Post-Game Analysis
What was good about our robot
There were a few nice features
about our robot. Physically, our two special features, the arm and
the ramp, worked very well and were masterfully constructed. I
believe we were the only robot that used a ramp to pick up balls.
Another nice feature of our robot was its brains, the code. Our robot
wasn't actually very smart, but we put timeouts on everything so that
it never just sat banging its head against the wall. Several times in
competition our robot was able to recover after getting off course due
to these timeouts. The very best thing about our robot was how it
oriented itself. We used three sensors and nifty code to figure out
orientation. In every single orientation trial from when we loaded
the code to when 6.270 was over the robot oriented correctly without
fail. Sadly, this was the only thing about our robot that was
perfect.
What was horribly, horribly wrong with our robot
Our robot
couldn't drive. This is a very bad problem to have. Basically, we
couldn't drive in a straight line, we couldn't use wall following
because of our robot's sharp corners, and we discovered this too late
to implement line following. And to top it off, we couldn't turn 90 degrees consistently. I think most of our problems were due to our robots design. Our gear ratio was probably too low and our center of gravity was too high. What made matters worse was that the strength of the motors would greatly decrease while the robot was driving, so if it started out going straight, it would probably end turning into a wall.
What happened in competition
Our robot's record was 0-2
and we managed to score ZERO balls in competition play. In both of
our matches, poor turns cost us. Our robot oriented correctly, went
to pick up the top ball with the ramp(worked beautifully), backed away
from the wall and did a 180. It was this turn that got us. Both
times the turn was WAY off and our robot went straight into the
obstacle. After seizing on the obstacle for awhile the robot would
carelessly drop the ball it had then flap its arm for the remainder of
the match. At least we didn't take the easy way out and just score
the easy bottom ball.
What we learned
The key to having a good robot is having a
good robot. Forget about strategy, forget about code. Those things
can't make up for a poor robot. If you have a robot that can drive
straight and turn perfectly, you can do anything with it. A reliable robot can beat a fast robot with a decent strategy anyday.
Back to the Main Page