AI-Based Smart Attendance System Using Face Recognition

AI-Based Smart Attendance System Using Face Recognition ESP32 + AI Agentic IoT + n8n Automation + Telegram Voice Alerts + Google Sheets + ThingSpeak Cloud Dashboard

AI-Based Smart Attendance System Using Face Recognition ESP32 + AI Agentic IoT + n8n Automation + Telegram Voice Alerts + Google Sheets + ThingSpeak Cloud Dashboard This project combines: Face Recognition Attendance ESP32 IoT Controller n8n Workflow Automation Telegram Alerts + Voice Notifications Google Sheets Logging ThingSpeak Cloud Dashboard AI-based Power Consumption Prediction Agentic AI Automation Logic 1. Project Overview Objective Build an intelligent attendance system that: Detects and recognizes faces Marks attendance automatically Sends data to cloud Stores records in Google Sheets Sends Telegram notifications and voice alerts Displays analytics on ThingSpeak Predicts power usage using AI logic Uses n8n as an automation brain 2. System Architecture Complete Workflow Camera Detects Face ↓ ESP32-CAM Captures Image ↓ Face Recognition Process ↓ Attendance Verified ↓ ESP32 Sends Data to n8n Webhook ↓ n8n Automation Executes ↓ ├── Google Sheets Entry ├── Telegram Message ├── Telegram Voice Alert ├── ThingSpeak Update └── AI Analytics Processing ↓ Dashboard Monitoring 3. Hardware Components List Component Quantity Purpose ESP32-CAM Module 1 Main controller + camera FTDI Programmer 1 Upload code OLED Display (Optional) 1 Status display Buzzer 1 Audio alert Relay Module (Optional) 1 Door control LED Indicators 2 Status LEDs Push Button 1 Enrollment mode Power Supply 5V 2A 1 Power Jumper Wires Several Connections Breadboard/PCB 1 Circuit setup WiFi Router 1 Internet connection 4. Software Requirements Software Purpose Arduino IDE ESP32 programming n8n Workflow automation Telegram Bot Notifications Google Sheets API Attendance logging ThingSpeak IoT cloud dashboard Python/OpenCV Face training Edge Impulse (Optional) AI model deployment 5. ESP32-CAM Pin Configuration ESP32-CAM Important Pins Pin Function GPIO0 Flash mode GPIO2 LED GPIO12 Camera GPIO13 Camera GPIO14 Camera GPIO15 Camera GPIO16 UART GPIO4 Flash LED 6. Circuit Schematic Diagram Basic Wiring ESP32-CAM -------------------------------- 5V → Power Supply 5V GND → Ground U0R → FTDI TX U0T → FTDI RX GPIO0 → GND (while uploading) Buzzer: GPIO15 → Buzzer + LED: GPIO2 → LED + Relay: GPIO14 → Relay IN 7. Face Recognition System Face Recognition Methods Option 1 — ESP32 Built-in Face Recognition Good for: Small attendance systems 5–20 users Option 2 — Python OpenCV Server Good for: Large databases Better accuracy Recommended: Use ESP32 for image capture Use Python/OpenCV for recognition 8. Face Enrollment Process Steps User presses enrollment button ESP32 captures multiple images Images stored in server/database AI model trains face embeddings Face ID assigned 9. Attendance Logic Workflow Face Detected? ↓ YES Face Recognized? ↓ YES Already Marked Today? ↓ NO Store Attendance Send Notification Update Cloud 10. ESP32 Source Code Arduino IDE Setup Install: ESP32 Board Package Camera libraries WiFi libraries ESP32 Attendance Code #include #include const char* ssid = "YOUR_WIFI"; const char* password = "YOUR_PASSWORD"; String webhookURL = "https://your-n8n-url/webhook/attendance"; void setup() { Serial.begin(115200); WiFi.begin(ssid, password); while (WiFi.status() != WL_CONNECTED) { delay(1000); Serial.println("Connecting..."); } Serial.println("WiFi Connected"); } void loop() { // Simulated recognized face String personName = "Rahul"; String timeStamp = "10:30 AM"; if(WiFi.status()== WL_CONNECTED){ HTTPClient http; http.begin(webhookURL); http.addHeader("Content-Type", "application/json"); String jsonData = "{\"name\":\"" + personName + "\",\"time\":\"" + timeStamp + "\"}"; int httpResponseCode = http.POST(jsonData); Serial.println(httpResponseCode); http.end(); } delay(10000); } 11. n8n Automation Workflow What n8n Does n8n acts as the AI automation brain. It receives attendance data and performs: Google Sheets update Telegram alert Voice notification ThingSpeak update AI prediction logic 12. n8n Workflow Architecture Webhook Trigger ↓ Data Validation ↓ Google Sheets Node ↓ Telegram Node ↓ Text-to-Speech ↓ ThingSpeak API ↓ AI Prediction Function 13. Install n8n Local Installation Using Docker: docker run -it --rm \ -p 5678:5678 \ n8nio/n8n Official Website n8n 14. Create Webhook in n8n Steps Open n8n Create new workflow Add Webhook node Method → POST Path → /attendance Copy webhook URL 15. Google Sheets Integration Create Google Sheet Columns: Name Date Time Status Setup Steps Open Google Cloud Console Enable Sheets API Create Service Account Download JSON credentials Connect credentials to n8n Official APIs Google Sheets API 16. Telegram Bot Setup Create Telegram Bot Open Telegram Search for BotFather Run: /newbot Copy API token Telegram Official Telegram Bot API 17. Telegram Alert Message Example Message ✅ Attendance Marked Name: Rahul Time: 10:30 AM Status: Present 18. Voice Notification Automation Method n8n → Google TTS → Telegram Voice Voice Message Example "Rahul attendance marked successfully." 