New Project: Autobot

I am helping out a good friend of mine with his PhD project.  We are still working on the general concept but it is going to be along the lines of an Adaptive Embedded System.  a.k.a. Robot.  We will have a Google Code repository for it soon.  In the meantime, while we are still refining some of the project requirements I decided to try to figure out a better way for traversing complex terrain.  i.e. Driving through rocks and junk and stuff.  Robots getting stuck in sand, rocks etc is nothing new.  

After shopping online for a good R/C truck platform I happened to stumble upon this.  This looks really cool but I don't know about the quality of it, since it's a clearance toy from Toys R Us. It might be a good product, I just don't know.  

Anyways I was definitely inspired by this, but I thought I could do better.  Here is the first iteration of my neat-o Expanding Wheel (better name to be determined better).  The idea is to actuate the "paddles" using linear solenoids, 1 per paddle.  In the software the robot would automatically determine when it was stuck and open the "paddles" on the stuck wheel, making it more "grabby".  There are definitely some issues to figure out with this design, such as the amount of force/travel required for the solenoids, how to fit them in a confined space (the ideal wheel diameter pre-expansion is 3") and how to power everything (slip-rings anyone?).  Irregardless, here are some pictures of a CAD model:

1.  Wheel normal size.

2.  One paddle extended.

3.  All paddles extended.

4.  All paddles extended again, front view.