Heart Rate Variability-Based Drowsiness Detection System

Abstract

Driving has become one of the most significant activities in people's lives. Driving while drowsy has become one of the core causes of fatal traffic accidents and human causalities. This article proposed a driver drowsiness detection system using edge computing and heart rate variability to avoid traffic accidents. An electrocardiogram sensor was used to retrieve the signal and was transferred to the processing device through a built-in Bluetooth module. Then the signal was processed using the hamming window and fast Fourier transform algorithm, and with the help of edge computing theory, this system calculates the average LF/HF value. The result shows that the average LF/HF ratio goes down when a person goes from the conscious state to the sleep state, and the corresponding value is 0.19 to 0.101 for males and 0.18 to 0.101 for females.