An experimental study was carried out to identify effective advisory and alarm stimuli to be used in a drowsy driver detection system. The envisioned system has three stages. In the first stage, previously developed detection algorithms would compute online drowsiness levels. If a driver's drowsiness level exceeds a predetermined threshold the system would proceed to stage two. At this point an initial advisory tone and a voice message would be played. If the driver does not respond, he or she would experience a realerting alarm. The third stage of the system would give the driver an option of using a drowsiness countermeasure to help maintain the re-alerted state.

The goal of the present research was to determine the effectiveness of possible stimuli to be used in the second and third stages of the envisioned system. Eight initial advisory tones, two voice messages, eight alarm sounds, and five peripheral stimuli were investigated as part of stage two. In addition, six drowsiness countermeasures to be used in stage three were investigated. Eight graduate students in the Human Factors Engineering program at Virginia Tech volunteered as subjects. Subjects drove the automobile simulator throughout the experimental session. Data were collected using subjective opinion, paired comparisons, and effectiveness ratings.

This study succeeded in answering many question regarding stimuli to be used in a drowsy driver detection, advising, and alerting system. The results of the study indicated very effective stimuli to be used in the advising and alerting stages of the envisioned system.