🚗 AI Road Condition Detection & SOS Alert System

An intelligent road monitoring system that uses AI, computer vision, and GPS to detect road anomalies such as potholes and speed breakers while automatically logging their locations to a backend server. The system also includes an SOS emergency alert feature that sends notifications to a Telegram bot with the vehicle’s location. Team - 11 - Suriyan Loganathan, Udhayadev G S, Dhakshnamoorthy J

Sample image for dashboard -

Sample image for maps - Telegram in emergency contacts phone when there is an SOS alert -

---

📌 Project Overview

This project uses Flutter, TensorFlow Lite, and a Node.js + MongoDB backend to build a real-time road monitoring system. The mobile app captures images, detects road conditions using an AI model, and sends the results with GPS coordinates to the backend. The server stores detections and visualizes them on a web dashboard and Google Maps.

An SOS emergency button allows users to send their location instantly, triggering a Telegram bot alert.

---

⚙️ System Architecture

Flutter App ↓ Camera + TensorFlow Lite ↓ GPS Location (Geolocator) ↓ Node.js API ↓ MongoDB Database ↓ Web Dashboard + Google Maps ↓ Telegram SOS Alerts

---

🚀 Features

📷 AI Road Detection

• Detects Normal Road
• Detects Potholes
• Detects Speed Breakers
• Runs using TensorFlow Lite

📍 GPS Tracking

• Captures latitude, longitude, timestamp
• Sends location with every detection

🗄 Backend Logging

• Stores detections in MongoDB
• Displays logs in a web dashboard

🗺 Map Visualization

• Google Maps showing detected road issues

• Color markers:

  • 🔴 Pothole
  • 🟠 Speed Breaker
  • 🟢 Normal Road

🚨 SOS Emergency Alert

• SOS button in mobile app
• Sends GPS location to backend
• Backend triggers Telegram bot notification

Example alert:

🚨 SOS ALERT Vehicle TN07-XYX-123 sent an emergency alert

Latitude: 13.0827 Longitude: 80.2707

https://maps.google.com/?q=13.0827,80.2707

---

🧠 Technologies Used

Mobile App

• Flutter
• Dart
• Camera
• Geolocator
• TensorFlow Lite

Backend

• Node.js
• Express
• MongoDB
• Mongoose
• Axios

Visualization

• HTML / CSS
• Google Maps API

Alerts

• Telegram Bot API

---

📂 Project Structure

road_detector_project
│
├── flutter_app
│   ├── main.dart
│   └── model.tflite
│
├── road_backend
│   ├── server.js
│   └── package.json
│
└── README.md

---

🛠 Installation

Clone repository

git clone https://github.com/yourusername/road-detector-ai.git

Flutter App

flutter pub get
flutter run

Required permissions:

• Camera
• Location
• Internet

Backend

cd road_backend
npm install
node server.js

Server runs at:

http://localhost:5000

---

🌍 Web Dashboard

Detection logs:

http://localhost:5000

Map visualization:

http://localhost:5000/map

---

📡 API Endpoints

Save Detection

POST /api/detection

Send SOS

POST /api/sos

---

👨‍💻 Author

Suriyan L