Robotics
It all started way back in 1993 when I started going to Carnegie
Mellon. I saw what cool things could be done with robotics and all the
different disciplines which it combines. That got me hooked. These are
some of the projects I've worked on in the past and even some of the hobby
robotics I work on now.
CMU Projects
Daedalus
- apparently this project is so old they don't even have an official website
on it anymore...
Nomad
- The Atacama trek happened the summer after I left the project. That wasn't
the end for this guy though. They took him to Antarctica a couple times
to search
for meteorites.
NASA Ames Research Center
Dexter - I worked with the Intelligent Mechanisms Group in the summer
of 1998 before things got rearranged and the IMG became part of the Autonomy
and Robotics Group. Dexter was a six-legged robot design that was aimed
towards low power consumption. The project never had a web page, but I'm
going to try and get an image up here from my computer models of it. I
developed a simulation environment that allowed me to test out control
strategies for the robot. The strategy I eventually used (in the simulation)
was based on the control of Robot
II from Case Western's Biologically
Inspired Robotics Laboratory. Check out their videos for this robot
if you haven't seen them before. This has been THE robot that has inspired
me over the years. The way it moves almost makes me believe it's alive.
Sidewinder [pic1][pic2]
- didn't work, but it's the idea that matters
Sidewinder II - The same idea as above, but properly implemented.
Completely overpowered computing-wise. The thing had a 68040 running VxWorks
along with ethernet to control 1 motor and 8 solenoids (to move pneumatic
cylinders for the legs). I can't seem to find a picture of this thing.
Senior Thesis Research Project at CMU
Black Widow - See my research
page.
Hobby Projects
I still have Black Widow and plan to bring her back to life some day. I
had to rip out her electronics when I left CMU because they weren't mine.
Oh, and just so you don't mistakenly think I did all the hard work, Ryan
Miller and I built the mechanism during my junior year at CMU.
My more recent project is ripping out the guts and reusing this little
guy:
It's
called a Motor Ball and was intended to be driven by remote control. It's
total lack of any kind of stability and only limited controls on the motor
speed (full ahead, stop, and full back on each side) make driving it nearly
impossible. Sounds like perfect robot material for me.
The wires sticking out of the top are connected to encoders that I'm
building inside each wheel/hemisphere. The hole the wires are sprouting
out of normally holds an antenna/3rd wheel that keeps the thing from just
sitting still while the motor and electronics box inside flips around and
around.
I'll update this as I add things to the little bugger. Current plans
are for some solid-state gyros or accelerometers to get some measurement
of the side to side oscillations this thing tends to get itself into. It
ought to be an interesting control problem.