Julia and I had a lot of fun working with the sciborgs, but it definitely got frustrating at times when both hardware and software were malfunctioning, and we couldn’t quite pinpoint our mistakes. This didn’t happen until the line following exercises, though.
1 – Making Functions
2 – Set up your sciborg
We used Sciborg #3. It needed some lubrication on the front ball, but was otherwise ready to roll.
3 – Work with your sciborg
b – The minimum speed (figured out empirically) is 73
c – Hard turn
d – Gentle turn
e – The sciborg took 18 seconds to travel 10 feet. When we ran the timed program, the sciborg stopped exactly where we thought it would stop, with the back wheels on the line. If it were to be inaccurate, I would attest it to not instantaneous communication.
4 – Feedback with sensors
MotorEncoder: The motor position doesn’t go back to zero for two reasons. First of all, the motor cannot stop instantaneously. Secondly (related to this) the motor requires time to slow and switch directions.
LightSensorNXT: Black reads as 0, while white reads as 97. THe LED light neither helps nor hurts the sensitivity. On the one hand, I would expect it to, since it’s another source of light, but on the other hand that would be a spectacularly impractical design. The optimal height from the ground where the light sensor doesn’t cast a shadow is about 2 cm.
5 – Drive Straight
Our results are rather consistent on the concrete, but the sciborg can’t turn on carpet. Our results are somewhat reproducible.
6 – Drive Straighter
The sciborg drives extremely straight, and these are very reproducible results. I prefer straight proportional to straight bang-bang.
7 – Bang-bang Light Following
We had an idea of how to carry this one out, but the sciborg kept getting stuck on the tape. Our original idea was for the sciborg to drives straight as long as it saw the line, then, if it got off, spin in one direction until it sensed the line brightness once more. Next, we tried to accomplish this with only two pieces of code; one told the sciborg to turn left if the light sensor reading was in certain range, otherwise it would turn right. The idea was for the sciborg to wiggle and weave down the white path. Both of these approaches were mildly successful.
8 – Proportional Line Following
We approached this similar to the bang-bang with only left and right turns, except the amount that the motor turned by was proportional by a factor kp to the error of the light reading.
