Smart Baby Monitor System - Real-time Cry Detection and Health Monitoring

This repository contains the code and documentation for a Baby Monitor System using two Arduino microcontrollers: Arduino Nicla Voice for real-time AI cry detection and Arduino Nano 33 IoT for health monitoring (body temperature, room temperature, and heart rate). The system provides caregivers with real-time information about the baby’s well-being and triggers alerts for different events.

Features

• Real-Time Cry Detection: AI-powered cry detection using Edge Impulse for accurate classification of baby cries.
• Health Monitoring: Monitors baby’s body temperature, room temperature, and heart rate.
• Mobile App Notifications: Sends real-time alerts to the mobile app via Wi-Fi to notify caregivers.
• Non-Invasive Measurement: Uses MLX90614 contactless infrared thermometer for body and room temperature monitoring.
• AI Integration: The cry detection model is deployed on the Arduino Nicla Voice microcontroller, leveraging machine learning models trained with Edge Impulse.

Hardware Setup

The Baby Monitor System consists of two main hardware units:

1. Arduino Nicla Voice:

   • Microcontroller used for AI-based cry detection using audio analysis.
   • It integrates with the Edge Impulse platform to run the trained AI model for detecting baby cries.

2. Arduino Nano 33 IoT:

   • Microcontroller responsible for transmitting data to the mobile app via Wi-Fi.
   • It handles temperature measurements using the MLX90614 infrared thermometer and monitors the heart rate sensor.

Sensors and Modules:

• MLX90614 Contactless Infrared Thermometer: Used for measuring the baby's body temperature and room temperature.
• Heart Rate Sensor: Used to monitor the baby's heart rate.

Software Setup

Required Libraries:

• ArduinoBLE - For Bluetooth Low Energy (BLE) communication.
• Edge Impulse - For training and deploying the AI model for cry detection.
• BLEDevice - For handling BLE communication between devices.

Deployment Options:

1. Deployment using the Edge Impulse Deployment Library (NDP120):

   • This option allows integration of the AI cry detection model into the firmware using the NDP120 library, ensuring real-time cry detection on the Nicla Voice the AI model available in cry_detect-nicla-voice-v12 file.

2. Deployment as a Binary File:

   • The AI model is compiled into a binary file that is uploaded to the Nicla Voice, making the system operate autonomously without needing a connection to the Edge Impulse platform.

Phases of Operation

1. Initialization Phase:

• The system powers up and initializes the Arduino Nicla Voice and Arduino Nano 33 IoT microcontrollers, as well as other sensors like the MLX90614 thermometer and heart rate sensor.

2. Cry Detection Phase:

• The Nicla Voice uses the AI model trained on Edge Impulse to analyze audio in real time and detect the presence of a baby’s cry. The result is transmitted to the Arduino Nano 33 IoT for further processing.

3. Data Transmission Phase:

• The Nano 33 IoT transmits the cry detection status and other health data to the mobile app via Wi-Fi.

4. Body Temperature Measurement Phase:

• The MLX90614 thermometer measures the baby’s body temperature without physical contact and sends the data to the Nano 33 IoT, which then forwards it to the mobile app.

5. Room Temperature Monitoring Phase:

• The MLX90614 also monitors the room temperature and sends alerts if it detects dangerous conditions like high temperatures (potential fire risk).

6. Heart Rate Monitoring Phase:

• The Nano 33 IoT acquires the baby’s heart rate data from the heart rate sensor and sends it to the mobile app for display and analysis.

System Requirements

• Arduino Nicla Voice for AI-based cry detection.
• Arduino Nano 33 IoT for health monitoring and Wi-Fi communication.
• MLX90614 Infrared Thermometer for body and room temperature monitoring.
• Heart Rate Sensor for monitoring the baby's heart rate.
• Mobile App for receiving notifications and alerts (the app should be developed separately).
• Edge Impulse account for training and deploying the AI cry detection model.

How to Use

1. Set up the hardware:

   • Connect the Nicla Voice and Nano 33 IoT according to the wiring diagrams provided in the repository.
   • Install the necessary sensors: MLX90614 for temperature monitoring and a heart rate sensor.

2. Train the AI model:

   • Use Edge Impulse to collect and label cry samples for training.
   • Deploy the trained AI model on the Nicla Voice either using the NDP120 library or by flashing the model as a binary.

3. Upload the Code:

   • Upload the provided code to the Nicla Voice and Nano 33 IoT using the Arduino IDE.
   • Ensure the code for both microcontrollers is correctly uploaded to enable proper communication and functionality.

4. Mobile App Integration:

   • Ensure the mobile app is configured to receive data from the Nano 33 IoT and display real-time alerts for cry detection, body temperature, room temperature, and heart rate.

Troubleshooting

• Problem: No cry detection or false positives:

  • Make sure the Edge Impulse model is properly trained with enough labeled cry data.
  • Ensure the microphone is clear of obstructions to allow accurate audio detection.

• Problem: No Wi-Fi connection:

  • Check the Arduino Nano 33 IoT Wi-Fi connection settings and ensure the network credentials are correctly configured in the code.

• Problem: Temperature measurements not accurate:

  • Verify that the MLX90614 sensor is correctly calibrated and positioned for accurate readings.

Contributing

Feel free to submit pull requests, report issues, or suggest improvements to this project. Contributions are welcome!

License

=======

57c94e3dd9cc55fd52d03ef0f4b5b7faab5f35f1

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