FabLab inventory and Beyond

[Marcin Jakubowski]

Friends,

I would like to call out further for participation in RepLab. The focus is economic significance, ie, applicability of open source tools to real production functions of society. Sam discussed this point in his post, On Replication.

I am posting this same message to anno...@lists.hackerspaces.org and dis...@lists.hackerspaces.org . If someone knows more effective means of contacting hacklabs, pass this on. We also need to pass this on to the FabLab people.
----
As such, the discussion on RepLab needs to start with components. We propose these here, and then ask collaborators explicitly what they can contribute to the RepLab tool inventory development process. We are looking primarily for people who have can research, design, and build equipment. The goal is producing tools that can lead to disruptive change by combination of open source business model and Type 3 replication (see above article). We should be clear that the development has openness as a priority - including the enterprise model for building respective machines. The enterprise model must include economic analysis and ergonomic analysis for the RepLab tool, to promote economically significant production of that tool or of its product. This is in the name of distributive economics - and to spawn a large number of production facilities worlwide that rely on open source tooling. This is a means to address bootstrap funding towards post-scarcity economics for many of the groups involved - by earnings from viable, open business models.

It should be said that the common ground between RepLab development and toolchain application towards economically significant production (such as automated circuit fabrication proposed by Sam) is the development of the toolchain components. This means that if we want to develop toolchains - we need tool chain components - and that's where the core mission of RepLab lies. We are hoping that a large pool of developers from many hackerspace-related initiatives come together on developing the various Open Source Fab Lab tools, since all of us are interested in at least some tools of production.

Sam proposes automated circuit fab as one viable enterprise. Erik brings RepRap developments to the table, and I hope that this becomes a viable open business model, possibly co-developed with Makerbot. FeF brings RepTab to the table - we have an untested prototype so far, and we welcome on-site collaborators.

The list of tools needed is below, with only the key items listed. Please get back to us specifically on what you can contribute. In particular:

(1), tell us the tool of interest to you;
(2), what specific technical developments you are willling to contribute to the project - such as design, calculations, research, fabrication, testing, documentation, web development, marketing, resource development, parts sourcing assistance, etc.
(3), resources and infrastructure that you have available, and what gruops you are involved with or that you can leverage for assistance
(4), suggestions on strategies and enterprise models that can be utilized in development
(5), suggestions on tools missing, and how you're willing to contribute to devloping them
(6), how much energy you can commit - do you 'have a job' or can you commit significant time?

TOOL LIST - from FabLab (see  detailedhis for a list)
1. Laser cutter - large DIY community exists for C02 lasers
2. ShopBot - RepTab is the Factor e Farm version
3. Precise router for milling circuits
4. Plasma cutter - power circuit is main point to opensource
5. Welder - power circuit is main point to opensource
6. Oscilloscope - can a computer oscilloscope cover most needs?

Beyond FabLab:
1. RepRap - fabrication streamlining and open business model needed
2. Mill
3. Drill
4. Lathe
5. Induction furnace - power electronics are main point to opensource
6. Ciruit fab - automated process including pick-and-place
7. Aluminum extrusion
8. Metal casting - of ingot from induction furnace, and other molds
9. Hot rolling
10. Cold rolling
11. Forging
12. Metal shear and hole punch for up to 1" steel
13. Wire drawing

These tools cover electronics, precision fabrication, heavy metal work, and ability to make other tools for producing any technology known in advanced civilization. Tools created from the above can even yield clean room technology. The above is quiate a limited set, but is sufficient to generate other tools. The latter parts focus on serious industrial process, with a bias towards building replicable, post-scarcity resilient communities with capacity to bootstrap largerly from scrap steel.

Comments and refinements of problem statement are welcome.

Thanks,

Marcin
mar...@replab.org



--
--------------------------------------------------
Marcin Jakubowski, Ph.D.
Open Source Ecology
http://openfarmtech.org
opensourceecology at gmail dot com
Skype: marcin_ose
--------------------------------------------------

Nobody said that building the world's first open source village would be easy.

-- Anonymous, 2009

A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects.

-- Robert A. Heinlein

NOTICE: All discussion in this communication is in the public domain, unless otherwise noted. If you are sharing proprietary, confidential, or otherwise privileged information, you must make that explicit. Otherwise, this discussion may be copied, republished, and otherwise used in the public domain - respectfully and with proper attribution. Furthermore, please consider that we are not interested in discussion as much as action. Therefore, we are particularly interested in discussion of ideas that both parties can commit to by acting on them.

[Smári McCarthy]

Hi guys,

Recent developments over here have been to the effect of a
Hackerspace potentially opening soonish in Reykjav�k. I'll keep you
posted as we move forward.

