AI Smart Multi-Language Voice Translation Device Using Raspberry Pi
Project Overview
This project is an AI powered IoT voice translation system.
It uses:
- Raspberry Pi - AI voice processing
- ESP32 - IoT controller
- n8n - AI automation workflow
- Telegram Bot - Voice alerts
- Google Sheets - Cloud logging
- ThingSpeak - IoT dashboard
System Architecture
User Voice
|
|
Raspberry Pi AI Engine
|
Speech Recognition
|
Translation Engine
|
Text To Speech
|
ESP32 IoT Controller
|
n8n Automation
|
-------------------------
Telegram
Google Sheets
ThingSpeak
Features
- Real-time voice translation
- Multi language support
- AI speech recognition
- Voice output
- ESP32 monitoring
- Telegram voice alerts
- Cloud dashboard
- Power prediction
Hardware Components
| Component | Purpose |
|---|---|
| Raspberry Pi 4/5 | AI processing |
| ESP32 | IoT control |
| Microphone | Voice input |
| Speaker | Audio output |
| INA219 | Battery monitoring |
| DHT11 | Temperature |
| OLED Display | Status display |
ESP32 Circuit Connection
INA219 VCC -> ESP32 3.3V GND -> GND SDA -> GPIO21 SCL -> GPIO22 DHT11 VCC -> 3.3V DATA -> GPIO4 GND -> GND
ESP32 Source Code
#include WiFi.h
#include HTTPClient.h
#include DHT.h
#define DHTPIN 4
void setup()
{
Serial.begin(115200);
WiFi.begin(
"SSID",
"PASSWORD"
);
}
void loop()
{
float temp = dht.readTemperature();
HTTPClient http;
http.begin(
"http://server.com/data"
);
http.GET();
http.end();
delay(10000);
}
Raspberry Pi AI Translation Python
import speech_recognition as sr from googletrans import Translator import pyttsx3 translator = Translator() engine = pyttsx3.init() recognizer = sr.Recognizer() while True: with sr.Microphone() as source: audio = recognizer.listen(source) text = recognizer.recognize_google(audio) result = translator.translate( text, dest='hi' ) engine.say(result.text) engine.runAndWait()
n8n Automation Flow
ESP32 Data
|
Webhook
|
AI Agent
|
Power Prediction
|
---------------------
Telegram Alert
Google Sheet
ThingSpeak
n8n Workflow JSON
{
"nodes":[
{
"name":"ESP32 Webhook",
"type":"webhook"
},
{
"name":"AI Prediction",
"type":"function"
},
{
"name":"Telegram Alert",
"type":"telegram"
},
{
"name":"Google Sheet",
"type":"googleSheets"
}
]
}
Telegram Bot Setup
- Open Telegram
- Search BotFather
- Create new bot
- Copy API Token
- Add token into n8n Telegram node
Google Sheets Integration
Columns: Time Temperature Battery Language Translation
ThingSpeak Dashboard
- Temperature Graph
- Battery Status
- Translation Count
- Power Prediction
AI Power Prediction
Remaining Time = Battery Capacity / Average Power Usage Example: 5000mAh Battery 500mA usage Result: 10 Hours Remaining
Voice Notification Automation
IF Battery < 20% Send Telegram Voice Alert IF Temperature > 40°C Send Warning Alert
Deployment Steps
- Install Raspberry Pi OS
- Install AI libraries
- Connect ESP32
- Upload firmware
- Setup n8n
- Create Telegram Bot
- Connect Cloud Dashboard
- Test Translation
Future Enhancements
- GPT based translation
- Offline AI model
- Camera translation
- Wearable version
- Solar charging
- Mobile application
Final Result
AI Translation Assistant + Agentic IoT Automation System
Using Raspberry Pi + ESP32 + n8n + Telegram + Cloud
