Driver/Vehicle Response Diagnostic System for the Vehicle-Following Case

It is well known that not all drivers drive the same and that the same driver has different driving characteristics with different vehicles. Identifying the characteristics that are unique to each driver/vehicle response opens the way for more personalized and accurate driver assistance systems. In this paper, we consider the problem of identifying the driver/vehicle characteristics by processing real-time driving response data. We propose the use of a Gaussian mixture model combined with the knowledge of dynamic characteristics modeled as probability distributions together with additional logic and appropriate thresholds in order to implement a real-time driver/vehicle response diagnostics system. We focus our efforts on the vehicle-following part of driving. The system is tested on a customized vehicle using different drivers under different driving conditions. We demonstrated that the system can distinguish between different drivers and can classify driver aggressiveness during vehicle following.