Sensitivity of heart rate variability as indicator of driver sleepiness

A number of research studies have been conducted on interventions to minimize accident risks while driving. Among ergonomic interventions is driver sleepiness detection based on biological signals. However, results seem to be inconclusive. This study investigated the sensitivity of sleepiness detection based on drivers' heart rate variability (HRV). Sixteen professional male drivers participated in a laboratory experiment using a driving simulator. Heart beat per minute and peak-to-peak heart beat (RR interval) were monitored during sixty minutes driving, along with theta brain wave activity derived from EEG measurements, Heart rate data were then processed in terms of time-domain, frequency-domain, and fractal (Poincaré plot method). Theta activity was used to determine sleepiness event. Finally, hit rates and false alarm rates were calculated for each heart rate measure to find out the sensitivity in detecting sleepiness. Results showed that the decrement of root mean square of successive differences (RMSSD) of RR interval for 28% and the decrement of short-term variability (SD1) in Poincaré plot for 27% were the two most sensitive parameters for sleepiness detection. Therefore, these biological signals can be considered in developing sleepiness detection system in the future study.