[Bomberofdoom]

Accelerometers and Gyros have other applications than for the drivetrains:

In the 2007 game, where at the end-term of the match (20 seconed before end), you would have to lift up two of your other alliance robots, our team has developed a jack system to lift the two robots.
The robot had two "wings" (flat plates from each side that would open up). In the inner corners of the robot were 4 poles that were attached to a motor (The robot. I'm afarid the good pictuers are gone, but it shows most of the robot). These poles were designed to be pushed downwards (more exactly, rotated downwards) when the motors where activated. At some point all 4 poles would reach the ground, and start pushing the ground, which, thanks to Newtown's third law (If object A exerts a force F on object B, then object B exerts an equal and opposite force –F on object A - http://www-istp.gsfc.nasa.gov/stargaze/Snewton3.htm), would result in the lifting of the robot itself, with the two other robots on each of the plates (overall lifting 2 robots "plus one more").

The major problem we had in our design is that we had only a total of 2 of the main motors we planned to use to rotate the poles, so we were forced to use 2 lesser powerful motors, which would result in that two poles would push the robot earlier than the two others, resulting in the tilting o the two other robots on our plates and thier fall or our poles braking from the pressure of 180 KG.

We resolved on using a 2-axis accelerometer to measure the the tilting to each direction - forwards/backwards and sideways. By integrating the change in each axis, we were able to come up with a mathematical algorithm for sending the correct speed to each motor, in order to compinsate for any tilts that would might have been caused from the varying dimension and weight of the alliance robots.

I've seen a team use a Gyro for thier robot's arm, to measure the rate of change in the arms movements to know what its current position is (by integeration, as described before me).

Note that usually a potentiometer is more effective and easier to use than a Gyro for this kind of arm positioning application. The potentiometer revolves with the arm, and as it revolves, it chances the resistance inside of it, and therefore the voltage that is recieved from it changes also. If the potentionmeter is constant, you get a constant voltage of that position. If your position would change, so would the voltage - even if you would stop in a new position, it would send the voltage for that position, instead of giving you the "rate of change" like the Gyro - no need for integration here in order to know what is your position.

Hope I gave a good lesson.