DIY Autonomous Talking Robot with Raspberry Pi

⚡ Circuit Lab

Capabilities Architecture Materials Circuit Build Software Code FAQ

🤖 Advanced Robotics Project · Raspberry Pi

Build a Robot That Moves, Sees & Talks — All On Its Own

A Raspberry Pi robot that rolls around a room by itself, steers clear of furniture, glances around with a camera, and answers your questions out loud. This is the big one — every skill from our earlier tutorials, combined into one real robot.

🧠 What This Robot Actually Does

"Autonomous" just means it makes its own small decisions without you holding a remote. This build stands on four skills working together:

🚗

Move

Drives forward and steers using two motors

📡

Sense

Measures distance with an ultrasonic sensor

👁️

See

Looks around using a camera module

🎤

Talk

Listens and answers spoken questions

📌 Setting expectations

This robot won't be a movie-style AI — it navigates using simple obstacle-avoidance rules, and it answers questions using a keyword-matching system you can expand yourself (with a real AI upgrade path in the Extensions section). That's exactly how real robotics engineers start: simple rules first, smarter behavior later.

🗺️ System Architecture

The Raspberry Pi is the robot's brain — every other part is a sense or a limb connected to it.

🧰 What You'll Need

• 1× Raspberry Pi 4 (or 3B+)
• 1× microSD card, 16GB+, with Raspberry Pi OS
• 1× Robot chassis kit (2 DC gear motors + wheels + caster wheel)
• 1× L298N motor driver module
• 1× HC-SR04 ultrasonic sensor
• 1× 1 kΩ + 1× 2 kΩ resistor (ECHO voltage divider)
• 1× Raspberry Pi Camera Module (or USB webcam)
• 1× USB microphone
• 1× Small USB or 3.5mm speaker
• 1× 7.4V battery pack for motors + separate 5V power bank for the Pi
• Jumper wires, screws/standoffs, on/off switch

🔋 Two power supplies, not one

Motors cause sudden power dips that can crash a Raspberry Pi. Power the motors from their own battery pack, and the Pi from a separate 5V supply, sharing only a common ground.

🔌 The Circuit Diagram

Just like our ESP32 glasses build, the ECHO line needs a voltage divider — Raspberry Pi GPIO pins are also only safe up to 3.3V.

🛠️ Hardware Build Steps

1

Assemble the chassis

Mount the two DC gear motors and the caster wheel onto the robot chassis frame, then attach the wheels.

2

Mount the L298N driver

Screw the motor driver board onto the chassis, then wire OUT1/OUT2 to the left motor and OUT3/OUT4 to the right motor.

3

Wire the driver to the Pi

Connect IN1–IN4 and ENA/ENB to the GPIO pins shown in the diagram, and tie the driver's GND to the Pi's GND.

4

Mount the ultrasonic sensor

Fix the HC-SR04 facing forward at the front of the chassis, and wire it in through the voltage divider.

5

Attach the camera

Connect the camera ribbon cable to the Pi's CSI port, then mount the camera facing forward, near the sensor.

6

Plug in mic and speaker

Connect the USB microphone and speaker to any free USB ports (or the audio jack for the speaker).

7

Connect both batteries

Wire the motor battery pack to the L298N's power input, and power the Pi separately from its own 5V supply — share only ground.

💾 Software Setup

A few one-time steps get the Pi ready before running any robot code.

1

Flash Raspberry Pi OS

Use Raspberry Pi Imager to write Raspberry Pi OS onto the microSD card, then boot the Pi and connect it to Wi-Fi.

2

Enable the camera

Run sudo raspi-config, open Interface Options, and enable the Camera.

3

Install the required libraries

Open a terminal and run the commands below.

sudo apt update && sudo apt full-upgrade -y
sudo apt install python3-pip python3-opencv espeak -y
pip3 install RPi.GPIO pyttsx3 SpeechRecognition pyaudio

💻 The Code

This is the robot's whole brain in one script. It runs two jobs at the same time: one thread drives around and dodges obstacles, while another thread listens for questions and answers them out loud.

autonomous_robot.py

# DIY Autonomous Talking Robot — Raspberry Pi
# Two brains running together: navigation + voice Q&A

import RPi.GPIO as GPIO
import time, threading
import speech_recognition as sr
import pyttsx3

