🔐 IoT-Lock Backend

This is the backend service for the IoT-Lock system, acting as the middle layer between:

• Raspberry Pi (IoT device) → captures faces / events
• Mobile App (React Native) → receives notifications, manages access
• Backend (FastAPI + PostgreSQL) → handles APIs, face recognition, notifications, and database

---

📂 Project Structure

IOT-LOCK-BACKEND/
│── .gitignore
│── requirements.txt
│── README.md
│── alembic.ini             # (if using migrations)
│
├── app/
│   ├── main.py             # FastAPI entrypoint (server)
│   │
│   ├── api/                # REST API endpoints
│   ├── core/               # configs, logger, security
│   ├── db/                 # database models + session
│   ├── ml/                 # ML models (face detection)
│   │   ├── face_recog.py   # prediction functions
│   │   ├── train_faces.py  # training script
│   │   └── data/           # trained models (face_trained.yml, people.npy)
│   ├── notifications/      # push notification logic
│   └── schemas/            # Pydantic request/response models
│
├── migrations/             # Alembic migration scripts
├── tests/                  # Unit tests
└── venv/                   # Local virtual environment (not pushed to git)

---

⚙️ Setup Instructions

1. Clone the repository

git clone https://github.com/Eahtasham/iot-lock-backend.git
cd iot-lock-backend

2. Create a virtual environment

python -m venv venv

Activate it:

• Windows (PowerShell)

  .\venv\Scripts\activate

• Linux/Mac

  source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Run the server

uvicorn app.main:app --reload

👉 Server will start at: http://127.0.0.1:8000

---

📸 Face Recognition Model

• Training script: app/ml/train_faces.py
• Trained data: stored in app/ml/data/face_trained.yml and people.npy
• Prediction: handled by app/ml/face_recog.py

To train model:

python app/ml/train_faces.py

---

🗄 Database (Postgres on Neon)

• ORM: SQLAlchemy + Alembic

• Connection settings are inside app/core/config.py

• Run migrations:

  alembic upgrade head

---

📲 Notifications

• Located in app/notifications/
• Will use Firebase Cloud Messaging (FCM) to push alerts to the mobile app

---

🧪 Testing

Run unit tests with:

pytest

---

🚀 Roadmap

• REST API endpoints for device, auth, visits, notifications
• Face detection pipeline integration
• Push notification triggers
• DB migrations with Alembic
• Deployment (Docker + server)

---

✅ With this setup, any team member can:

1. Clone repo
2. Setup venv + install requirements
3. Train ML model (optional)
4. Run FastAPI with uvicorn
5. Start contributing 🚀