Analysis of Electrocardiogram and Photoplethysmogram Signals to Detect Car Driver Drowsiness Using the Threshold Method

Abstract

Drowsiness in drivers can be detected through human physiological signals. Heart Rate and Respiration Rate are physiological parameters that can be used for detecting drowsiness. This study aimed to see the accuracy of the drowsiness detection system using the threshold method with heart rate and respiration rate parameters. The sensors used were a MAX86150 and a pulse sensor to get ECG and PPG signals as it uses dry electrodes so that there is no need to attach electrodes to the driver’s body. To find out the drowsiness conditions in drivers, a subjective assessment with the Karolinska Sleepiness Scale (KSS) consisting of 9 sleepiness scales (levels) was used. The correlation value between the drowsiness condition levels and respiration rate is low, with a value of −0.26 and the correlation value between the drowsiness condition levels and heart rate is moderate, with a value of −0.49. The drowsiness detection system using the threshold method with heart rate and respiration rate parameters has a low accuracy value of 81% when drivers are in a condition between drowsy and not drowsy. Therefore, it is still difficult to apply to actual driving conditions. This study is very useful to find out the relationship between human physiological signals and drowsiness conditions when driving and to find out the most appropriate drowsiness classification method to apply to driving conditions, as seriously as possible, to help reduce the number of traffic accidents due to drowsiness.