Future version ideas

I've never felt the need to get involved in the diy fabricator community before, but I've been thinking lately about the future of what fabricators will look like.

I have an idea for a design that I think will be more flexible and more powerful than the current servos-in-a-frame version. The idea is to use an arm or "tentacle" that can operate like the human arm, extending itself out and around. It should be nimble enough to move around significantly from one mount point, be capable of repositioning itself around the object it's building, have different tool heads so that when they work in groups they can do tasks independent of one another or cooperatively, and should be made of as many similar parts as possible so that a single arm can be extended to arbitrary length or can make new arms simply by copying the same part over and over.

Meh?

Such might be useful for creating overhangs. Otherwise, it'd probably be more effort for no more gain.
Multiple toolheads, working in tandem, though, sounds good.

Actually, it sounds like you're talking about the reprap altering itself on the fly, while making another object, so as to make that other object. This isn't a new idea, (see "Diamond Age" by Neil Stephenson, specifically the comment on the mooring tower for the airship), but it is a good one. Probably a ways off yet. By the time the machine can alter itself to make toolheads, it can also make most of the things we're interested in with the basic toolheads. If it can create the windings of the motors, (or other actuators,) the heating element, or sintering tool, the structural elements, it can already do plastic appliances, with embedded electronics.

It might be a useful addition if the tool can fabricate its own chip foundry on the fly, or create, for itself, the ability to work with ceramics or metals. I suspect the percentage of incidences where the reprap temporarily expands its repertoire to create something, will be rather low. Permanently expanding its range, however, should be common. Once it gives itself the ability to work in ceramic paste, it shouldn't have to go back to not having that ability, (unless the owner finds the need to scale the thing back for portability, or loans out that toolhead to a friend.)

Augur, you mean like a welding robot?

Image: [www.kobelco.co.jp]
Another image: [www.iai.csic.es]

Edited 1 time(s). Last edit at 11/24/2007 12:32PM by RoundSparrow.

@ Sean Roach: What if you want to fabricate something larger than your fabricator? I think it'd be easier to manage that with a print head and articulators that _aren't_ fixed to a frame in a permanent way.

@ RoundSparrow: To some extent, yeah. Tho I was thinking with significantly more versatility.

You make a good point. So, a future reprap may be built on an automatic tractor...
Not a bad way, if you can get the positional accuracy under control, and can consistently get on top of what you've already made to lay more material.

Personally, I figure the future reprap will probably be a gantry-crane mechanism, with multiple heads, and the ability to interface with additional gantries for cooperative construction. You make the cranes just smaller than the room that holds them...or you make a new gantry, folded up or disassembled, that you move outside.

An arm is a workable solution. I can see three problems with it, with differing levels of difficulty.

1. Different accuracy at full extension than at partial extension. I don't see the solution to this, but others are doing something similar which also faces this issue.

2. Leverage. For deposition modeling, less of a problem, but for machining, potentially a more serious one. The cage, at least a version of it, can be braced for this better. As can the gantry.

3. Position calculation. A solved problem, but it involves more math than does the cartesian coordinate system, where each axis position is effected by one, and only one, variable. Of course, once programmed, infinitely duplicatable.

It provides one advantage, that I see, that you pointed out.

1. Small storage/transport footprint.

As far as making an object larger than itself. It is still limited by its reach. It can only build as high, or as wide, as its arm can extend.

A valid path, none the less. You might want to read up on some of the things Victor Dirks is working on. He's facing some of the same issues your arm would.

The best path would probably be your arm, on some mechanism to climb with. Then it could be involved in creating any size object, simply by climbing the object as it built it, putting the finishing touches on one "face" as it climbed back down.

Actually, I'd like to have a swarm of tonka-truck sized crawler repraps.

Just my opinion on the matter, of course.

Hi augur,

... here is an article about a 'parallel scara design', whis is well suitable for high accuracy (look at figure 5) [www.harmonicdrive.de]

With special kinematics you can made astonishing things, but it's a very wide area ...

Viktor

On the making things larger than itself issue...

Imagine an arm with an extruder head on it. This arm could travel along a rail--single solid piece dictating the build area of the machine. The machine could then build make a rail for itself that is as long as the original rail plus two arm lengths--actually slightly less but almost that long. Thus the machine can make larger and larger versions of itself even without segmentation.

Or you could do a circular rail and print the next one outside of the first and increase the size that way. I'm sure there are many ways to do this.

Demented

Nice link viktor!

Demented

I like the idea of a combined arm and gantry. Perhaps a sufficiently long arm could use it's length to act as a gantry by locking into position, and then unlocking when it needs to act in a more versatile manner.

augur,

Locking arm/gantry...nice idea to increase rigidity at longer extensions. Could be useful for house-building RepRaps! Or any other large structure that you don't want to have to build a lot of support structures for as it would only come down anyway.

Demented

Roach_S Wrote:
-------------------------------------------------------
>
> Actually, I'd like to have a swarm of tonka-truck
> sized crawler repraps.
>
> Just my opinion on the matter, of course.


That's another way around the problem of fabrication large objects. Put the unit on a mobile system so it can move as it works. Introduces new levels of placment math though...

@ DentroMan765: I forgot to add above: with the arm-gantry design, you could use the device's swappable toolheads as graspers so it could move around the object, or to manipulate them or whatever.

Also you could use them to assembly whatever parts are printed, so the whole thing becomes an automated manufacturing plant.

I wonder if also it'd be possible to do some sort of non-printed autoconstruction when that method would be faster and easier. Some sort of generic milling (which would work well as a toolhead on the arm-gantry), or molding (which could use the arm-gantry to shape wax or some other stuff that could be used to form a mold around and then use the arm-gantry to manipulate the molds and pour the cast substance).

I was thinking last night, it might make sense to have design more of a spec for an arm-gantry type fabricator, specifying things like joint connection design, communication and power, software interface, etc. so that you don't have to have a finite set of part designs, but instead you could design your own parts and design the software interface and bam, done.