Afiatur Rochmah, Siti Sendari, Ilham A.E. Zaeni
People move from one area to another through land, sea, or air transportation. Land transportation includes buses, trains, bicycles, and cars. The duration of time needed to drive is not short, even when using a car; the time can be used to sleep for the passengers, but not so for the driver. Driving is a very monotonous job that results in fatigue and drowsiness. Fatigue and drowsiness can have a big effect on safety and security on the road. It can be prevented by using technological capabilities. Development of drowsiness detection uses the reading mechanism of electroencephalogram (EEG) with the classification of artificial neural networks. The method of the artificial neural network used is ANN Backpropagation. ANN Backpropagation method is a supervised artificial neural network. The data used in this study wass the value of eSense attention, theta waves, low alpha waves, and high alpha waves obtained from brain wave sensor output. The research framework used included data collection, data processing, data analysis, and conclusions. The architecture used in this study was 4 input neurons, 8 first hidden layer neurons, 4 second hidden layer neurons, and 1 output neuron. The other parameter was logsig-logsig-tansig for the use of the activation function, the learning rate of 0.1, and momentum of 0.85. The process was managed and produced the best output in the form of Mean Absolute Percentage Error (MAPE) of 0.02%. The results of the classification of drowsiness detection have an error rate of 10% and an accuracy rate of 90%. © 2019 IEEE.
Universitas Negeri Malang, Department of Electrical Engineering, Malang, Indonesia
