AI Smart Automatic Exam Paper Evaluation System
ESP32 + AI Agent + n8n + Telegram Voice Alerts + Google Sheets + ThingSpeak
1. Project Overview
This project automatically evaluates student answer sheets using AI, stores marks in Google Sheets, updates ThingSpeak dashboards, and sends Telegram voice notifications.
2. System Architecture
Student Answer Sheet
|
V
ESP32-CAM
|
V
WiFi
|
V
n8n Server
|
-----------------
| | |
OCR AI Database
| |
---------
|
V
Google Sheets
|
V
ThingSpeak
|
V
Telegram Voice Alert
3. Components List
| Component | Quantity |
|---|---|
| ESP32-CAM | 1 |
| ESP32 Development Board | 1 |
| OV2640 Camera | 1 |
| WiFi Router | 1 |
| USB TTL Converter | 1 |
| Breadboard | 1 |
| Jumper Wires | As Required |
| Power Supply | 5V 2A |
4. Flowchart
START | Initialize ESP32 | Connect WiFi | Capture Image | Upload to n8n | OCR Processing | AI Evaluation | Generate Marks | Store Results | Send Telegram Alert | END
5. ESP32 Source Code
#include <WiFi.h>
#include <HTTPClient.h>
const char* ssid = "YOUR_WIFI";
const char* password = "YOUR_PASSWORD";
void setup()
{
Serial.begin(115200);
WiFi.begin(ssid,password);
while(WiFi.status()!=WL_CONNECTED)
{
delay(500);
}
}
void loop()
{
HTTPClient http;
http.begin(
"https://yourserver.com/webhook/exam");
http.addHeader(
"Content-Type",
"application/json");
String data =
"{\"student\":\"101\"}";
http.POST(data);
http.end();
delay(30000);
}
6. n8n Workflow
Webhook | OCR | AI Evaluation | Score Calculation | --------------------- | | | Google ThingSpeak Telegram Sheets Voice
7. Google Sheets Structure
| Student ID | Subject | Marks | Similarity | Feedback |
|---|---|---|---|---|
| 101 | Maths | 85 | 90% | Good Performance |
8. ThingSpeak Dashboard Fields
| Field | Description |
|---|---|
| Field1 | Marks |
| Field2 | Similarity |
| Field3 | Evaluation Time |
| Field4 | AI Confidence |
| Field5 | Power Consumption |
9. Telegram Voice Alert
AI Result
|
Text Message
|
Google TTS
|
MP3 Audio
|
Telegram Send Audio
Example Voice: "Student 101 scored 85 marks. Performance is excellent."
10. AI Power Consumption Prediction
Formula:
Power = Voltage x Current Example: 5V x 0.24A Power = 1.2 Watts
11. Database Table
CREATE TABLE students( id INT, name VARCHAR(50), subject VARCHAR(30), marks FLOAT, similarity FLOAT, feedback TEXT, timestamp DATETIME );
12. Future Enhancements
- Handwritten OCR
- Face Recognition
- AI Proctoring
- WhatsApp Alerts
- Mobile Application
- TinyML Deployment
- Offline Evaluation
13. Deployment Architecture
ESP32-CAM
|
Internet
|
Cloud Server
|
n8n Automation
|
OpenAI / Gemini
|
Google Sheets
|
ThingSpeak
|
Telegram Alerts
