Applications of high-frequency telematics for driving behavior analysis

Processing driving data and investigating driving behavior has been receiving an increasing interest in the last decades, with applications ranging from car insurance pricing to policy-making. A popular way of analyzing driving behavior is to move the focus to the maneuvers as they give useful information about the driver who is performing them. Previous research on maneuver detection can be divided into two strategies, namely, 1) using fixed thresholds in inertial measurements to define the start and end of specific maneuvers or 2) using features extracted from rolling windows of sensor data in a supervised learning model to detect maneuvers. While the first strategy is not adaptable and requires fine-tuning, the second needs a dataset with labels (which is time-consuming) and cannot identify maneuvers with different lengths in time. To tackle these shortcomings, we investigate a new way of identifying maneuvers from vehicle telematics data, through motif detection in time-series. Using a publicly available naturalistic driving dataset (the UAH-DriveSet), we conclude that motif detection algorithms are not only capable of extracting simple maneuvers such as accelerations, brakes, and turns, but also more complex maneuvers, such as lane changes and overtaking maneuvers, thus validating motif discovery as a worthwhile line for future research in driving behavior. We also propose TripMD, a system that extracts the most relevant driving patterns from sensor recordings (such as acceleration) and provides a visualization that allows for an easy investigation. We test TripMD in the same UAH-DriveSet dataset and show that (1) our system can extract a rich number of driving patterns from a single driver that are meaningful to understand driving behaviors and (2) our system can be used to identify the driving behavior of an unknown driver from a set of drivers whose behavior we know.