Risk assessment of vehicle cornering events in GNSS data driven insurance telematics

We propose a framework for the detection of dangerous vehicle cornering events, based on test statistics related to the no-sliding and no-rollover conditions. The input variables are estimated using circle fitting of position measurements (giving the instantaneous radius of the driving trajectory), and a Rauch-Tung-Striebel smoother applied to a state-space model describing the vehicle dynamics. The accuracy of the resulting test statistic is evaluated in a simulation study. In addition, we present the results of a field experiment where three smartphones were used as measurement probes. Since the estimation only uses position and speed data from global navigation satellite system receivers, the framework is particularly well-suited for smartphone-based insurance telematics applications, aiming to avoid the logistic and monetary costs associated with e.g., on-board-diagnostics or black-box dependent solutions. The design of the estimation algorithm allows for instant feedback to be given to the driver, and hence, supports the inclusion of real time value added services in usage-based-insurance programs.