Edge AI Home Automation

A smart home system with customizable controls for lights, switches, temperature, and doors, featuring secure login, automatic database backup and error detection, and LSTM-based predictions every 10 minutes. Supports integrations with surveillance, voice-gesture control, and energy optimization systems for enhanced automation

Features

• Predictive Analytics: Implements an LSTM-based model for accurate predictions every 10 minutes, optimizing system efficiency.
• Intuitive User Interface: Provides seamless control over lights, switches, temperature, and doors, ensuring ease of use.
• Customizable Modules: Allows personalized home automation configurations for tailored user experiences.
• Secure Authentication: Enforces password-protected login to safeguard user access.
• Database Management: Ensures automatic backups and error detection to maintain data integrity.
• Smart System Integrations: Supports:
  • Secure Access Surveillance System for enhanced security monitoring.
  • Voice Gesture Based Control System for hands-free smart home interaction.
  • Energy Consumption Optimization System for efficient energy management.

Screenshots

Installation

• Run the following command to clone the required repositories:

  git clone https://github.com/Smart-Home-Automation-System-Project/Edge-AI-Home-Automation

For Ubuntu:

• Install Mosquitto and client tools:

  sudo apt install -y mosquitto mosquitto-clients

• Enable and start the Mosquitto service:

  sudo systemctl enable mosquitto
  sudo systemctl start mosquitto

• Navigate to src:

  cd Edge-AI-Home-Automation/src/

• Generate a password file for your system. Default username and password are admin and 1234. You can generate a new password file:

  sudo mosquitto_passwd -c /config/passwordfile.txt [your_username]

  Then enter a secure password when prompted.
• Install dependencies

  pip install -r requirements.txt

• Start system
  • Make executable:

    chmod +x start-system.sh
    chmod +x backup-db.sh
    chmod +x start-mqtt.sh

  • Modify 'config/.env' and update IPs, JWT_SECRET_KEY, username and passwords:

    MQTT_USERNAME=admin
    MQTT_PASSWORD=1234
    MQTT_PORT=1883
    JWT_SECRET_KEY=MTI3LjAuMC4xIC0gLSBbMjAvQXByLzIwMjUgMTA6NTU6MDFdICJHRVQgL2F
    UI_PASSWORD=abcd
    IP_SAS_SYSTEM=[IP OF Secure-Access-Surveillance-System]
    IP_ECO_SYSTEM=[IP OF Energy-Consumption-Optimization]

  • Run the system:

    ./run.sh

• Setup auto-start
  • Create a systemd service

    sudo nano /etc/systemd/system/ai-home.service

  • Paste the following:

    [Unit]
    Description=AI Home Automation Startup Script
    After=network.target
    
    [Service]
    ExecStart=/bin/bash [/path/to/start-system.sh]
    WorkingDirectory=[/path/to/]
    Restart=always
    User=[your_username]
    
    [Install]
    WantedBy=multi-user.target

• Setup auto-backup and auto-train-model
  1. Open the crontab editor:

     crontab -e

  2. Add these two lines at the bottom of the file:

     # Backup DB daily at 4:00 AM
     0 4 * * * [/full/path/to]/backup-db.sh
     
     # Train model every Monday at 12:00 AM
     0 0 * * 1 /usr/bin/python3 [/full/path/to]/database/train.py

For Windows:

• Install Mosquitto and client tools: https://mosquitto.org/download
• Add install directory to PATH

  setx PATH "%PATH%;[C:\path\to\install]"

• Navigate to src:

  cd Edge-AI-Home-Automation/src/

• Generate a password file for your system. Default username and password are admin and 1234. You can generate a new password file:

  mosquitto_passwd -c /config/passwordfile.txt [your_username]

  Then enter a secure password when prompted.
• Install dependencies

  pip install -r requirements.txt

• Start system
  • Modify 'config/.env' and update IPs, JWT_SECRET_KEY, username and passwords:

    MQTT_USERNAME=admin
    MQTT_PASSWORD=1234
    MQTT_PORT=1883
    JWT_SECRET_KEY=MTI3LjAuMC4xIC0gLSBbMjAvQXByLzIwMjUgMTA6NTU6MDFdICJHRVQgL2F
    UI_PASSWORD=abcd
    IP_SAS_SYSTEM=[IP OF Secure-Access-Surveillance-System]
    IP_ECO_SYSTEM=[IP OF Energy-Consumption-Optimization]

  • Run the system:

    run.bat

Authors

• @Sahan Gunasekara
• @Malaka Gunawardana
• @Himala Gunathilaka
• @Isurika Gunasekara
• @Nesandu Sithnuka