Change one thing at a time: Tune only GAIN, run, watch, repeat. If you change
GAIN, speed, and the threshold all at once, you can’t tell what helped.
Slow it down, speed it up: Lower DRIVE_SPEED and the robot follows more
forgivingly; raise it and a too-high GAIN starts to wobble. They work together.
Re-measure your readings: Read reflection() on the tape and floor again under
the room’s real lighting. Small changes in light change the numbers.
Try the Stretch
Two-sensor following: With two color sensors (one each side of the line), steer
from the difference between them. That follows the center of the line instead
of an edge — and stays centered better.
A loop course: Tape a line that loops back on itself and see if your tuning
survives the curves. Tight curves need a higher GAIN or a slower speed.
Bring the Week Together
You taught your robot to drive, solve a maze, steer by the gyro, dodge obstacles, and
follow a line. Pick two skills and combine them — like a line-follower that
stops when the distance sensor sees something ahead.
Write down, in your own words, what calibration meant in each session (wheel
size, cell size, gyro turn, trigger distance, line threshold). It was the same habit
every day.
Explore More (If Curious)
What’s the sharpest curve your robot can follow? What happens at a right-angle corner?
