Journals Proceedings

Abstract

Accidents due to drowsy driving have shown a steep increase in the past decade. An efficient safety measure is required to mitigate the impact of accidents. This paper details the system architecture for real-time, non-obstructive, automatic detection and alarming for driver drowsiness. Data/sensor fusion technology incorporated with the system monitors the driver drowsiness including fatigue and cardiac problems. Alertness of the driver is monitored by analyzing the heart rate by a non-invasive method. The system reads the pulse rate of the driver through multimodal physiological sensor unit embedded on steering wheel. Formulated data processing algorithm incorporated with the system measures the heart rate of the individual and invokes an emergency alarm, if it falls below a specified threshold value. A second level of alarm is issued to the concerned authorities and rescue forces, if the heart rate variation is found to be consistent. The second level of warning incorporates alert messages constituting vehicle identification number and GPS coordinates. One of the novel ideas incorporated is the development of multiple sensors embedded in the steering wheel and unique heart rate calculation algorithm. The system is capable of measuring the heart rate and dynamically alerts the driver or the rescue team about the driver drowsiness, to avert accidents. At the end of this paper, an analysis on the various methods for failure analysis and prevention is also provided.