# ---------- Motor setup (L298N) ----------
IN1, IN2, IN3, IN4 = 17, 27, 22, 23
ENA, ENB = 18, 13  # PWM-capable pins

GPIO.setmode(GPIO.BCM)
GPIO.setup([IN1, IN2, IN3, IN4, ENA, ENB], GPIO.OUT)
pwmA = GPIO.PWM(ENA, 1000)
pwmB = GPIO.PWM(ENB, 1000)
pwmA.start(70)
pwmB.start(70)

def forward():
    GPIO.output(IN1, GPIO.HIGH); GPIO.output(IN2, GPIO.LOW)
    GPIO.output(IN3, GPIO.HIGH); GPIO.output(IN4, GPIO.LOW)

def stop():
    GPIO.output(IN1, GPIO.LOW); GPIO.output(IN2, GPIO.LOW)
    GPIO.output(IN3, GPIO.LOW); GPIO.output(IN4, GPIO.LOW)

def turn_left():
    GPIO.output(IN1, GPIO.LOW);  GPIO.output(IN2, GPIO.HIGH)
    GPIO.output(IN3, GPIO.HIGH); GPIO.output(IN4, GPIO.LOW)
    time.sleep(0.4)
    stop()

# ---------- Ultrasonic sensor ----------
TRIG, ECHO = 5, 6
GPIO.setup(TRIG, GPIO.OUT)
GPIO.setup(ECHO, GPIO.IN)

def get_distance_cm():
    GPIO.output(TRIG, False)
    time.sleep(0.05)
    GPIO.output(TRIG, True)
    time.sleep(0.00001)
    GPIO.output(TRIG, False)

    start = time.time()
    timeout = start + 0.03
    while GPIO.input(ECHO) == 0 and time.time() < timeout:
        start = time.time()

    stop_t = time.time()
    timeout2 = stop_t + 0.03
    while GPIO.input(ECHO) == 1 and time.time() < timeout2:
        stop_t = time.time()

    elapsed = stop_t - start
    return (elapsed * 34300) / 2  # cm

def navigate_loop():
    while True:
        dist = get_distance_cm()
        if dist < 20:
            stop()
            turn_left()
        else:
            forward()
        time.sleep(0.1)

# ---------- Voice Q&A ----------
engine = pyttsx3.init()
recognizer = sr.Recognizer()

ANSWERS = {
    "your name": "I'm RoboBuddy, your homemade robot friend!",
    "how are you": "All circuits running smoothly, thanks for asking!",
    "what can you do": "I can roam around, dodge obstacles, and chat with you!",
}

def speak(text):
    engine.say(text)
    engine.runAndWait()

def voice_loop():
    mic = sr.Microphone()
    while True:
        with mic as source:
            recognizer.adjust_for_ambient_noise(source)
            audio = recognizer.listen(source)
        try:
            question = recognizer.recognize_google(audio).lower()
            reply = "I don't know that yet, but I'm learning!"
            for key, ans in ANSWERS.items():
                if key in question:
                    reply = ans
                    break
            speak(reply)
        except sr.UnknownValueError:
            pass  # didn't catch that — keep listening

# ---------- Run both brains at once ----------
if __name__ == "__main__":
    t1 = threading.Thread(target=navigate_loop, daemon=True)
    t2 = threading.Thread(target=voice_loop, daemon=True)
    t1.start(); t2.start()
    try:
        while True:
            time.sleep(1)
    except KeyboardInterrupt:
        stop()
        GPIO.cleanup()

🧠 How can it drive AND listen at once?

This uses threading — the Pi runs two loops "at the same time" by rapidly switching between them, just like how you can walk down a hallway while chatting with a friend without stopping either activity.

❓ Frequently Asked Questions

Does the robot need the internet to answer questions?

The starter version uses Google's free online speech-recognition service, so yes, it needs Wi-Fi. An offline engine like Vosk can replace it if you want the robot to work without internet.

Is this "real" artificial intelligence?

Not yet — the starter code just matches keywords in a dictionary. It's a great foundation to upgrade with a real AI language model, covered in the Extensions section below.

How does it "identify surroundings" exactly?

The ultrasonic sensor only measures distance to the nearest object — it can't tell what something is. The camera adds true "seeing" capability, which you can start simple (detecting colors or shapes) and grow into full object recognition.

The robot keeps resetting when it starts moving — why?

This almost always means the Pi is sharing power with the motors. Give the motors their own separate battery pack, and keep the Pi on its own supply.

Why does the ultrasonic ECHO pin need a voltage divider here too?

Just like the ESP32, Raspberry Pi GPIO pins are only safe up to 3.3V, while the HC-SR04's ECHO output is 5V — the resistor divider steps it down safely.

🚀 Take It Further

Upgrade to real AI answers

Swap the keyword dictionary for a call to an AI language model API, so the robot can genuinely answer open-ended questions.

Add real object recognition

Run a lightweight TensorFlow Lite model on the camera feed so the robot can say what it actually sees — "I see a chair" instead of just a distance number.

Map the whole room

Add a rotating ultrasonic sensor or a small LIDAR module to build a simple map of a room as the robot explores it.

Control it from your phone

Add a small Flask web server on the Pi so you can drive the robot manually from a browser on your phone when you want to.

🤖 Built for curious young makers exploring robotics and AI fundamentals. Build and test with an adult nearby, especially around moving parts and battery wiring.