The fields of navigation and motion inference have rapidly been transformed by advances in computing, connectivity, and sensor design. As a result, unprecedented amounts of data are today being collected by cheap and small navigation sensors residing in our surroundings. Often, these sensors will be embedded into personal mobile devices such as smartphones and tablets. To transform the collected data into valuable information, one must typically formulate and solve a statistical inference problem. This thesis is concerned with inference problems that arise when trying to use smartphone sensors to extract information on driving behavior and traffic conditions. One of the fundamental differences between smartphone-based driver behavior profiling and traditional analysis based on vehicle-fixed sensors is that the former is based on measurements from sensors that are mobile with respect to the vehicle. Thus, the utility of data from smartphone-embedded sensors is diminished by not knowing the relative orientation and position of the smartphone and the vehicle. The problem of estimating the relative smartphone-to-vehicle orientation is solved by extending the state-space model of a global navigation satellite system-aided inertial navigation system. Specifically, the state vector is augmented to include the relative orientation, and the measurement vector is augmented with pseudo observations describing well-known characteristics of car dynamics. To estimate the relative positions of multiple smartphones, we exploit the kinematic relation between the accelerometer measurements from different smartphones. The characteristics of the estimation problem are examined using the Cramér-Rao bound, and the positioning method is evaluated in a field study using concurrent measurements from seven smartphones. The characteristics of smartphone data vary with the smartphone's placement in the vehicle. To investigate this, a large set of vehicle trip segments are clustered based on measurements from smartphone-embedded sensors and ...


    Access

    Download


    Export, share and cite



    Title :

    Sensor Fusion for Smartphone-based Vehicle Telematics


    Contributors:

    Publication date :

    2017-01-01


    Type of media :

    Theses


    Type of material :

    Electronic Resource


    Language :

    English



    Classification :

    DDC:    629




    Smartphone-Based Vehicle Telematics: A Ten-Year Anniversary

    Wahlstrom, Johan / Skog, Isaac / Handel, Peter | IEEE | 2017


    Smartphone-based Vehicle Telematics For Naturalistic Driving Studies

    Tarlochan, Faris / Dun, Susan / Mohammed, Semira Omer et al. | IEEE | 2022


    A secure vehicle-based telematics environment used for enhanced smartphone communication

    Conradi, Peter / Graefe, Dennis / Lueders, Timo et al. | VDE-Verlag | 2016


    System and method for actual and smartphone telematics data based processing

    FERNANDES STEVEN J / FEERO STEVEN R | European Patent Office | 2017

    Free access