Naturalistic Observations of Driver and VRU Interactions

Abstract: Due to the high number road traffic related fatalities of VRU (Vulnerable Road Users) in the EU, it is necessary to improve the performance of the AEB (Autonomous Emergency Brake) systems. This thesis covers the pre-stage of the AEB-sensor development, from the pre-study of different European accident databases to get a better understanding of which type of accidents are the most common and which cases are the most severe, to the Naturalistic Driving Study performed in Barcelona, to record a big amount of critical situations between the driver and the VRU.
The Naturalistic Driving Study was performed with a vehicle equipped with a Lidar (Light Detection And Ranging), two cameras, CAN (Controller Area Network)-system information and a VBOX. All the critical situations was then analyzed using 58 different parameters, from VRU head and torso orientation, weather conditions to intersection type and relative velocity.
The main objective of this thesis is to present a conclusion of how the VRU and the driver react to a critical situation. This data will be useful for future sensor development for more precise Autonomous Emergency Brake systems and the development of Automatic Emergency Steering.