EARLY DEVELOPMENT OF EMBEDDED FATIGUE MONITORING SYSTEM BASED ON HEART RATE
Earn Tzeh Tan, Zaini Abdul Halim, Vincent Kok · 2015
The seriousness of road traffic injuries had been aware by public since the past decades. Nevertheless, it is very difficult to avoid driver fatigue or drowsy driving while fatigue is suspected to be the primary cause in more than 20% of road fatalities. In this article, an embedded based detection system is proposed to alert the driver when drowsiness or driver fatigue is detected. The conducted statistical analyses in this study show that the normal heart rate of a person is significant difference when he/she falls in fatigue. Thus, the framework of the proposed system uses heart rate as the classification feature for the fatigue detection. In overall, the proposed system consists of an Arduino UNO microcontroller integrated with a pulse sensor, GPS module, and GSM module. The measured heart rate from the pulse sensor is further processed with programmed Artificial Neural Network (ANN) algorithm and a warning SMS is sent with current GPS location when fatigue was detected. From several conducted testing and evaluations, the functionality of the developed system is verified and the mean accuracy achieved for the recognition of fatigue is 94%. In summary, the study reveals that besides of vision solutions, classification based on biometric signal can be an alternative approach for fatigue detection and the proposed framework has a potential to reduce the occurrence of road crashes due to drowsy driving.