In this project, the feasibility of using mobile traffic sensors for binary vehicle classification (i.e., to distinguish passengers from trucks) on arterial roads is investigated. Here mobile sensors refer to those that move with the traffic flow they are monitoring such as global positioning system (GPS), smart phones, among others. Features of vehicle dynamics (e.g. speed related, acceleration/deceleration related, among others) are extracted from vehicle traces collected from real world arterial roads. Machine learning techniques are applied to distinguish passenger cars from trucks using these features.