This Project aims to promote a method of conserving both time and energy by allowing a person to turn on a fan without having to touch it using hand gestures and turn it off with hand gestures as well! This Project also measures the distance between a person and the fan and then dictates the speed of the fan based on that! This project utilizes multiple different parts with an Arduino board to truly create a project that acts as a culmination of various different lessons compiled into one project!

Supplies

Breadboard

Small Breadboard

LCD 16x2

UNO R3 Power Supply Module

Fan blade and 3- 6V Motor

Button

Arduino UNO R3

Breadboard Jumper Wire

1 Green LED Light

1 Red LED Light

Computer Webcam

Arduino Ultrasonic Distance Sensor

Clear Tape

Laptop/Monitor

Step 1: Assemble the Materials!

Collect the following materials to use!

Breadboard

Small Breadboard

LCD 16x2

UNO R3 Power Supply Module

Fan blade and 3- 6V Motor

Button

Arduino UNO R3

Breadboard Jumper Wire

1 Green LED Light

1 Red LED Light

Computer Webcam

Arduino Ultrasonic Distance Sensor

Clear Tape

Laptop/Monitor

Step 2: Put It All Together Like Here!

After arranging all the parts like so, let's go over a couple of reasons as to what each part plays in the actual Project!

Breadboard - To hold all the Wires together!

Small Breadboard - To hold the Ultrasonic Distance Sensor in place on top of a Laptop/Monitor! (We will get into why later!)

LCD 16x2 - To show how fast the Fan is spinning!

UNO R3 Power Supply Module - To regulate the power going through the Breadboard.

Fan Blade and 3 - 6V Motor - To give us the cool air we need!

Button - To act as a reset button in case the project is having some technical difficulties!

Arduino UNO R3 - To serve as our little microcomputer for this project!

Breadboard Jumper Wire - To connect the many components of our project together!

1 Green LED Light - To indicate whether the Fan is On!

1 Red LED Light - To indicate whether the Fan is Off!

Computer Webcam - if a person were to point the number 1 on their hand with their index finger and show it to the camera, the AI model would tell the fan to start working! Likewise, if there are all 5 fingers on display towards the camera, the AI model would tell the fan to stop working!

Arduino Ultrasonic Distance Sensor - This is used to tell how far away you are from the fan and then dictate the speed at which you will receive air!

Clear Tape - To steady the Small Breadboard in place atop the laptop/monitor.

Laptop/Monitor - To hold the Ultrasonic distance Sensor and be used to determine the distance between a person and a screen!

Make sure to tape the small breadboard with the Ultrasonic Distance Sensor to the top of a Laptop/Monitor!

Step 3: Arduino Code!

For This Specific Project, We will use 2 pieces of code! One for the Arduino Itself and One for the AI Model! This piece of code right here is for the Arduino itself!

String inputString = "";

#include <LiquidCrystal.h>

LiquidCrystal LCD(11,10,9,2,3,4,5);

#define trigPin 0

#define echoPin 1

//rs ,en d4 d5 d6 d7

int rs=10;

int e=9;

int d4=2;

int d5=3;

int d6=4;

int d7=5;

int speedPin= 6;

int directionPin1= 7;

int dirctionPin= 8;

int mSpeed;

int green=13;

int red=12;

String lastClass = "";

void setup() {

pinMode(trigPin, OUTPUT);

pinMode(echoPin, INPUT);

pinMode (speedPin,OUTPUT );

pinMode (directionPin1, OUTPUT);

pinMode (dirctionPin, OUTPUT);

pinMode(red, OUTPUT);

pinMode(green, OUTPUT);

Serial.begin(9600);

LCD.begin(16,2);

}

void loop() {

long duration, distance;

digitalWrite(trigPin, LOW);

delayMicroseconds(2);

digitalWrite(trigPin, HIGH);

delayMicroseconds(10);

digitalWrite(trigPin, LOW);

duration = pulseIn(echoPin, HIGH);

distance = (duration/2) / 29.1;

LCD.setCursor(0,0);

LCD.print("CURENT DISTANCE :");

