Project number: PSR-17-11

Funding source: Caltrans

Contract number: 65A0674 TO 006

Funding amount: $105,000

Performance period: 2/1/2019 to 1/31/2020

Over the next several decades, highly automated driving systems (HADS) will become increasingly common on our roads, greatly reducing traffic accidents and road congestion. However, for the foreseeable future, the human driver will be required to take control when automation fails. Therefore, it is critical to understand how take-overs from HADS affect driver performance and whether driver take-over performance will affect crash rates rate and related issues of congestion. We review extant research on driver take-over performance, identifying key factors influencing driver performance. We then develop a virtual-reality driving simulator designed to evaluate specific differences in driver performance during take-overs from HADS versus current full-manual driving. In a study using this simulator, we find no clear difference in driver performance during obstacle avoidance maneuvers and identify several benefits that HAD take-overs provide to driver performance.

Link to project data: https://doi.org/10.7910/DVN/BWCCET