FallWatcher

FallWatcher is an academic project that encompasses a complete IoT-based system for real-time fall detection. The solution was developed using ESP8266 devices, an MQTT broker, and a Flutter mobile application. This system integrates various components to achieve efficient communication and instant notifications. I was responsible for orchestrating the development and integration of all components, while the rest of the team constructed the physical device using IMU sensors and a 3D-printed casing. The system is composed of the following components:

1. ESP8266 Fall Detection Device: Firmware running on ESP8266 devices that detects falls and registers with the MQTT broker.
2. MQTT Broker (Node.js): A broker built using Aedes and Node.js, which manages the communication between devices and the Flutter app. It also generates a QR code for each device's MAC address when they register.
3. Web Server (Express): Serves the generated QR code that contains the device's MAC address so it can be scanned and registered in the Flutter app.
4. Flutter Mobile Application: A mobile app that allows users to receive fall detection notifications, manage the devices, and confirm their status.

System Components

1. ESP8266 Fall Detection Device

• The ESP8266 devices are responsible for detecting falls using sensors like an accelerometer.
• When a device detects a fall, it sends a message to the MQTT broker.
• When a device registers with the broker, the MAC address is used to generate a QR code, which can be scanned by the mobile app.
• The firmware is located in the esp8266/ directory.

2. MQTT Broker (Node.js)

• Built using Aedes, a lightweight MQTT broker, the MQTT broker handles all messages between the ESP8266 devices and the Flutter app.
• It listens for device registration on the topic devices/register.
• Upon receiving a registration message with the device’s MAC address, the broker generates a QR code that represents the MAC address and stores it.
• The broker runs on port 1883.

3. Express Web Server (Node.js)

• The Express server is responsible for serving the QR codes generated by the MQTT broker.
• It exposes a route /qrcode/:mac, where :mac is the MAC address of the device.
• The mobile app scans this QR code to register the device and start receiving fall detection notifications.
• The server is configured to run on port 80 by default.

4. Flutter Mobile Application

• The Flutter app displays the list of devices, shows notifications when a fall is detected, and allows the user to confirm that the device is OK after a fall.
• Devices are added by scanning the QR code generated by the MQTT broker.
• The app subscribes to MQTT topics to receive real-time updates about the status of each device.

System Architecture

The following diagram illustrates how the system components interact with each other:

graph LR
    A[ESP8266 Device] -->|Register via MAC| B(MQTT Broker)
    B -->|Generate QR Code| C[Express Web Server]
    C -->|Serve QR to Mobile App| D[Flutter App]
    A -->|Send Fall Detection| B
    B -->|Send Status Updates| D
    D -->|Confirm Device OK| B

Loading

System Flow

1. Device Registration: When a new device (ESP8266) powers up, it sends its MAC address to the MQTT broker using the topic devices/register.
2. QR Code Generation: The broker generates a QR code using the MAC address and stores it.
3. QR Code Scanning: The Express server serves the QR code, and the user can scan it with the Flutter app to register the device.
4. Fall Detection: When a device detects a fall, it sends a message to the broker, which forwards the information to the Flutter app. The app displays a notification to the user.
5. Acknowledge Status: The user can confirm the status of the device in the app, and the confirmation is sent back to the MQTT broker.

Setup Instructions

1. ESP8266 Fall Detection Device

1. Navigate to the esp8266/ folder to access the device firmware.
2. Flash the firmware to the ESP8266 devices.
3. Configure the Wi-Fi credentials and the MQTT broker IP in the firmware.
4. The device will send its MAC address to the broker when it starts.

2. MQTT Broker Setup

1. Navigate to the mqtt-server/ directory.

2. Install the dependencies:

   npm install

3. Start the MQTT broker:

   node index.js

4. The MQTT broker will now run on port 1883 and listen for device registrations and fall detection messages.

3. Web Server Setup

1. The web server is also part of the mqtt-server/ folder.
2. It will start automatically when you run the node index.js command.
3. The server will be available at http://localhost:80 and serve QR codes at the route /qrcode/:mac.

4. Flutter Mobile Application Setup

1. Navigate to the fallwatcherapp/ folder.

2. Install the Flutter dependencies:

   flutter pub get

3. Configure the MQTT broker IP in mqtt_service.dart:

   static final MqttServerClient client = MqttServerClient('', '');

   Replace with the actual IP address of your broker.

4. Run the app:

   flutter run

MQTT Configuration

The ESP8266 devices register with the broker using the topic:

devices/register

The broker listens on port 1883 for incoming messages. When a fall is detected, the device sends a message to the topic:

nodemcu/{device_mac}/fall

The mobile app subscribes to this topic to receive real-time updates and display fall detection notifications.

Flutter App Pattern

The Flutter app uses the Bloc/Cubit pattern for managing the app’s state. Here is a breakdown of the architecture:

• Cubit: The Cubits handle the state of the device list, device status updates, and fall detection notifications.
• BlocBuilder: The UI rebuilds based on the state changes managed by the Cubits.
• MQTT Integration: The app uses MQTT to receive messages about device statuses and fall detections in real time.

Features:

• Device Management: Devices are added by scanning a QR code with the device’s MAC address.
• Notifications: The app notifies the user when a fall is detected by any device.
• Acknowledge Fall: The user can confirm that the device is OK after a fall is detected.

License

This project is licensed under the GPL v3 License. See the LICENSE file for more details.