19. ThingSpeak Dashboard Setup Create ThingSpeak Channel Fields: Field Purpose Field1 Attendance Count Field2 Power Usage Field3 Recognized Faces Field4 WiFi Strength Official Website ThingSpeak 20. Sending Data to ThingSpeak HTTP Request String url = "http://api.thingspeak.com/update?api_key=YOUR_KEY&field1=1"; 21. AI Power Consumption Prediction Goal Predict system power usage using AI logic. Parameters Parameter Description Camera usage time Active duration WiFi transmission Network activity CPU load Processing usage Flash LED usage LED activity Simple Prediction Formula P total ​ =P camera ​ +P wifi ​ +P cpu ​ +P led ​ AI Logic Example predicted_power = (camera_time * 0.5) + (wifi_packets * 0.2) + (cpu_usage * 0.1) 22. AI Agentic Features Smart AI Behaviors AI Agent Can: Detect duplicate attendance Predict abnormal activity Notify low power state Detect unauthorized access Recommend energy optimization Generate daily reports 23. Advanced Attendance Validation Anti-Spoofing Features Use: Eye blink detection Face movement analysis IR sensor validation Multi-frame recognition 24. Database Design Attendance Table ID Name Date Time Confidence 25. Security Features Recommended Security HTTPS webhook Token authentication Face encryption Local backup API rate limiting 26. n8n Workflow JSON Example { "nodes": [ { "name": "Webhook", "type": "n8n-nodes-base.webhook" }, { "name": "Google Sheets", "type": "n8n-nodes-base.googleSheets" }, { "name": "Telegram", "type": "n8n-nodes-base.telegram" } ] } 27. Deployment Guide Local Deployment Good for: College projects Labs Small offices Cloud Deployment Use: AWS Railway Render VPS Docker 28. Production Architecture ESP32 Devices ↓ Cloud API Gateway ↓ n8n Server ↓ Database Cluster ↓ AI Analytics Engine 29. Future Enhancements Advanced Features AI Features Emotion detection Mask detection Crowd analytics Face aging adaptation AI attendance prediction IoT Features RFID backup Fingerprint backup Smart lock integration Battery monitoring Offline synchronization Cloud Features Mobile app Admin dashboard Multi-school support Analytics reports 30. Testing Procedure Step-by-Step Testing Hardware Test Power ESP32 Verify camera Test WiFi API Test Trigger webhook Verify Google Sheets update Check Telegram alert AI Test Train face model Test recognition accuracy 31. Troubleshooting Guide Problem Solution Camera not detected Check power supply WiFi disconnects Improve signal Face mismatch Retrain model Telegram not sending Verify bot token Sheets update fails Check API permissions 32. Recommended Folder Structure project/ │ ├── esp32_code/ ├── face_dataset/ ├── python_ai/ ├── n8n_workflow/ ├── dashboard/ ├── docs/ └── deployment/ 33. Recommended Technology Stack Layer Technology Hardware ESP32-CAM AI Vision OpenCV Automation n8n Notifications Telegram Database Google Sheets Cloud IoT ThingSpeak Backend Flask/FastAPI 34. Complete End-to-End Workflow Person Arrives ↓ Face Captured ↓ AI Recognition ↓ Attendance Verification ↓ n8n Webhook Trigger ↓ Google Sheets Updated ↓ Telegram Alert Sent ↓ Voice Notification Sent ↓ ThingSpeak Dashboard Updated ↓ AI Analytics Generated 35. Suggested Enhancements for Final Year Projects Add These for Higher Innovation Edge AI inference Real-time analytics dashboard MQTT communication Firebase integration AI chatbot assistant Voice-controlled admin system Generative AI attendance summaries 36. Recommended Learning Resources ESP32 ESP32 Official Documentation Arduino IDE Arduino IDE OpenCV OpenCV ThingSpeak ThingSpeak Documentation 37. Final Output of System The completed system will provide: ✅ AI face recognition attendance ✅ Real-time cloud monitoring ✅ Telegram alerts ✅ Voice notifications ✅ Google Sheets logs ✅ ThingSpeak analytics ✅ AI-based energy prediction ✅ Fully automated IoT workflow ✅ Smart attendance intelligence ✅ Scalable enterprise architecture 38. Conclusion This project combines: Artificial Intelligence IoT Automation Edge Computing Cloud Analytics Workflow Automation Smart Notifications into a modern smart campus/office solution suitable for: Final year projects Research projects Smart classrooms Offices Industrial attendance systems AIoT demonstrations