Building an Obstacle-Avoiding Robot with Raspberry Pi and Python

You don’t need to buy a fancy robot vacuum — you can build a two-wheel obstacle-avoiding robot yourself with a Raspberry Pi. It’s cheap, and you’ll learn both the hardware and software side.

Hardware

• A Raspberry Pi (3B+ or newer is comfortable)
• A two-wheel robot chassis + DC motors + motor driver (L298N)
• An HC-SR04 ultrasonic distance sensor
• A power bank for the Pi, and a separate battery pack for the motors

Control logic

The main loop is simple: measure the distance ahead → if something’s close, stop and turn → otherwise keep driving forward.

import RPi.GPIO as GPIO
import time

TRIG, ECHO = 23, 24
MOTOR_LEFT_FWD, MOTOR_RIGHT_FWD = 17, 27

GPIO.setmode(GPIO.BCM)
GPIO.setup(TRIG, GPIO.OUT)
GPIO.setup(ECHO, GPIO.IN)
GPIO.setup([MOTOR_LEFT_FWD, MOTOR_RIGHT_FWD], GPIO.OUT)

def read_distance_cm():
    GPIO.output(TRIG, True)
    time.sleep(0.00001)
    GPIO.output(TRIG, False)

    start = time.time()
    while GPIO.input(ECHO) == 0:
        start = time.time()
    while GPIO.input(ECHO) == 1:
        stop = time.time()

    return (stop - start) * 34300 / 2

def drive_forward():
    GPIO.output(MOTOR_LEFT_FWD, True)
    GPIO.output(MOTOR_RIGHT_FWD, True)

def stop():
    GPIO.output(MOTOR_LEFT_FWD, False)
    GPIO.output(MOTOR_RIGHT_FWD, False)

try:
    while True:
        distance = read_distance_cm()
        if distance < 20:
            stop()
            time.sleep(0.3)
            # turn right to avoid the obstacle — drive left/right wheels separately
        else:
            drive_forward()
        time.sleep(0.1)
finally:
    GPIO.cleanup()

Real-world gotchas

• Power the motors and the Pi separately — sharing one supply tends to brown-out and reset the Pi when the motors kick in.
• Ultrasonic sensors reflect poorly off angled or soft surfaces, so add a timeout to read_distance_cm.
• Cheap wheels rarely spin at exactly the same speed on both sides — calibrate with per-side PWM instead of the simple on/off used in this example.

Where to go from here

Once obstacle avoidance is reliable, extend it: log the path travelled for a basic map, add a Pi camera for object detection, or drive it remotely from a small Flask web app. This is a great project for seeing the gap between “writing code” and “code that controls something physical” — hardware quirks will teach you things pure software never will.