A_FAB Robot in action

First documented operation of the A_FAB robot. It sits idle, stands, walks, changes the position of its end effector, and extrudes a material in specific pat...

10 members like this

Get Embed Code

Tags:

Comment

You need to be a member of Grasshopper to add comments!

Comment by ben Fruehling on April 11, 2012 at 8:38pm: If you now have a working 3d printer, you can probably have now found out that initial Z height of the nozzle is critical to the print quality of the finished artifact. If you could incorporate a sensor on the nozzle to find z=0 I think you will be getting much more exciting results.

I also can't tell if you are printing paste or plastic. The specific design of your nozzle/ extrusion is in someways similar to a "bowden's extruder" which is used in reprap 3d printers. This is interesting because the long filament drive allows massive retracts of the material which causes negative pressure in the nozzle. this reduces the 'ooze' that comes during your chuckle at the end of the video, and will help quality as well.

You have me very curious; what material is it printing?

I would like very much to know more, and to add more if you find my suggestions helpful.

Comment by Dicky Ferd on April 18, 2011 at 7:37pm: super awesome

Comment by Brian Harms on April 18, 2011 at 12:28pm: @Tomasz

Haha, yeah that's the idea. A few other students and I actually just purchased a PP3DP (like a makerbot, but a bit more expensive and I think a bit nicer) I got to use one of these @ Future Cities Lab not too long ago and I was really surprised with how great of a resolution you can get with it - and also how awesome it is with automatically laying down support material.

Needless to say we are on a bit of a 3d printing stint here in our studio.

Comment by Brian Harms on April 18, 2011 at 12:16pm: Hey Andy, Thanks!

It is sort of a hybrid of grasshopper, firefly, and the native scripting language of a Pololu Maestro Servo Controller. The movement of the end effector and the extrusion nozzle is all defined by grasshopper. The definition takes any simple geometry and creates a basic toolpath which the nozzle follows. Alternatively you can manually input any other curve (like in my other video, the delta bot writing test).

The standing, sitting, and walking sequences are all recorded and stored on the servo controller as sequences (individual servo positions are key-framed manually) and then using the controller's native script editor I'm calling different sequences based on signals from arduino.

I never really had the time to figure out how to get arduino and the servo controller to share serial data (which I know people have done) and with firefly I believe it might involve a bit of firmata editing. So for now the servo controller receives 4 High/Low signals from arduino, essentially using binary combinations to trigger certain sequences (#0000 = Sit sequence, #0001 = stand sequence, etc.). It's a pretty rough workaround, so if anyone has any helpful resources it would be really appreciated!

So although it's not an entirely firefly driven project, you can control each function entirely from the grasshopper interface. This is great because everything is controlled via iPhone using touchOSC and gHowl.