Smart Wearable Fall Detection and Caregiver Monitoring System Using Arduino Nano 33 IoT

Falls can be dangerous for elderly people or patients living alone, especially when immediate help is not available. In this project, I developed a smart wearable fall detection and caregiver monitoring system using Arduino Nano 33 IoT, Raspberry Pi, MQTT, Node-RED, and Telegram alerts.

The wearable device continuously monitors movement using the built-in IMU sensor in the Arduino Nano 33 IoT. If a possible fall is detected, the system activates a buzzer alarm, sends an emergency alert through MQTT, updates a caregiver monitoring dashboard, and can also trigger Telegram notifications.

This project was designed as a practical healthcare monitoring prototype that combines embedded systems, IoT communication, and real-time monitoring.

Hardware

1. Arduino Nano 33 IoT – wearable device with built-in IMU sensor
2. Raspberry Pi – central caregiver monitoring system
3. Buzzer – for local emergency alerts
4. Slider Switch – manual reset to cancel false alarms
5. Jumper Wires – for connecting components
6. Breadboard – prototype assembly
7. Portable Power Bank / USB power source – for wearable portability
8. Compact Enclosure (optional) – to house the wearable device

Software

1. Arduino IDE – to program the wearable device
2. Node-RED – for caregiver and patient dashboards
3. MQTT Broker – for lightweight IoT messaging
4. Telegram Account (optional) – for remote caregiver notifications
5. Required Arduino libraries:
6. Arduino_LSM6DS3
7. WiFiNINA
8. PubSubClient

This project is a smart wearable healthcare monitoring system designed to improve patient safety, especially for elderly people or individuals who may require emergency assistance.

The main purpose of the system is to automatically detect possible fall events and immediately notify caregivers in real time.

The system works using two connected parts:

Wearable Patient Device:

The wearable device is built using an Arduino Nano 33 IoT. It uses the built-in IMU sensor (accelerometer + gyroscope) to continuously monitor movement and detect abnormal impact patterns that may indicate a fall. If a fall is confirmed, the wearable activates a local buzzer alarm and sends an emergency alert.

Caregiver Monitoring System:

A Raspberry Pi acts as the monitoring station. It receives emergency messages through MQTT communication and updates a Node-RED caregiver dashboard in real time. The system also includes a patient assistance dashboard where manual help requests such as Feeling Unwell or Need Help can be sent. Optional Telegram alerts can also notify caregivers remotely.

System Workflow

The basic workflow is:

1. Arduino wearable continuously monitors movement
2. Sudden impact is detected
3. System waits briefly to confirm if movement stops
4. If confirmed as a fall:
5. buzzer alarm activates
6. MQTT emergency message is sent
7. Raspberry Pi receives alert
8. caregiver dashboard updates
9. Telegram notification can be triggered
10. Patient can manually reset the alarm if it was a false trigger

Purpose of each component

Arduino Nano 33 IoT:

Reads movement data from the built-in IMU sensor and runs the fall detection logic.

Buzzer:

Activates immediately when a fall is confirmed so the patient gets an instant local alert.

Slider Switch:

Allows the patient to manually reset the alarm if the fall detection was a false trigger.

Power Bank:

Makes the system portable instead of requiring constant USB connection to a laptop.

Hardware Assembly

Assemble the Arduino wearable prototype as shown in the image.

Make sure:

1. buzzer is connected to the configured digital output pin
2. slider switch is connected to the reset input pin
3. Arduino receives stable power
4. Raspberry Pi is connected to the same Wi-Fi network

In this step, we will program the Arduino Nano 33 IoT, which acts as the wearable patient monitoring device.

1. Install Arduino IDE on your computer.
2. Connect the Arduino Nano 33 IoT to your computer using USB.
3. Go to Library Manager and install Arduino_LSM6DS3, WiFiNINA and PubSubClient
4. A full video walkthrough of the Arduino code logic is provided. (https://youtu.be/GwdYINLCmjU)
5. After completing the code click on Verify/Upload. Look out for error message if any fix in and again upload the code.
6. Try moving the Arduino Nano 33 IoT and observe the readings in the Serial Monitor.
7. The use the Serial Monitor for testing and debugging to verify sensor readings, Wi-Fi connection, MQTT communication, and fall detection behaviour.

Once the wearable Arduino device was working correctly, I moved to setting up the caregiver monitoring system using a Raspberry Pi.

1. Power on the Raspberry Pi and make sure it is connected to the same Wi-Fi network as the Arduino Nano 33 IoT.
2. Open the terminal on the Raspberry Pi and start Node-RED using the command: node-red-start
3. Wait for Node-RED to fully start, then open the provided web address in a browser to access the Node-RED editor.
4. Install the required Node-RED dashboard package by opening the menu, selecting Manage Palette, then searching for and installing node-red-dashboard.
5. Add an MQTT input node and configure it with the same MQTT broker details and topic used in the Arduino code.
6. Add a function node or switch node to process incoming messages such as SAFE and FALL_DETECTED.
7. Add dashboard text or status display nodes to visually show the patient’s current condition on the caregiver dashboard.
8. Connect the nodes as shown in the dashboard flow image and configure each node accordingly.
9. Click Deploy to activate the caregiver monitoring dashboard.
10. Test the dashboard by triggering a fall event from the Arduino wearable device and checking whether the patient status updates in real time.

1. Open Node-RED on the locally on your desktop.
2. Create a new flow for the patient assistance dashboard.
3. Add dashboard button nodes for manual emergency requests such as Feeling Unwell and Need Help.
4. Configure these buttons so that when pressed, they send emergency messages through MQTT communication.
5. Create a Telegram bot using BotFather and keep the bot token and chat ID ready.
6. Install the required Telegram notification nodes in Node-RED if needed.
7. Add the Telegram notification node and configure it using your bot token and chat ID.
8. Connect the dashboard buttons to the Telegram notification flow so emergency requests trigger caregiver alerts.
9. Deploy the flow and test by pressing the dashboard buttons.
10. Confirm that the caregiver receives Telegram notifications successfully.

1. Power on the Arduino Nano 33 IoT wearable device.
2. Verify successful Wi-Fi connection through the Serial Monitor.
3. Confirm MQTT communication between the Arduino and Raspberry Pi.
4. Open the caregiver monitoring dashboard in Node-RED.
5. Simulate a sudden impact movement to test fall detection logic.
6. Confirm the buzzer activation when a fall is detected.
7. Verify that the caregiver dashboard updates with the emergency alert in real time.
8. Test the patient assistance dashboard by triggering manual help requests.
9. Confirm that Telegram emergency notifications are received successfully.
10. Test the reset switch to ensure false alarms could be cleared and the system returned to SAFE monitoring mode.

The final prototype successfully combined embedded sensing, IoT communication, and real-time caregiver monitoring into a practical wearable healthcare safety system.

The completed system includes:

1. wearable fall detection using Arduino Nano 33 IoT
2. built-in IMU motion sensing
3. local buzzer emergency alert
4. manual false alarm reset
5. MQTT communication
6. Raspberry Pi caregiver monitoring dashboard
7. patient manual help request dashboard
8. Telegram emergency notifications
9. portable wearable operation using power bank

This project demonstrates how embedded systems can be applied in real-world healthcare monitoring scenarios to improve patient safety and emergency response.

This tutorial is part of an assignment submitted to Deakin University, School of IT, Unit SIT210 - Embedded Systems Development.