LCD.setCursor(0,1);

LCD.print("");

LCD.setCursor(0,1);

LCD.print(distance);

LCD.print(" cm");

delay(250);

while (Serial.available()) {

char inChar = (char)Serial.read();

inputString += inChar;

if (inChar == '\n') {

inputString.trim(); // Remove newline and spaces

Serial.println("Received: " + inputString); // Debug

if (inputString == "Class1") { //Start

digitalWrite(red, LOW);

digitalWrite(green, HIGH);

digitalWrite(directionPin1, LOW);

digitalWrite(dirctionPin, HIGH);

digitalWrite(speedPin , 255);

lastClass = "Class1";

} else if (inputString == "Class2") { //Stop

digitalWrite(red, HIGH);

digitalWrite(green, LOW);

digitalWrite(directionPin1, HIGH);

digitalWrite(dirctionPin, LOW);

digitalWrite(speedPin , 0);

lastClass = "Class2";

} else if (inputString == "Class3" && lastClass == "Class1") { //Repeat

digitalWrite(red, LOW);

digitalWrite(green, HIGH);

digitalWrite(directionPin1, LOW);

digitalWrite(dirctionPin, HIGH);

digitalWrite(speedPin , 255);

} else if (inputString == "Class3" && lastClass == "Class2") { //Repeat

digitalWrite(red, HIGH);

digitalWrite(green, LOW);

digitalWrite(directionPin1, HIGH);

digitalWrite(dirctionPin, LOW);

digitalWrite(speedPin , 0);

}

inputString = "";

}

LCD.clear();

}

Step 4: AI Model Code!

<!DOCTYPE html>

<html>

<head>

<title>Thumbs Detection</title>

</head>

<body>

<h2>Thumbs Gesture Detection (Offline + Arduino)</h2>

<button onclick="init()">Start</button>

<p id="status">Waiting...</p>

let model, webcam, writer;

let lastClass = "";

let lastSendTime = 0;

const sendInterval = 400; // milliseconds between predictions

async function init() {

try {

document.getElementById("status").innerText = "Loading model...";

const modelURL = "model.json";

const metadataURL = "metadata.json";

model = await tmImage.load(modelURL, metadataURL);

webcam = new tmImage.Webcam(200, 200, true);

await webcam.setup();

await webcam.play();

document.body.appendChild(webcam.canvas);

const port = await navigator.serial.requestPort();

await port.open({ baudRate: 9600 });

const encoder = new TextEncoderStream();

encoder.readable.pipeTo(port.writable);

writer = encoder.writable.getWriter();

document.getElementById("status").innerText = "Camera + Serial Ready!";

window.requestAnimationFrame(loop);

} catch (err) {

console.error("Setup failed:", err);

document.getElementById("status").innerText = "Error: " + err;

}

async function loop() {

webcam.update();

const now = Date.now();

if (now - lastSendTime > sendInterval) {

await predict();

lastSendTime = now;

}

window.requestAnimationFrame(loop);

}

async function predict() {

const prediction = await model.predict(webcam.canvas);

let maxIndex = 0;

for (let i = 1; i < prediction.length; i++) {

if (prediction[i].probability > prediction[maxIndex].probability) {

maxIndex = i;

}

// If you have more Classes add them here. [..., "Class4", "Class5", ...], also add DisplayName#...

const classNames = ["Class1", "Class2", "Class3"];

const displayNames = ["start ", "stop", "rpeat "];

const selectedClass = classNames[maxIndex];

// Only send to Arduino if changed

if (selectedClass !== lastClass && writer) {

await writer.write(selectedClass + "\n");

lastClass = selectedClass;

}

// Show a friendly name on screen

document.getElementById("status").innerText = displayNames[maxIndex]; // You can change this text

}

</script>

</body>

</html>

F7I8NW2ILCHIOTO.webp
Download

Step 5: Final Product!

Voila! That should be how it looks when in use!

009fb216-a4ea-4bbd-9754-6a9c6d7b324b.MP4
Download

Introduction: Fanning Away Inefficiency! (FAI)