Application of Hidden Markov Model on Car Sensors for Detecting Drunk Drivers

Abstract

The ability to detect drunk driving behavior on roadways enhances road safety by significantly reducing the risk of fatal accidents. In this paper, a set of measurements, readily available via on-board vehicle sensors, was selected to detect drunk driving behaviors based on learning in accordance with certain drunk driving cues. A Hidden Markov Model (HMM) method was applied for each of the collected time series data, which correspond to the selected measurements. The prediction accuracy attained using each measured variable was derived and analyzed. The longitudinal acceleration achieved the best average prediction accuracy, for detecting both drunk and normal driving behaviors, with an accuracy that is equal to about 79%.