A simulator-based approach to assess take-over performance in a conditionally automated vehicle

Abstract

The interaction between the driver and the automated driving systems (ADS) will remain a key element of automated driving because drivers are expected to be available to take over control for the case of system failure or limitation in a conditionally automated vehicle, i.e. SAE level 3. A number of studies reported that various factors such as the time budget, the traffic complexity, and the driver’s inattention may influence take-over time and quality. Therefore, the driver’s take-over performance must be carefully analyzed to ensure a safe transition. This study aims to propose a take-over performance assessment method using a driving simulator. A systematic review was conducted to design a driving scenario for unintended take-over events. As a result, a take-over performance test protocol with four take-over situations such as missing lines on a straight and a curved road, road construction, and system failure was designed. Visual and cognitive non-driving related tasks, which influence a driver’s situation awareness and take-over performance, were also considered. It will be proposed as Korean Traffic Safety Regulation to assess the safety of the take-over control in a conditionally automated vehicle from the perspective of the driver.