Following lines with a P controller

The problem with on and off

A line follower should drive along a black line. The simplest idea: “If I see black, steer left, otherwise steer right.” That works – but the robot keeps jerking back and forth.

A controller is better: it steers hard when the robot is far off the line and gently when it’s almost right.

The P controller

The colour sensor sits exactly on the edge of the line. There it measures a medium brightness, the target value. The difference from it is called the error. The P controller simply calculates steering = KP * error.

from pybricks.hubs import PrimeHub
from pybricks.pupdevices import Motor, ColorSensor
from pybricks.parameters import Port, Direction
from pybricks.robotics import DriveBase
from pybricks.tools import wait

hub = PrimeHub()

left_motor = Motor(Port.A, positive_direction=Direction.COUNTERCLOCKWISE)
right_motor = Motor(Port.B, positive_direction=Direction.CLOCKWISE)
color_sensor = ColorSensor(Port.C)
drive_base = DriveBase(left_motor, right_motor,
                     wheel_diameter=56, axle_track=114)

TARGET_BRIGHTNESS = 50        # sensor value exactly at the line's edge (0-100)
KP = 1.2                      # the larger, the stronger the correction
BASE_SPEED = 120              # mm/s - better to start slowly!

while True:
    brightness = color_sensor.reflection()
    error_value = brightness - TARGET_BRIGHTNESS   # deviation from the line's edge
    steering = KP * error_value                    # only the P component
    drive_base.drive(BASE_SPEED, steering)
    wait(10)

Press Run and watch the robot follow the line. Then experiment:

Remember: drive_base.drive(speed, turn_rate) keeps driving until a new command comes. That’s why it fits perfectly into a control loop.

See the readings: In the example above, add print("error:", error_value, "steering:", steering) below the steering line (same indentation). The simulator then draws the values as curves while the robot drives – so you can see how much it swings back and forth with KP = 6. With 💾 CSV you save the readings for a spreadsheet (the file stays on your device).

The book then continues with the PID controller, which also learns from the past (I) and from the change (D) – so the robot drives fast and cleanly even through tight curves.