Structure
Our final design was not all that different from our original design
except for a couple of fixes here and there. Fortunately, since
David was an expert with legos, we didn't have any major design
flaws. Pictures are worth more than a thousand words so just look
at the picture of our robot to check out our design. But basically,
we had a beautifully symmetric robot with 4 big wheels, two on each
side, in the back and a smaller wheel in the middle. Dimensions
were 1 ft. *1 ft. * 5 in. In the front, our robot had 2 gates
that opened and closed like the claws of a crab. Just look at the
pictures, you'll understand.
DK31 Front
DK31 Rear
DK31 Side
DK31 Top
Drive/Steering
Drive/Steering DK31 used a differential drive combined with a steering
wheel in the middle. The steering wheel in the middle was powered
by a servo which allowed the robot to have smoother and more accurate
turns. We mainly used shaft encoders on the back wheels to monitor
the rotational distance travelled when making turns, and driving
straight. Also, we were able to "veer" - not drive straight,
but lean towards one side - using the shaft encoders, by making
one wheel rotate faster than the other.
Gears/Motors
DK31 had a 45:1 wheel ratio. We used two bi-directional motors
to power each side. Our robot had very little noise and gear friction
when driving - DK31 was a smoothie. DK31 was probably one of the
most quiet robots in the competition. We had almost no problem with
our gears compared to other teams. Whereas other teams had gears
grinding and breaking, we never had a problem with the gears, as
we followed the "book" in gear ratios and gear structuring.
Sensors
Sensors. Here was our robot's biggest flaw - no sensors. Ok, we
used a few - we used leds and phototransistors (for finding out
starting orientation and our team color) and the start light detector
(for the start code). We had many ideas - we had two digital sensors
within the gates (claws) to determine whether we had a ball, and
we had planned to attach digital sensors on the back and the side
of our robot to give it the ability to detect and follow walls more
efficiently. Unfortunately, because of an error that occurred 5
hours before impounding (and we couldn't find the cause for 3 hours),
we couldn't incorporate our ideas and the numerous sensors in our
robot - so essentially, our robot lacked a lot of knowledge. But
thanks to the claws, we were able to follow walls very effectively
- the opening of the claws gave the robot an effective turning and
wall-following (sticking to the wall rather) ability. However, making
use of more sensors would have made our robot definitely smarter
and capable of winning.
|