Effectiveness of Warning Signals in Semi-Autonomous Vehicles 2019-01-1013

The rise of automation in the automotive industry has ensured significant progress in vehicle safety and infrastructure. During the transition to full autonomy the driver is often the redundancy and safety feature in the event of a hazard or automation error, accommodations must be made to ensure the best behavior from the driver in these events. Properly handling these events will be more critical as these events become less common and people begin to trust automated driving systems more. This research investigates the case of SAE level-3 automated driving systems, where the driver need not constantly pay attention but is responsible for reaction during hazards. This description can also be applied to abused level-2 systems and prototype systems for higher levels. Findings show valuable quantitative and qualitative assessment of various warning modes for a distracted driver injected into an automated driving failure. Driver response time and behavior for these events are compared to instances with minimal warning systems. The results offer insight into human-vehicle interfaces during the transition towards full autonomy.