Flexibot 0.1 Update

Here's a short post to show the progress I've made on the flexibot development. I've been spending the majority of my time lately fixing things around the house, and haven't put a lot of effort into the project, but the first version is now what I'd call complete.

First off, I finished re-tooling the development platform by getting some little cable connectors made by Du-bro, cutting an adapter for an aluminum tubing leg on my ShapeOko, and making a terrible looking, but functional bent tubing leg.  The resulting robot has one active leg with two degrees of freedom, and demonstrates that a Sugru flexible link joint will support at least a little weight (this one is lifting four AAA ni-cad batteries) and is useful for walking.

I didn't edit the video down, so it's kind of long - sorry about that.  It does showcase most of what flexibot 0.1 can do right now.  I hope you enjoy, at least part of it.

Drawbot 3.0 Makes Its Mark

I finally got the finishing touches on the new version of the Drawbot, and it drew its first images today.

I will be fully documenting the build, and perhaps offering boards and such.

Update on flexibot

A few days ago I cast two more flexible links for my flexibot development platform.  My initial link was a #1 SRFLG (standard RobotGrrl flexible link guage) - about 4 mm diameter.  The two new ones I cast are a #2 SRFLG - about 6 mm in diameter, and #3 SRFLG - about 7 mm in diameter.

After they cured, I started flexing them with my fingers, just to get a feel for the flexibility.  My #3 quietly failed.

Close inspection of the crack seems to indicate there was internal structure to the link, so the Sugru isn't homogeneous.  Either I didn't get the Sugru squished together well enough, or didn't let it cure long enough. Oh well.  I've made 4 links, and experienced 50% failure rate.  I guess I have something to learn about casting Sugru.

Now, on to the good part.  I put the first flexibot frame together really quickly, and without much planning.  So, I discovered that hot glue isn't real good at take-apart-and-put-back-together prototyping.  I went back to CAD, and modified the frame design to incorporate interchangeable link base plates.

Here you can see my #1 and #2 links on their plates...

Here's the new frame...

and the frame with a link installed, from the front...

and the back...

I haven't yet figured out how I'm going to attach the control cables (actually I'm using polyester thread originally designed to stitch sails together) and still make them interchangeable.

If you have any ideas, leave me a comment.  Thanks for reading.

New Line of Research

One thing I've been thinking about for years, is flexible robots.  I'm a pretty good fabricator, but the precision  and complexity, required to make useful multi-degree of freedom joints (like your shoulder) has always seemed a little daunting, especially if you try to make them compact.

Some time ago Erin Kennedy, a brilliant, creative, visionary robot builder, and long distance friend posted a couple of tidbits on a tentacle mechanism she was pursuing.  The first iterations I saw were 3D printed parts connected with machine screws.  After seeing a video put out by the folks who make Sugru illustrating the use of 3D printed parts as molds to form Sugru parts, I suggested to Erin she try using Sugru as a flexible link between her 3D printed parts.  The folks from Sugru caught wind of the project, and donated some material to the cause, and Erin ran like mad, quickly bringing the idea into a working model.

Shortly after Erin worked out her design, she started offering 3D printed parts for sale.  I immediately placed an order, and was quickly rewarded with dozens of electric green and orange parts, direct from Erin's lab.

That night I cast a couple of links, and in the morning had one good one to start with.  The other one stuck in the mold, and tore coming out.  Once again, let me shout out to the folks who make Sugru, they sent me a donation as well, and I'm finally putting it to good use.

This weekend I used Sketch-up to design a framework, cut the parts on my ShapeOko, and TADA - Flexibot 0.1 is born.

Right now I've got the servos hooked up to an RC receiver, and I'm just driving it around with the joystick on my transmitter.  This is just the beginning, and once I get a little further on the mechanics, I'll turn it into a proper robot with autonomous control.

Drawbot update, and ShapeOko's first real work

Last night I spent some time in SketchUp designing a pen carrier for the new drawbot.  This morning I exported all the parts to .dxf, pulled them into CamBam+, generated the gcode, and cut parts on the ShapeOko.  Here's the results.

My New Toy Takes its First Steps

About three weeks ago my wife bought me a really cool gift, and tonight my new ShapeOko CNC mill took made its first moves.  The crappy video is it running the ShapeOko "Hello World" job.

Drawbot software on it's way

To paraphrase "Cool Hand Luke" - what we have here, is the ability to communicate!

In this video the comms board and the pen carrier board are connected up and working for the first time.  The comms board connects to my desktop via USB, and is connected to the pen carrier board via RS-485. When I send a command out the USB, you can see the transmit and receive LEDs flash, then the pen carrier board responds by changing the servo's position.

The protocol the boards are running over the 485 link is dead simple right now, and not capable of multi-drop communication.  So, the next step is to implement a new protocol.  The carrier board also only implements binary servo positions up, and down.  The new protocol will give the ability to send data through the link, so I can control how far up/down the pen is.  With this I'll be able to use brush tip markers, and vary the line thickness drawn by the 'bot.