Friday, 17 July 2026

AI Smart Voice Controlled Wheelchair for Disabled Persons

"GPIO 26", "Left Motor IN2" => "GPIO 27", "Right Motor IN1" => "GPIO 14", "Right Motor IN2" => "GPIO 12", "Ultrasonic TRIG" => "GPIO 5", "Ultrasonic ECHO" => "GPIO 18", "Battery Voltage Sensor" => "GPIO 34", "Current Sensor" => "GPIO 35", "Temperature Sensor" => "GPIO 32", "Emergency Button" => "GPIO 4", "Buzzer" => "GPIO 2" ]; $workflowSteps = [ "User gives voice command", "Speech is converted to text", "AI Agent identifies the command", "n8n validates the command", "ESP32 receives the command", "Safety sensors are checked", "Motor driver controls the wheelchair", "Sensor data is sent to the cloud", "Google Sheets stores data", "ThingSpeak displays data", "Telegram sends notifications", "AI predicts power consumption" ]; function showSectionTitle($title) { echo "

" . htmlspecialchars($title) . "

"; } ?> <?php echo $projectTitle; ?>

AI + ESP32 + IoT + n8n Automation + Telegram Voice Alerts

Google Sheets + ThingSpeak Cloud Dashboard + AI Power Prediction

The wheelchair is controlled through natural voice commands. The AI Agent understands the user's spoken command and converts it into an appropriate wheelchair movement command.

The ESP32 acts as the main controller. It communicates with motors, sensors, cloud platforms and automation services.

Project Working

  • Control wheelchair using voice commands.
  • Improve mobility for disabled persons.
  • Detect obstacles automatically.
  • Monitor battery voltage and percentage.
  • Measure motor current and power consumption.
  • Send emergency notifications.
  • Use AI for intelligent decision making.
  • Store IoT data in Google Sheets.
  • Visualize data using ThingSpeak.
  • Use n8n for automation.

USER / PATIENT | v VOICE COMMAND | v AI VOICE PROCESSING | v n8n AI AGENT | v ESP32 CONTROLLER | +----------------+ | | v v SENSORS MOTOR DRIVER | | v v CLOUD PLATFORM DC MOTORS | +----------------+ | | | v v v ThingSpeak Google Telegram Sheets Alerts
+----------------------+ | BATTERY | | 12V / 24V | +----------+-----------+ | +--------------+--------------+ | | v v +-------------+ +--------------+ | DC-DC BUCK | | MOTOR DRIVER | | 12V to 5V | | HIGH CURRENT | +------+------+ +------+-------+ | | v v +------------------------------------------+ | ESP32 | | | | GPIO 26/27 ------ LEFT MOTOR CONTROL | | GPIO 14/12 ------ RIGHT MOTOR CONTROL | | | | GPIO 5/18 -------- ULTRASONIC SENSOR | | GPIO 34 ---------- VOLTAGE SENSOR | | GPIO 35 ---------- CURRENT SENSOR | | GPIO 4 ----------- EMERGENCY BUTTON | | GPIO 2 ----------- BUZZER | +------------------------------------------+
$pin): ?>
Device ESP32 Pin
START | v INITIALIZE ESP32 | v CONNECT TO WIFI | v READ SENSORS | v EMERGENCY BUTTON PRESSED? | YES | v STOP MOTORS | SEND TELEGRAM ALERT | v LOOP NO | v OBSTACLE LESS THAN 30 CM? | YES | v STOP MOTORS | SEND WARNING | v LOOP NO | v RECEIVE AI COMMAND | v VALIDATE COMMAND | v CONTROL MOTORS | v SEND DATA TO CLOUD | v LOOP
Voice Command AI Intent 

#include <WiFi.h>
#include <HTTPClient.h>
#include <ArduinoJson.h>

const char* ssid = "YOUR_WIFI_NAME";
const char* password = "YOUR_WIFI_PASSWORD";

const char* n8nWebhook =
"https://YOUR-N8N-DOMAIN/webhook/wheelchair-data";

