New user, several new ideas.

I'm new here, but have a few ideas:

- The ability for the next version of RepRap to have, say, three slots off to the side with different write heads to be slotted into. The RepRap could send its current write heads off to an empty slot, and pick up a new one, through the process. This would not only allow the RepRap to set a nozzle off to cool while still extruding with another, it would also allow different nozzles using different materials to be set. Just like a set of colored pens to choose from. This will come in handy as more materials are used, particularly if the software could be told what's in each slot. If I understand correctly, this is something being worked on - but I wanted to mention it in case I misunderstood.

- Speaking of different materials, it would be easy to tell the thing to solder electronics, and give it a special write head that would heat and release soldier.

- I don't fully understand the RepRap yet, but I read that people were looking for sensors to give it some feedback. There are now cheap EM sensors that are smaller than a dime, often used to detect tiny material and structural cracks. They would seem ideal. Here's more information on them, in case it's useful.

[archives.sensorsmag.com]

[electronics.sensorsmag.com]

- Using these sensors on a write head would allow a couple of other neat things to happen. By placing an item on the work surface, the RepRap could move its write head around it and collect sensor data about the item into a file. At that point, any small item could be replicated based on an exact model.

- Additionally, those sensors on a write head would, when combined with a write head that could sinter, allow people to use sheets of plastic for ultracheap, long-term data storage. The RepRap would determine a write format, and sinter holes through, or shapes into, the plastic sheet. This data could be retrieved by scanning the sheets of plastic with a sensor-equipped write head. A similar technique (http://www.theregister.co.uk/2006/11/23/rvd_system/) has been used to use an ordinary sheet of paper as a 450GB storage medium, but paper is easily damaged. And if eventually we can turn bottles and cans into sheets of raw material (see below), we would have a nearly free storage medium we could churn out. Also, the sensors can sense -through- material, allowing us to stack sheets of it and still sense through it. We would then have very compact 3D data cubes made of layers of plastic.

- Machining in plastic will allow people to cast a mold around them, and then use the mold to cast in other materials - like molten metal. This would allow third world countries to cast machine parts, bootstrapping their villages in a hurry.

- Recycling - Imagine not only being able to recycle fabricated parts, but also make RepRapable granules out of different kinds of materials, such as bottles, cans, and old discarded plastic items. For plastics, you'd just cut it up, toss it in a blender, and throw the chunks into a RepRap recycler. All you'd need for a recycler would be RepRap-made components, some electronics, and a basin that could withstand high temperatures. Users would check online to find out what the melting point of their material was, chuck it into the recycler, and basically it would heat the material into extrudable goo, cutting off extruded chunks of it into uniform chunks usable in different RepRap extrusion cartridges. Voila, bottles and cans become RepRap-able material! See WillItBlend.com to see what plastic does in a blender.

- With the ability to RepRap, comes the ability to RepRap things like automated spinners and textile machines! Imagine taking wool and dye, and churning out bolts of cloth! If a later version of RepRap used a laser cinderer as a slottable write head, it could also cut fabric (or plastic, wood or metal!) into different shapes, and basically that's clothing if you RepRap a sewing machine. Download the latest Internet fashions, and make them from wool and dye? Glass-blowing heads that will allow us to make chemistry equipment, and automated machines to convert widely-available materials into usable chemicals for goods and RepRapping materials? Where does it end?

I'd be interested to know what the current obstacle is to making RepRap perform very finely and accurately, which will be necessary for microcircuitry. I suspect those sensors can also be used to determine the X and Y positions along guidelines very precisely as well - see this thread. [forums.reprap.org]

Thanks - hope I was able to contribute something of value.


Be well,

- Satori

Edited 3 time(s). Last edit at 09/14/2008 02:04AM by Satori.

Quote

I'm new here, but have a few ideas:

Hi! I'm going to be a bit mean and shoot some of them down, I'm afraid.

Quote

The ability for the next version of RepRap to have, say, three slots off to the side with different write heads to be slotted into

Multiple toolheads are intended for v2 of the reprap, I believe. Work has been done on them, see here: [blog.reprap.org]

I think the current goal is to make the best use of one head before adding in the serious complexities of additional ones, let alone hot-swappable ones.

