Modern vehicles with Advanced Driver Assistance Systems (ADAS) functions and Autonomous Vehicles (AV) in the future are envisioned to increase the safety of drivers and other road users. The objective of this paper is to elaborate on the importance of training drivers of vehicles with ADAS or autonomous functions in order for them to execute those functions properly and yield better road safety. This paper also aims to identify the optimum number of training sessions required for ordinary drivers to operate those technologies properly. In this study, a group of participants was invited which covers different age groups, genders, years of driving experience, and regularities in using ADAS functions in real-life driving. In our simulation-based driving scenario, we incorporated most of the key features which can be found on common road systems. Each participant was requested to drive a virtual Subject Vehicle (SV) several times separated at different time intervals. In each session, participants were instructed to activate and/or deactivate various ADAS functions. In return, their accuracies, as well as their reaction time, have been measured. It is observed that a significant correlation between training frequency and participants’ performance can be found from the data collected. It is also observed that the most significant improvement can be achieved after the first two training sessions. The outcomes of this research will help the automotive industry regulatory bodies around the world to design effective and essential training structures, accordingly. Local licensing authorities can then adapt them and design training and licensing frameworks for AV drivers.

RMIT University: Mohsin Murtaza, Chi-Tsun Cheng, Mohammad Fard; La Trobe University: John Zeleznikow