Project Goals

• Build a functioning car that can drive and turn
• Control with two DC motors, each separately controlled with a different Pi Pico
• Train and use a machine learning model that can recognize hand signals to start and stop the car
• Use an OpenMV camera to scan custom AprilTags to control movement parameters of the car

Project Documentation

Left to right: the steering wheel to control car’s rotation, the pyscript UI for the machine learning model to start and stop the car, the car

https://youtu.be/AvQDAMbo5R4

https://youtu.be/jLMJ0VvUBS0

Successes and Challenges

• Successes
• Challenges

Codespace

while True:
    
    clock.tick()
    img = sensor.snapshot()
    for tag in img.find_apriltags():
        img.draw_rectangle(tag.rect(), color=(255, 0, 0))
        img.draw_cross(tag.cx(), tag.cy(), color=(0, 255, 0))
        
        tag_id = tag.id()
        tag_name = tag.family()
        tag_rotation = (180 * tag.rotation()) / math.pi
        
        currentTag = f'{tag_id}, {tag_name}'

        print("Tag Family %s, Tag ID %d, rotation %f (degrees)" % (tag_name, tag_id, tag_rotation))

    if currentTag == f'42, 16':
        client.publish("ME35-24/kai", str(tag_rotation))

This is the main code logic for the camera which reads the AprilTag. It captures images from a camera and searches for AprilTags within each frame. When a specific tag (ID 42, family 16) is detected, it publishes the tag's rotation angle to the MQTT topic "ME35-24/kai.”

try:
    recent_msg_float = float(recent_msg)
    if 85 < recent_msg_float < 95:
        if counter < 0:
            stepper.right(0 - counter)
        elif counter > 0:
            stepper.left(counter)
        counter = 0
            
    elif 75 < recent_msg_float < 85:
        if counter < -1: 
            stepper.right(abs(counter + 1))
        elif counter > -1:
            stepper.left(abs(counter - 1))
        counter = -1
    elif 95 < recent_msg_float < 105:
        if counter < 1:
            stepper.right(1 - counter)
        elif counter > -1:
            stepper.left(abs(counter - 1))
        counter = 1
       
except ValueError:
    if recent_msg == 'start':
	    left_motor.forward()
    elif recent_msg == 'stop':
      left_motor.stop()

This code snippet is the main logic for the MQTT commands and the steering. It adjusts the direction and steps of a stepper motor based on the value of recent_msg. It ensures the motor moves within specific ranges of rotation and handles special commands like "start" and "stop" to control a left motor's operation, which are sent by the machine learning model.