Quote

it would be easy to tell the thing to solder electronics

Easy with the handwaving there. There is pretty much nothing about making precision CNC machines from scratch that is easy.

Darwin may well be precise enough to do surface mount assembly work, but it requires at least two new toolheads (solder dispenser and a part holder) to be designed and implemented first, and if not a head hotswapping mechanism, at least a head calibration system so when the paste dispenser is swapped out for the part holder, differences in the XYZ offset of the tip can be taken into account.

This is a goal of reprap, but not one you're likely to see terribly soon

Quote

but I read that people were looking for sensors to give it some feedback

Of the ones you linked, the magnetometer looked interesting but presumably is only of use with conductive materials. The accellerometer probably isn't of use here.

Someone has already designed and assembled a co-ordinate measuring head (I forget who, sorry ) which has the advantage of being very simple, and so anyone can make one cheaply. Firmware to use the head and digitise and object does not currently exist though, I believe.

Quote

Machining in plastic will allow people to cast a mold around them

This is already done. The example paste extruder on the wiki here [reprap.org] uses a rapid-prototyped part to create a mold, for example.

Quote

Recycling - Imagine not only being able to recycle fabricated parts, but also make RepRapable granules out of different kinds of materials, such as bottles, cans, and old discarded plastic items

People imagine this quite frequently, as obtaining feedstock isn't exactly convenient or cheap right now. Unfortunately, extruding a filament from a granules or powder is not a trivial process, due to the difficulty of keeping even flow rates and a bubble-free filament and so on. Experiments have been done of course, but nothing useful has come to light yet. I'm sure it will in time, though.

Quote

the ability to RepRap things like automated spinners and textile machines

Weaving fabric isn't rocket science, I guess. But looms are hefty devices, and I don't fancy printing one out I'm not convinced that lack of raw material for clothing is stopping people from making their own clothes. Automatic cutting and stitching sounds kinda neat though... there are such things as CNC controlled sewing machines about, but I suspect again it would be easiier to retrofit an old machine than build a new one from scratch.

Quote

I'd be interested to know what the current obstacle is to making RepRap perform very finely and accurately

Hard to say. In a few months time when the current crop of rep(st)raps grows up we'll be able to tell! Currently, the problem is not necessarily one of positioning. but precise flow control rate of the extruder. This will be (hopefully) sorted out by people using a rotary encoder on their extruder drives. We shall see.

Anyway, there you go. I appreciate that it isn't easy to find answers to all these questions without having been on this place for a little while. I hope I have been informative!

I'd say for most of the ideas the summary answer is
'Yes, it's more complex than that but we are working on it.'
;-}

I have a few thoughts about creating a shape as a data storage medium.

The register article's claims seem very dodgy. 450Gb on a sheet of A4 would require storing 7KB per square mm! which is about 35Gb/in2. That's about the same as turn of the millennium hard-drives.

I don't think reprap would come close to useable for data storage for a long time. If each Voxel (0.1)mm3 could be independently controlled, which it can't as the shape needs to be connected and we don't have that level of print accuracy, you would only get about 3MB of storage. I also suspect the reading process would need to be destructive to see the inner detail.

As I don't want to discourage you or anyone else from sharing their ideas here is a good adage;

To have a great idea, have a lot of them. - Thomas Edison

It also occurs to me that the choice of material for a long term storage medium is a very important one. If you can't outlast a DVD or magnetic tape, then your storage capacity is academic. There's something to be said for widely supported standards after all

Thanks so much for all the specific information and feedback. This is just the kind of information I needed. I'm glad to know that so many things are in the works - as the open source software people start getting into the open source hardware development community, the developmental curve should spike in ways that traditional technology corps can only have wet dreams about - I'm specifically looking at Linux development as a parallel. I wanted to reply to some of these points, not to argue about them but simply to discuss them.

- Headswaps: I'd been handing out thoughts specifically for Mendel - I should've been more clear on that. How exciting! I checked out the blog link you suggested, Ru - dead sexy! Looks like they're doing hotswaps, using a bank of heads in a row, and that I'd been pretty much been reinventing the wheel - about a year behind, but that's pretty good for having only heard about RepRap for a few hours. I'll catch up, then overtake, and have better suggestions.

