SLEEPINESS EVALUATION BASED ON FACIAL EXPRESSSION AND EYE BLINK : A STUDY IN TRAIN SIMULATOR

• Main Author: AFIANO HIDAYAT NIM 13414041, CALVIN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/26182
• Institution: Institut Teknologi Bandung

Based on the Komite Nasional Keselamatan Transportasi (KNKT), train accidents due to human factor are still high. One of the main causes of accidents from human factors is sleepiness. Sleepiness assessment is done through facial expressions and winks using train simulator. This study aimed to quantify the level of drowsiness characteristics to reduce accidents that occur due to drowsiness. The sleepiness level classification process was carried out using several measuring instruments such as Observer Rated Sleepiness (ORS), Karolinska Sleepiness Scale (KSS), blink rate, and blink duration. Facial expressions and blinks showed good results in describing and predicting the level of sleepiness. Assessment using ORS and SSC is done by giving a questionnaire that has been equipped with details of the level of sleepiness. Assessment is done every 20 minutes to see the change in sleepiness. In addition, the measurement of eye parameters is done by watching at recording of driving activities using a simulator. Eye parameter data collection is done at the last minute on every 20 minutes driving. The results of the study showed significant differences for sleepy people based on facial expressions and blinks. The results of this study are useful to provide a better classification of facial expressions and blinks so that the study results can be used to monitor work in real time. Data obtained from this study can be useful to predict the level of sleepiness that felt by machinists. In addition, research can be applied to technology that can determine the level of sleepiness and predict the level of drowsiness.