Smartphone Enabled Nonintrusive T&D Detection

Abstract

Now-a-days the most top dangerous behaviors for drivers is Texting-while-Driving(T&D).There has been many mobile phone applications and interesting systems that have been used to detect and reduce T&D. However, to make the System practical, there must be a key property that will be used to distinguish Driver�s mobile phone from passenger�s. There are many existing solutions to this problem which involve user�s manual input, or utilize specific localization devices to determine driver�s location. Contrast to all of this solution, In this paper, we propose a method which automatically detect T&D without using any additional devices. The basic idea of our system is when a user is using the mobile phone, the smart phone embedded sensors (i.e., gyroscopes, accelerometers) will collect the connected information that includes touch-strokes, holding orientation and vehicle speed. . The collective information will then be examined to see if there exists any specific T&D patterns. In addition with the infrastructure free and high accuracy, the method does not access the content of messages and therefore is privacy-preserving.