- Sensors: I don't know electronics, and I won't pretend to know anything about magnetometers - learning as I go, here - but according to the Wiki [en.wikipedia.org] "Apart from the direct measurement of the magnetic field on Earth or in space, these magnetometers prove to be useful whenever a phenomenon is linked with variations of magnetic field in space or in time, such as anomalies arising from submarines, skiers buried under snow, archaeological remains, and mineral deposits". Because buried skiers aren't generally magnetic, and because the first article said, "They can be used discretely or deployed in arrays for perimeter security, vehicle detection, and unattended sensor systems." and "These new applications could include physiological monitoring [...]" (again, humans not generally being magnetic), my limited understanding of this technology leads me to think that it senses mass vs. non-mass extremely sensitively by picking up the distortions mass creates in the earth's electromagnetic field. This is very exciting stuff when it's picking up a crack 3 mm long and 18 mm below the surface of an aluminum plate - evidently not because it's made of metal, but because it has mass. Good for cheaply sensing what it's doing and what it's done if it's mounted on a tip, and it could also be good for sensing where it is along the X and Y axes by sensing where the mass at its current point along an elongated triangular guiderule. Like a very simple potentiometer using the mass of the triangular guiderule instead of electrical resistance. If not, a simple potentiometer using the resistance of electricity at various points along the X and Y axes would be very simple, and probably very cheap as well.

- Weaving: I suppose what you'd want to do to make a loom is machine-sinder the big parts and RepRap the joiners and fiddly bits. I realize it's a pretty trivial task, but I'm thinking of what RepRap could do for developing countries and self-sustaining hippie communes -er, communities. Live off the grid, churn out cheap, inefficient solar panels in droves from readily-available materials, automate a farm, and sell clothing to the outside world for the few things you actually need to buy. RepRap's prospects for sustainability and economic independence quickly turns into political independence. Retrofitting old sewing machines is an excellent idea, by the way.

- Optical storage: The high end of the estimate (450gigs on an A4) seems pretty suspect to me as well. From the reading I've done on it elsewhere, it encodes data into a) the shapes, b) their colors, c) their sizes, d) their rotation, and e) their orientation relative to each other. The digital equivalent of "a picture's worth a thousand words", basically - visual encoding. RepRap may not have the ability to draw lines that touch, and obviously colors are right out (at least until people develop write heads that can paint). But when Mendel comes out, with the ability to plug in different write heads, and people design new ones, a laser cutter or even small drill write head isn't out of the question. My understanding is that layers of plastic sheets could be read non-destructively with those mass-detecting sensors I was talking about. As far as this medium being inferior to CDs or magnetic tape, I was thinking more of two applications where that wouldn't matter - developing countries that can neither produce nor afford such a medium, and the ability to eventually escape the trap of embedded Digital Rights Management and other such totalitarianism combined with being locked into proprietary corporate technologies. In other words, with with RepRap comes the eventual ability to make our own increasingly-complex computers (cheaply, when we can eventually turn bottles and cans into tons of low-end hardware) and with open source hardware and manufacturing comes a renegotiation of the social contract - let's face it, the system isn't getting any better. When anyone can create a sustainable community somewhere in the wild and set up automated systems to grow crops, manufacture and compute, we're pretty much set. I can't help looking at some of Darwin's earlier results and thinking of where computer graphics were back in 1985 - consider what we have now. With internet-linked open source development, that kind of thing really isn't that much to anticipate, nor very far off.

Anyway, thanks for the specific information! I noticed that RepRap needed a website built. Is this it, or does it still need one? I can do them quickly and easily in Drupal. What about a FAQ with sections about where RepRap is on aspects like these, so that new users can quickly and easily get up to speed and pitch in?

Edited 5 time(s). Last edit at 09/14/2008 04:52PM by Satori.

Quote

my limited understanding of this technology leads me to think that it senses mass vs. non-mass extremely sensitively by picking up the distortions mass creates in the earth's electromagnetic field. This is very exciting stuff when it's picking up a crack 3 mm long and 18 mm below the surface of an aluminum plate - evidently not because it's made of metal, but because it has mass.

According to my own limited understanding of physics, a non-ferrous object is going to have approximately 0 impact on the earth's electromagnetic field. The technology discussed in the sensor magazine sounds like a fundamentally different thing than the subject of the wikipedia article you linked to.

