Abstract:

This article describes the principle of a system designed to improve the quality of teaching and training of motorcyclists. From the user's point of view, the system consists of telemetry data acquisition hardware, a mobile application and a web application. The system should help driving school instructors to better and faster recognize the mistakes of the motorcycle driver in connection with his control. The data from the drive mainly consists of the values of the accelerometer and gyroscope in three axes and the exact geographical position (coordinates from the navigation system). These data are important for evaluating the behavior of the motorcycle driver in certain model situations, such as correct cornering, correct behavior before and after the curve, intersection, etc. The system also includes a neural network model, which, with a sufficient amount of data from driving on the polygon, can evaluate the quality of the student's driving by itself.