One thing I intend on doing as soon as possible is developing a 3
axis articulated arm that can do FDM - similar to RepRap. I've been
thinking about the design for quite a while and I believe it would be
fairly easy to build in such a way that it can self-replicate most of
its parts. Attached is a very rudimentary sketch of the idea; I've not
figured out most of the mechanical details yet I'm afraid.


- Sm�ri

Marcin Jakubowski wrote:
> Friends,
>
> I would like to call out further for participation in RepLab

> <http://openfarmtech.org/weblog/?p=1254>. The focus is economic

> significance, ie, applicability of open source tools to real production
> functions of society. Sam discussed this point in his post, On
> Replication

> <http://groups.google.com/group/replab/browse_thread/thread/bba7bd0881b10fc1#>.

>
> I am posting this same message to anno...@lists.hackerspaces.org

> <mailto:anno...@lists.hackerspaces.org> and
> dis...@lists.hackerspaces.org <mailto:dis...@lists.hackerspaces.org> .

> RepTab <http://openfarmtech.org/index.php?title=RepTab> to the table -

> we have an untested prototype so far, and we welcome on-site collaborators.
>
> The list of tools needed is below, with only the key items listed.
> Please get back to us specifically on what you can contribute. In
> particular:
>
> (1), tell us the tool of interest to you;
> (2), what specific technical developments you are willling to contribute
> to the project - such as design, calculations, research, fabrication,
> testing, documentation, web development, marketing, resource
> development, parts sourcing assistance, etc.
> (3), resources and infrastructure that you have available, and what
> gruops you are involved with or that you can leverage for assistance
> (4), suggestions on strategies and enterprise models that can be
> utilized in development
> (5), suggestions on tools missing, and how you're willing to contribute
> to devloping them
> (6), how much energy you can commit - do you 'have a job' or can you
> commit significant time?
>
> TOOL LIST - from FabLab (see detailedhis for a list

> <http://openfarmtech.org/index.php?title=MIT_Fab_Lab_item_list>)

> mar...@replab.org <mailto:mar...@replab.org>

[Marcin Jakubowski]

2009/11/25 Smári McCarthy <sm...@fabfolk.com>

Hi guys,

 Recent developments over here have been to the effect of a Hackerspace potentially opening soonish in Reykjavík. I'll keep you posted as we move forward.

 One thing I intend on doing as soon as possible is developing a 3 axis articulated arm that can do FDM - similar to RepRap. I've been thinking about the design for quite a while and I believe it would be fairly easy to build in such a way that it can self-replicate most of its parts. Attached is a very rudimentary sketch of the idea; I've not figured out most of the mechanical details yet I'm afraid.

Do you think implementation of cartesian motion in a noncartesian mechanical system will be a difficult problem?

Also, what is the main list to spread the RepLab message to the FabLab people?

Marcin

 - Smári

http://openfarmtech.org/weblog
http://www.replab.org
marcin at replap dot org

[Sam Putman]

2009/11/25 Smári McCarthy <sm...@fabfolk.com>:

> Hi guys,
>
>  Recent developments over here have been to the effect of a Hackerspace

> potentially opening soonish in Reykjavík. I'll keep you posted as we move

> forward.
>
>  One thing I intend on doing as soon as possible is developing a 3 axis
> articulated arm that can do FDM - similar to RepRap. I've been thinking
> about the design for quite a while and I believe it would be fairly easy to
> build in such a way that it can self-replicate most of its parts. Attached
> is a very rudimentary sketch of the idea; I've not figured out most of the
> mechanical details yet I'm afraid.
>
>

>  - Smári
>

This kind of robot arm is called a SCARA and there are a lot of people
building them, including Aaron Nielson of oomlout in the UK.

If you add a parallelogram linkage to the 'forearm', then your
extrusion head will stay perpendicular to the work surface.

This sort of robot requires inverse kinematics, but I'm fairly certain
this poses no challenges that LinuxCNC/EMC2 can't meet. You do lose
precision as you approach the perimeter, and get it back as you
approach the base: simple consequence of polar coordination.

You may do better to build a SCARA that has linkages parallel to the
base rather than perpendicular to it, and drive the Z axis with the
usual leadscrew. That would mean your extrusion head would stay
perpendicular without active controls or a potentially fussy
parallelogram.

Have fun either way :-) It looks like an entertaining project.

Cheers,
-Sam.

Cheers,
-Sam Putman.

[Smári McCarthy]

>
> Do you think implementation of cartesian motion in a noncartesian
> mechanical system will be a difficult problem?
>
> Also, what is the main list to spread the RepLab message to the FabLab
> people?
>
> Marcin
>

As Sam pointed out, accuracy will drop as you reach the perimeter, but
if the overall accuracy is good enough this won't matter much.

