoting point but *would* include other block that it's properly connected to. You would treat each molecube as having two independent parts that are part of bigger systems that in turn make up the entire reconfigurable bot. Seeing each cube as two systems would reduce the need for interconnection for anything but power and Dallas single wire communication (3 pins in total?). Better connections can be made between blocks than within them (via the prohibitively expensive slip ring). A 3 wire slip ring would be fab-able, even with a coarse solder extruded board and a few bended wires that slide along it (or would that cause too much interference?)

---

Regards,

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

I quite enjoy looking at Lipson's stuff. It's always flashy and cool. When I look harder, though, invariably there is little if anything beyond flash and coolness.

Think about molecubes, for example. It's really fun to watch. Try thinking of what sorts of situations such an "innovative" technology would be useful in, though, even if it could be made cheaply, which, be honest, it can't. Yeah, a gripper on the end of a series of molecube units can address quite a bit, but by no means all of a decent volume. what it can't do, however, is guarantee that the gripper can approach a target from a particular direction. That makes the technology flashy and cool but from a practical standpoint, largely useless.

Fab@home was much the same way. Notice that the positioning 'bot is really cool, made, as it is, of laser-cut acrylic. It's also mindbogglingly expensive, like pretty much anything Lipson's outfit does. Where it really sucks, however, is in the much less than brilliant extruder concept.

Adrian Bowyer with reprap took exactly the opposite tack. He and the team knew damned well that a successful positioning 'bot could be built in a large number of ways and that when one built one cost and accuracy were going to be the driving criteria. Adrian went to the heart of the matter, however, when he concentrated on building a practical extruder first and foremost. The old Mk II was ready to go at the end of 2005. THEN Adrian and the team went after the question of the positioning 'bot.

Darwin, in kit form, costs maybe 40% max of what a Fab@home costs and the percentage is being aggressively reduced month by month. Fab@home still costs and looks like what it did when Malone got his degree with it.

Figure out which approach is better, flash and cool or practical and steady.

---

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

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

I totally agree with your point: 'better practical and steady vs. cool and flashy'. On the other side, I am sure that there is a difference in what they communicate and what they plan to do with this research. The communication with their examples and youtube movies looks flashy, cool and whatever because thats what 'communicates' best to the broader public (sex would be even better). The research will most probably not be used to commercialize shiny blimps, swimming penguins (they run linux?) or rubicslike robotarms...
I think many OS projects could benefit from some extra sexiness in their look, feel and communication, once they try to gain brother acceptance. Offcource there's a balance to be found and I think for example Makerbot is doing a good job with the cupcake.

Cupcake is a good design, if it could actually make candy it would be wonderful

The Darwin design is in no way comparable with the iFab & fab@home. The darwin is based on evolution, while the fab@home is based on rapid prototyping. Thats why darwins are getting cheaper and cheaper, while fab@home still is so expensive.

Seeker, there will be a syringe on my BfB machine. Ill make you some candy corner blocks

All you have to do is do an alexa rating on fabathome.org and reprap.org and you well see very quickly what sexiness gets you. ><

---

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

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

I think the reprap's slightly ugly/scary look and feel(as described by a former partner of mine who was definately not a pro-tech geek like most of us) but capacity to improve is better to have over something that looks good, but is something that might become a dead end in development terms. The reprap is pretty much based on the idea of biology, both in terms of it's reproduction ability and it's evolution. Reproduction is great for the cost effects because of the inherantly different model(hense why Dr bowyer has been sometimes reffering the reprap as an agriculture project as well as an engineering one). Plus the danger of your customers replicating everwhere makes sure that all the suppliers keep honest with pricing to be compeditive.

I made a choice a few years back of whether to look at the fab@home or the reprap as the printer I would work on. The fab@home at the time could do more and was more complete, but the reprap seemed like it could progress more in the long term. I think it's somewhat telling that while the reprap has improved itself alot since then (the curcuit printing is just one great thing it couldn't do a while ago, while print quality has gone way up) the fab@home hasn't really pushed itself forward that much. It does it's job well, but I don't think that it's really the way to go unless you want a printer right now and aren't interested in getting better builds in the future.

