WattBot is a full-stack web application that lets you upload images of electricity meters and automatically extract readings using a custom-trained YOLO model.
The project includes a Python FastAPI backend and a React frontend with support for dark mode, currency selection, and price tracking.

• 📸 Upload electricity meter images
• 🔍 Detect digital readings using a trained YOLO model
• 🌙 Dark mode support
• 💰 Track electricity prices and consumption history
• 🛠 Built with FastAPI (Python) and React + Material UI
• 🐳 Easy deployment via Docker

• Frontend: React, Material UI, TypeScript
• Backend: FastAPI, Python 3.13, Uvicorn, Pydantic
• Machine Learning: YOLOv11s-obb (custom-trained on real electricity meter data)
• Database: MongoDB
• Containerization: Docker, Docker Compose
• CI/CD: GitHub Actions

• Docker & Docker Compose

1. Create a new folder and save the contents of docker-compose.yml from this repo into a file named docker-compose.yml.

2. Open a terminal in that folder and run:

docker compose pull
docker compose up -d

3. Navigate to the app:

• Frontend UI: http://localhost:80
• Backend API docs: http://localhost:8000/docs

If you're deploying on a remote server, replace localhost with your server's IP (e.g., http://123.123.123.123:5173).

The YOLO model was trained on a limited dataset of real electricity meters. If your meter's layout or font style is significantly different, the reading might be misinterpreted.
If you encounter inaccuracies and want to improve precision for your meter type, feel free to open an issue contributions are welcome!

Make sure you have the following installed on your machine:

• Python 3.11+
• Node.js 18+
• Docker & Docker Compose

Runs on http://localhost:8000

cd backend
pip install -r requirements.txt
uvicorn main:app --reload

Swagger UI available at: http://localhost:8000/docs

Runs on http://localhost:5173

cd frontend
npm install
npm run dev

FastAPI automatically generates interactive API docs:

• Swagger UI – Try out endpoints directly from the browser
• ReDoc – Clean reference-style documentation

This project is licensed under the MIT License – see the LICENSE file for details.