Quote

I noticed that RepRap needed a website built. Is this it, or does it still need one? I can do them quickly and easily in Drupal. What about a FAQ with sections about where RepRap is on aspects like these, so that new users can quickly and easily get up to speed and pitch in?

Well, there's still a standing request for a really well-made objects library, which can be seen at [forums.reprap.org]. I think that the ideas we've discussed for Mendel could be better exposed as well, but that would probably fit well within the existing wiki structure and not necessitate a new layout or site.

Quote

leads me to think that it senses mass vs. non-mass extremely sensitively

Not even slightly. It senses changes in magnetic fields, and so with a suitable piece of control logic you can locate various magnetic or conductive things. Sure, people aren't generally magnetic, but the stuff they carry is much more likely to be so. Skiing equipment has various metal fastenings, for example.

Quote

I'm thinking of what RepRap could do for developing countries

What they generally don't need is assistance with weaving and the like... I suspect that everywhere that gets chilly enough to require clothing susses out making clothes avery, very long time ago.

Quote

Live off the grid, churn out cheap, inefficient solar panels in droves from readily-available materials, automate a farm, and sell clothing to the outside world for the few things you actually need to buy

Solar panels that are efficient enough to be useful are particularly awkward to manufacture. I'm not convinced that they could be made by any kind of cottage industry any time soon.

Also, automating stuff on a large scale is complex... sensors and effectors are cheap but the logic to drive them all in a useful fashion is not so simple. The ability to trivially make spare parts is a more useful near-term goal, and one that is most definitely with reprap's reach.

Quote

developing countries that can neither produce nor afford such a medium, and the ability to eventually escape the trap of embedded Digital Rights Management and other such totalitarianism combined with being locked into proprietary corporate technologies

Neither CDs nor magnetic tapes imply DRM. They are also sufficiently old that they aren't really proprietary any more.

And the last thing that anyone needs is a storage system incompatible with the ones everyone else uses. Developing countries especially will need and want to be able to use exactly the same systems that more developed nations use, DRM and lock-in be damned. Doing otherwise does them no favours.

Ru Wrote:
-------------------------------------------------------
> What they generally don't need is assistance with
> weaving and the like... I suspect that everywhere
> that gets chilly enough to require clothing susses
> out making clothes avery, very long time ago.

Sure, but Europe knew how to make clothing for a good long while before the Industrial Revolution and it still turned out to be a big hit. Just because the capability exists doesn't mean that it can't be made easier/faster/cheaper. Using RepRap to facilitate everyday tasks (weaving in this case) is probably a much more realistic goal to implement than even trying to roll your own filament from corn, for example.

I think that Satori has a legitimate point here. It certainly merits further investigation.

... some companies as Airbus and developer of military vehicles are experimenting with carbon-fiber-fabrics - they sew CNC-cutted parts together, inflate to the final 3D-form and imbue with epoxy to get the rigidity ...

So weaving, stitching, crocheting and knitting evolves to high-tech-methods!

So why not experimenting with similar techniques for rigid, usefull and extreme lightweight frames and mechanics?

Viktor

Quote

I think that Satori has a legitimate point here. It certainly merits further investigation.

I'm not convinced. The production of fabric very readily lends itself to mass production, and indeed that is what is done. It is the conversion of such raw materials into parts and finished items which is complex, expensive, time consuming, etc etc.

This is where the benefits of automation show, and having the tools in your hands is a tangible economic benefit. Sure, it might take a long time to get right, but in the end you've accomplished something that cannot be done so easily or cheaply without significant amounts of skill or expensive machinery. Being able to hook up a roll of fabric to your clothing-o-matic and have it spit out a tshirt is an example of such a thing.

The same goes for the production of fabric. You've expended lots of time and effort and expense, only this ime you've produced something that you could already obtain elsewhere cheaply and quickly. The benefits are significantly smaller, and therefore the effort that should be reduced.

Quote

So why not experimenting with similar techniques for rigid, usefull and extreme lightweight frames and mechanics?

It would seem to be vastly easier to obtain and use carbon fabric than it is to get raw filament and weave it yourself. It also avoids the stage of needing to build a carbon loom, which I imagine is quite tricky; significantly more so than making one for more conventional fibres.