Printable Robots | Hacker News Reader

Printable Robots

spectrum.ieee.org

25 pointsjnazario10 comments

The article and project come across as so much being about building robots, as building tools which are capable of building physical things (including robots). The emphasis is on the tools, not the product. One can then ask the question "where does the tool stop and the design begin"?

With proprietary CAD/EDA systems it's reasonably clear, in that the tools are typically in the form of binaries, so the tool is "the stuff for which there is no source".

With open source tools, it's much greyer. Which side of the boundary does a library with source lie? How does the boundary change with time? Say you design a product with free software tools, then others build on top of that product. Is your product now a part of the tool?

This, to my thinking is a competitive advantage of the open source design tools: products can be incorporated back into the tools used to design them, thus extending the tools and allowing the effort expended in the design to be reused. It's also an area that seem to be largely neglected. Library management is generally the weakest part of the open source EDA/CAD tools, at best trying to emulate the commercial library offerings, but ignoring the unique ability of open source to gather extensions in from the associated ecosystem. Software tools at least some of this, in the form of collections like CPAN, github and so on.

I'm meandering here, but is anyone aware of an effort to do for the EDA/CAD world what github does for the software world, but even more than that to offer integration with the tools? Sort of like Diaspora (peer-to-peer) meets github (distributed RCS with social) meets CPAN (integrated) meets OpenCores (EDA/CAD).

I'm currently working on something similar. http://www.cubehero.com

If you have any insights into the problem space, I'd love to talk to you. email me. iamwil@gmail.com

What you are working on could be a good match for BRL-CAD?

It's not exactly the same thing - these are decorative robots, not working robots - but My Robot Nation is a lot of fun. You design and decorate your robot and they print it and mail it to you.

Their online robot CAD program is worth checking out for design inspiration even if you don't order a robot:

http://www.myrobotnation.com/

Yes, but how long until you can print a 3D printer?

depends on the quality you want, and what parts you require to be printed.

the current reprap style uses all printed fittings for threaded rods. you could replace the horizontal rods with printed racks (http://www.thingiverse.com/thing:14537), but your xy precision would suffer. This is homemade, scrounged parts kinds of solutions. If you have access to a nice commercial printer, you can print everything but the electronics and steppers at very high quality. I don't think anyone prints conductive inks as part of the process. I think you can steal this inkjet print head (http://www.thingiverse.com/thing:5152) and incorporate it into an existing system without to much effort.

There's nothing standing in the way of starting with magnets and wires and a 3d printer for motors. you'd probably want some cleverly placed metal sleeves for heat dissipation.

The electronics might be really hard, but they might not. Lots of universities used to let EE students fab their own VLSI chips. I don't think it's outrageously hard, it's just toxic. a lot of this might be solved with cool inks. I know there have been a few printable electronics doctorates handed out over the years.

so.. not long. not long at all.

Do you have a 3D printer? Sounds like you do. Do you also design 3D models?

I do, i have a cupcake.

I have, but they kind of suck. my programming language snobbery makes it hard for me to use openscad. i keep meaning to write a haskell dsl to emit openscad, but i'm unbelievably lazy.

OpenSCAD isn't meant to be a general-purpose programming language. It's a declarative language, like HTML.

That said, there's OpenJSCAD if you like variables. And also, speaking of the devil, there is a programming language CAD in Haskell! It's call ImplicitCAD.

Would you ever have a need for visual source control for your 3D models? If so, I'd be interested in hearing from you. iamwil@gmail.com

that is one of the goals of the reprap project.

https://en.wikipedia.org/wiki/RepRap#Construction

fingers crossed they get there.

Reply on news.ycombinator.com