#define LEFT_IN1 26
#define LEFT_IN2 27

#define RIGHT_IN1 14
#define RIGHT_IN2 12

#define TRIG_PIN 5
#define ECHO_PIN 18

#define EMERGENCY_PIN 4
#define BUZZER_PIN 2

String movement = "STOP";

void setup() {

    Serial.begin(115200);

    pinMode(LEFT_IN1, OUTPUT);
    pinMode(LEFT_IN2, OUTPUT);

    pinMode(RIGHT_IN1, OUTPUT);
    pinMode(RIGHT_IN2, OUTPUT);

    pinMode(TRIG_PIN, OUTPUT);
    pinMode(ECHO_PIN, INPUT);

    pinMode(EMERGENCY_PIN, INPUT_PULLUP);
    pinMode(BUZZER_PIN, OUTPUT);

    stopMotors();

    WiFi.begin(ssid, password);

    while (WiFi.status() != WL_CONNECTED) {

        delay(500);

        Serial.println("Connecting to WiFi...");

    }

    Serial.println("WiFi Connected");

}

void loop() {

    bool emergency =
        digitalRead(EMERGENCY_PIN) == LOW;

    float distance = readDistance();

    if (emergency) {

        stopMotors();

        digitalWrite(BUZZER_PIN, HIGH);

        movement = "EMERGENCY_STOP";

    }

    else if (distance > 0 && distance < 30) {

        stopMotors();

        digitalWrite(BUZZER_PIN, HIGH);

        movement = "OBSTACLE_STOP";

    }

    else {

        digitalWrite(BUZZER_PIN, LOW);

    }

    delay(5000);

}

void moveForward() {

    digitalWrite(LEFT_IN1, HIGH);
    digitalWrite(LEFT_IN2, LOW);

    digitalWrite(RIGHT_IN1, HIGH);
    digitalWrite(RIGHT_IN2, LOW);

    movement = "FORWARD";

}

void moveBackward() {

    digitalWrite(LEFT_IN1, LOW);
    digitalWrite(LEFT_IN2, HIGH);

    digitalWrite(RIGHT_IN1, LOW);
    digitalWrite(RIGHT_IN2, HIGH);

    movement = "BACKWARD";

}

void turnLeft() {

    digitalWrite(LEFT_IN1, LOW);
    digitalWrite(LEFT_IN2, HIGH);

    digitalWrite(RIGHT_IN1, HIGH);
    digitalWrite(RIGHT_IN2, LOW);

    movement = "LEFT";

}

void turnRight() {

    digitalWrite(LEFT_IN1, HIGH);
    digitalWrite(LEFT_IN2, LOW);

    digitalWrite(RIGHT_IN1, LOW);
    digitalWrite(RIGHT_IN2, HIGH);

    movement = "RIGHT";

}

void stopMotors() {

    digitalWrite(LEFT_IN1, LOW);
    digitalWrite(LEFT_IN2, LOW);

    digitalWrite(RIGHT_IN1, LOW);
    digitalWrite(RIGHT_IN2, LOW);

    movement = "STOP";

}

float readDistance() {

    digitalWrite(TRIG_PIN, LOW);

    delayMicroseconds(2);

    digitalWrite(TRIG_PIN, HIGH);

    delayMicroseconds(10);

    digitalWrite(TRIG_PIN, LOW);

    long duration =
        pulseIn(ECHO_PIN, HIGH);

    float distance =
        duration * 0.034 / 2;

    return distance;

}
ESP32 | v n8n WEBHOOK | v JSON DATA | v AI AGENT | +------------------+ | | v v LOW BATTERY OBSTACLE | | v v TELEGRAM ALERT MOTOR STOP | v GOOGLE SHEETS | v THINGSPEAK CLOUD

