What are the different data sources used for the driving analysis?

Our applications aim to provide an objective assessment of driving behavior. To perform this type of analysis, we use data measured by smartphone sensors as well as contextual information. Thus, our algorithms exploit:

• GPS sensor measurements recorded with a period of one point per second. This information makes it possible to quantify the vehicle's dynamic and energy trajectory throughout the trip;
• Measurements from the phone's inertial unit provide useful information to determine the type of activity and mode of transport used;
• Position data (latitude/longitude) are not essential, but they make it possible to verify that the route taken does indeed use roadways;
• The vehicle characteristics, entered by the driver in the app, are input parameters of the analysis algorithms;
• External data are added to all these quantities to contextualize driving scores: road type and meteorological (weather) data;
• Phone brightness and call detection are used in the calculation of the distraction score.