There is no Fab Lab mailing list. This has annoyed me for YEARS. I have
on two occasions tried to start such mailing lists to no avail. For some
reason Fab Labs seem to attract people who are allergic to e-mail (and
sensible virtual collaboration tools in general).

- Sm�ri

[Erik de Bruijn]

Hi Smari,

Having a toolhead perpendicular to the layer orientation in layered fabrication is more a result of most mechanical setups than it is a design rule. Especially for FDM or FFF (which we RepRappers call it), it would be very useful to have more degrees of freedom. This would allow you to fabricate much more objects without support material (waste). Whille a recycler should also mitigate that problem, it still adds up to the build time (supports take time as well, for some objects this might even be more than the object itself).

Since extruded thermoplasts are very viscous, gravity is not affecting them as significantly as the direction it's being extruded from. An extruder will produce an outward force from its orifice. This means that you could print to the side of a plastic objects and have 90 degree overhangs easily, instead of the 45 degrees that we now use as a rule of thumb.

Smari, what would make your mechanical setup slightly easier is to have all mechanical bots work in the same plane (same height). You would need only one actuator to lower the build platform, instead of several extra degrees of freedom in the several heads that you intend to make. There is a point where scaling up the machine vertically (making it faster) become harder in proportion to having more machines (horizontal scalability). If all robot arms are in the same plane, you could do with 3 actuators per arm. The arms need to be as light as possible of you want to have rapid controlled motion and still be energy efficient. This can be achieved through an extrusion setup that I'm currently using. The thermoplastic filament is pushed through a tube or guide, which  is possibly printed itself (I recently confirmed that that will work!). The extruder will only be the heater and extrusion head, while the drive mechanism is on a static part of the machine. I've achieved a big speed advantage with this method already. I haven't even tested its limits yet because I don't have the time and my setup time is more valuable for me than actual production throughput.

Erik

-- 
Groeten,

Erik de Bruijn

--
My latest blog post(s): 
RepRap @ NPOX 2009, Hilversum
  
    We demonstrated the RepRap at NPOX (15 nov. 2009).

It was in the "Beeld en Geluid" building which was a huge building, but beautiful at every scale.

We were positively surprised by the way our project was advertised on pamphlets that were passed around and people seemed to know about it in advance. The visitors were eager to learn what was going on and posed questions ranging from how the t...
Read more: http://blog.erikdebruijn.nl/archives/121-RepRap-NPOX-2009,-Hilversum.html

[P.M.Lawrence]

On Nov 26, 1:43 am, Marcin Jakubowski <joseph.dolit...@gmail.com>
wrote:

> 2009/11/25 Smári McCarthy <sm...@fabfolk.com>
>
> > Hi guys,
>
> >  Recent developments over here have been to the effect of a Hackerspace
> > potentially opening soonish in Reykjavík. I'll keep you posted as we move
> > forward.
>
> >  One thing I intend on doing as soon as possible is developing a 3 axis
> > articulated arm that can do FDM - similar to RepRap. I've been thinking
> > about the design for quite a while and I believe it would be fairly easy to
> > build in such a way that it can self-replicate most of its parts. Attached
> > is a very rudimentary sketch of the idea; I've not figured out most of the
> > mechanical details yet I'm afraid.
>
> Do you think implementation of cartesian motion in a noncartesian mechanical
> system will be a difficult problem?

It is and it isn't. The direct version of the problem is, "given
cartesian co-ordinates, what manipulator parameters - angles,
displacements, whatever - will position there?", and you could add in
orientations etc. to the cartesian co-ordinates as well. In general
that is difficult to solve, because in general the equations going
from manipulator parameters to cartesian co-ordinates, although
straightforward, involve trigonometric functions. That means you have
to invert those equations, which is generally hard even with
computers.

But that's not the best and most appropriate way to set up the
problem. A better way is, "starting from here, what path/trajectory
takes us there?" - with here implied and only there specified in
cartesian co-ordinates. But you already have both cartesian co-
ordinates and manipulator parameters for here, so (with one kind of
exception) it is practical to use a generalisation of Newton's
Algorithm to work out the path/trajectory; you just step along it in
small increments, using the manipulator parameters from each step as
the seeds for iterating to get those for the next step. With small
enough steps, the continuity that comes from the physical problem
means that each seed is within the zone of convergence needed. It may
even be practical to optimise the path/trajectory - subject to its end
points being determined - for some other measure, e.g. time taken or
maximum torque, say with dynamic programming.

But there is one outstanding sort of problem: in general, this sort of
physical system can turn up analogues of gymbal/gimbal lock, so it is
important to detect and avoid the approach of such things. This can
also be considered an optimising issue and folded into whatever else
is done along those lines, but it is a difficulty that doesn't simply
go away. P.M.Lawrence.