Example n8n JavaScript Code


const data = $json;

let alert = "NORMAL";
let message = "";

if (data.emergency === true) {

    alert = "EMERGENCY";

    message =
    "Emergency stop activated.";

}

else if (data.distance < 30) {

    alert = "OBSTACLE";

    message =
    "Obstacle detected. Wheelchair stopped.";

}

else if (data.battery_percentage < 20) {

    alert = "LOW_BATTERY";

    message =
    "Battery level is low.";

}

return [

    {

        json: {

            ...data,

            alert: alert,

            message: message,

            timestamp:
            new Date().toISOString()

        }

    }

];
  1. Open Telegram.
  2. Search for BotFather.
  3. Use the command /newbot.
  4. Create a bot name.
  5. Copy the Telegram Bot Token.
  6. Add Telegram credentials to n8n.
  7. Use Telegram Send Message node.
  8. Use Text-to-Speech for voice alerts.
Example Telegram Alert:

🚨 AI SMART WHEELCHAIR ALERT
Status: EMERGENCY STOP
Battery: 62%
Distance: 18 cm
Movement: STOP
Timestamp Battery Current Power Distance Movement Alert
10:00 78% 3.8 A 47 W 145 cm FORWARD NORMAL
10:05 65% 5.1 A 62 W 20 cm STOP OBSTACLE

ThingSpeak can be used to display real-time sensor information.

  • Field 1: Battery Voltage
  • Field 2: Battery Percentage
  • Field 3: Current
  • Field 4: Power Consumption
  • Field 5: Distance
  • Field 6: Temperature
  • Field 7: Movement Status
  • Field 8: Emergency Status

The system calculates electrical power using:

Power = Voltage × Current Example: Voltage = 12.4 V Current = 3.8 A Power = 12.4 × 3.8 Power = 47.12 Watts

Historical data from Google Sheets and ThingSpeak can be analyzed to predict future energy consumption.

HIGH CURRENT + LOW BATTERY + HIGH TEMPERATURE | v HIGH POWER CONSUMPTION | v AI PREDICTION | v TELEGRAM WARNING
The safety system has the highest priority.

Emergency Button → Immediate Motor Stop
Obstacle Detected → Automatic Motor Stop
Wi-Fi Failure → Safe Stop Mode
Low Battery → Warning Alert
Invalid AI Command → No Movement
  • AI Camera-Based Object Detection
  • GPS Tracking
  • Fall Detection
  • Offline Voice Recognition
  • Advanced AI Power Prediction
  • Health Monitoring Sensors
  • Mobile Application
  • Autonomous Navigation
  • Obstacle Avoidance Using AI Vision
  1. Assemble the wheelchair motors.
  2. Connect motor driver to ESP32.
  3. Install ultrasonic and safety sensors.
  4. Install emergency stop button.
  5. Upload ESP32 firmware.
  6. Connect ESP32 to Wi-Fi.
  7. Create n8n webhook.
  8. Configure Telegram Bot.
  9. Connect Google Sheets.
  10. Create ThingSpeak channel.
  11. Configure AI Agent.
  12. Test voice commands.
  13. Test obstacle detection.
  14. Test emergency stop.
  15. Perform complete safety testing.

The AI Smart Voice-Controlled Wheelchair combines artificial intelligence, IoT, ESP32, voice recognition, n8n automation, Telegram notifications, Google Sheets and ThingSpeak Cloud.

The system follows an Agentic IoT architecture:

SENSE ↓ UNDERSTAND ↓ ANALYZE ↓ DECIDE ↓ ACT ↓ NOTIFY ↓ PREDICT
Important Safety Notice:

This project is intended for prototype and educational development. A real human-carrying wheelchair requires high-current electrical protection, mechanical braking, emergency-stop redundancy, battery protection, professional engineering validation and extensive supervised testing.

AI Smart Voice-Controlled Wheelchair Project

ESP32 | AI | IoT | n8n | Telegram | Google Sheets | ThingSpeak

on

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)