As forrest described in a blog post a few months ago, you can either carry out centralised selection of the next species, or mutate and carry out the equivilant of horizontal gene transfer to produce the best results along with the central method. Fab@home seems to be really centralised in it's development, while reprap is a hybrid. Basically, we have both the main blog and the builders blog and it's the mix of the two that lets us go forward so fast.

Hey Forrest. Please understand me well, you DEFINITELY need both. As a design teacher (interaction design, physical computing, 3d) I always stretch my students that you cant make a good design if the product/project sucks but on the other hand a good design in not something you can slap on like a layer of varnish in the end. They should go hand in hand. Thats always a hard discussion between engineers and designers but I think the good ones are those that understand each other and cooperate.
I also was examining RepRap and Fab@home for my project. I liked Fab@home better at first because it looked simpeler (a closed box vs. an open structure of 22 rods) and because I need a syringe for my project. But I chose RepRap because of the momentum around it, the price and later the capabilities. Many people are attracted to RR and there's brewing a lot on many, many fronts. The F@H site didn't change at all in more then a year while RR change a lot. Momentum is soooo important for any OS project, if it doesn't attract involved people it dies. And with involved people I mean developers, engineers, hobbyists and geeks but also web designers, propaganda writers, graphic wizards, animation makers, interaction designers, end users and many others who can contribute on the human side of Human-Computer-Interaction. What if we have the perfect self replicating machine and it has replicated itself until every hardware-lover has one?

ps. This is definitely NOT a rant on how things are developing here, I think that things go fine and the flexibility of the RR design leads to many spinoffs that try to lower the barrier while not obstructing the main goal of the project. Yes Darwin AND Intelligent Design, who would have thought that!

ps.2 My girlfriend liked the reprap design more then the F@H

I agree on momentum. That I always figured was really what made OS projects work so well. Because there is always the threat that someone else will do a better job of it than you if the momentum lets up. If you start to slack off then someone will start working on it more than you and become the main person running the project. I find it very unlikely of course, but if Dr Bowyer and the main developers were to all give up tomorrow, the project has enough steam in it to keep kicking and going forward, alibeit at a slower pace (there is already experimental non-core developer versions of half the electronics and software floating around the forums).

All that said, if I had the time (and if I hadn't made my X-axis stepper driver explode last night) I'd really love to work on the molecubes, especially on making the things easier and cheaper to make (they are currently pretty expensive for a hobbyist to make. Give you need at least 3 or 4 to do anything with them) with the reprap, especially now they've been open sourced. I have a copy of the Freitas book [www.amazon.com] which details how machines like these can be used to form replicating systems.

... some years ago i invented a sort of oszillation-driven linear/radial motor concept what's capable of replacing a motor-gear-combination with two plane sheets and a PZT-oszillator - look here for some drawings: [depatisnet.dpma.de] - from Patent-# DE00001994820A1 (or go to [depatisnet.dpma.de] and search for author "Viktor Dirks")

I made some samples and demonstartors with a coarse 'directional' nylon-sticky-tape - look here: [www.ctbot.de] the OSZ-images (the gripper-designs are covered with the same sticky-tape)

With finer structures (maybe aligned carbon nanotubes instaed of the nylon-fibres) this would result in ultrasmall (until some microns effective sizes) driving sheets for all sorts of moving, turning or extracting telescope-like rods or such.

The applied forces are extreme high compared with normal friction-driving - it's the stiffnes of the fibres and the count of hooked fibres per active area what works. With my demonstartors i had some Newtons per square-centimeter maximum force until the fibres bend free - with carbon nanotubes we estimated a possible applied force of some ten kilonewtons per square-centimeter!

With the nylon-sticky-tape the drive-scales are in square-millimetres, with smaller fiber-structures this should go smaller too - so imagine molecubes in some millimetres diameter

Maybe someone is interested in development of this sort of stuff?