WakeUp IOT Smart Alarm, Weight and Sleep Monitoring

A cloud based smart alarm clock using IoT sensors. This system aims to integrate various hardware and software components to provide an efficient and user-friendly alarm clock that leverages cloud connectivity for enhanced functionality.

Features

• Web based application - set alarm from any device
• Alarm rings only when necessary. (If user is on his bed around the scheduled time)
• Smart Weight Tracking
• Sleep Pattern Tracking

Techstack

Demonstration

• This alarm system is designed to help you stay on top of your meetings.
• You input all your meetings into the app, and the hardware detects if you're still in bed before a meeting starts.
• If you're in bed 5 minutes before a scheduled meeting, the alarm will ring continuously until you get up.
• The alarm detects your presence by measuring an increase in weight reading through a load cell and a decrease in distance reading using a TOF and ultrasound sensor.

Setting-alarm.mp4

Important

For Demonstration purposes, A 20kg loadcell was used only for the head. However for the actual Project a combination of four 20 kg load cells were placed under the bed to weigh the whole body.

UI Screens

UI.mp4

Data Flow

1. ESP32 Microcontroller connects to WebApplication Server websocket, gets alarm data in realTime
2. During NightTime, User Weight and Movement data from TOF, Ultrasound and HX711 Weight Sensors are sent to InfluxDB via Telegraf, HiveMQ.
3. Live real time monitoring of data and database data are available through the cloud Grafana dashboard
4. Data from InfluxDB is read by AWS Sage Maker for model based data Analytics using Machine Learning Models
5. All User Alarms and Aggregated Statistics are stored in MYSQL Database in Azure.
6. NextJS Frontend Server and NodeJS(ExpressJS) Backend Server are passed through a reverse proxy for user Abstraction and Security.
7. Google OAuth 2.0 with JSON Web Token is implemented for authenticating users.

Data_flow.mp4

IOT System Design

• By leveraging sensors, microcontrollers, communication modules, MQTT brokers, and real-time data processing, we have developed a cohesive system that offers valuable insights and enhanced user interaction.
• The project highlights the potential of IoT technology in everyday applications, transforming a traditional clock into a multifunctional device capable of improving user awareness and decision-making.
• Our system integrates various components seamlessly. The end devices, represented by the ESP32 microcontroller, ensure accurate data collection using sensors like the HX711 for weight measurement and the VL53L0X for distance measurement.
• These devices collect essential data, which is then transmitted to the MQTT broker for real-time processing and analytics.
• The OLED display and buzzer provide immediate feedback to users, enhancing the interactivity and functionality of the system.
• The integration of WebSocket communication further enables real-time updates and remote monitoring capabilities.

Circuit Diagram

Sensors:

• HX711: This sensor is an analog-to-digital converter (ADC) which is used for weight sensing features, switching off the alarm only when weight is removed from the bed.
• VL53LOX: VL53L0X is a Time-of-Flight (ToF) ranging sensor. It is used for movement recognition, which gives the sleep scores based on user’s movement during sleep.
• HC-SR04: This sensor provides 2cm - 400cm non-contact measurement function. It measures distances using ultrasonic waves. This is used to detect if someone is there, magnitude of change in conjuction with the TOF sensor for double check(since the frequency of waves are different in both).

Actuators:

• OLED Display: An OLED (Organic Light Emitting Diodes) display shows time information. OLEDs have a much higher contrast ratio compared to traditional displays, making them visible from all angles and even in low-light situations.
• Buzzer: The component that generates the alarm sound. It's the crucial part of the alarm clock, notifying or waking the user at the designated times.

Controller:

ESP32 Microcontroller: ESP32 is a feature-rich SoC with integrated Wi-Fi and Bluetooth connectivity for a wide-range of IoT applications. In the given circuit it is connected to the sensors, collects sensor data, transmits it using the WiFi to the Cloud.

Analytics

Sage_maker_analytics.mp4

• We have used AWS Sagemaker for deploying, training and modelling of several models and datasets.
• We have added few attributes such as Start time the user’s sleep, End time of the User’s sleep, sleep quality, time in bed, weight and Distance (Movement of the person).
• We have calculated the User’s time in bed by the start and end time of sleep by the user.
• Weight monitoring is done by the loadcell which is placed under the bed.
• The tof sensor, VL53LOX and Ultrasound sensor HC-SR04 which is placed near the bed measures the distance of the User moving and tracking the percentage of deep sleep he/she is in.
• The sleep quality of the user is trained by models to give a precise quality by checking the total time duration of the sleep and the movement the user while sleeping.
• For modelling, we have used, Preprocessing, Visualization, EDA, Linear regression, KNN, Logistic regression, Decision tree, Random Forest and SVM for better accuracy and got successful results.

Note

We have used our sleep data from 10.4.2024 to 10.5.2024 to train and test the models.

Libraries needed for ESP32 main.ino file

Name	URL
Websocket	https://github.com/gilmaimon/ArduinoWebsockets?tab=readme-ov-file
WiFi	https://github.com/esp8266/Arduino/tree/master/libraries/ESP8266WiFi/src
WiFiClientSecure	https://github.com/esp8266/Arduino/blob/master/libraries/ESP8266WiFi/src/WiFiClientSecure.h
MQTT	https://github.com/adafruit/Adafruit_MQTT_Library
Adafruit_VL53L0X	https://github.com/adafruit/Adafruit_VL53L0X
ArduinoHTTPClient	https://github.com/arduino-libraries/ArduinoHttpClient
HX711	https://github.com/bogde/HX711
Adafruit_SSD1306	https://github.com/adafruit/Adafruit_SSD1306
NTPClient	https://github.com/arduino-libraries/NTPClient
Time	http://playground.arduino.cc/Code/Time/
Adafruit_GFX	https://github.com/adafruit/Adafruit-GFX-Library