Heart Rate-Based Fatigue Monitoring System with K-Nearest Neighbor Algorithm for Burnout Prevention

Abstract

The purpose of this study is to design a fatigue level monitoring system consisting of a monitoring device based on a Wemos D1 Mini microcontroller and a MAX30102 sensor, and a monitoring server with Thingsboard IoT Platform and Laravel PHP Framework using K-Nearest Neighbor Algorithm for burnout prevention. Design based research was used in this study. The monitoring system reads the activity level of the heart rate and the time interval between beats using the k-Nearest Neighbor algorithm, the heart rate variability value, and the questionnaire results. The monitoring results were delivered to the user's email and Telegram app. The monitoring system test results reveal that the entire system is functioning properly. However, there are still weaknesses in reading activities that require a lot of movement. This was attributed to the appearance of motion artifacts in heart rate sensor data obtained using photoplethysmography and/or PPG procedures at high wavelengths. From this research, a new alternative method was obtained to help maintain and monitor fatigue levels in order to prevent burnout. A fatigue monitoring system can be an alternative to preventing burnout in someone based on data received by the user.