Vehicles with lower level of autonomy (L2-3) require constant awareness of the human driver during automated cruising, since they are not able to solve all the possible situations that may arise during operation. When the car faces an unresolvable traffic situation, it hands over the control to the human driver. The speed and quality of the handover process is crucial in terms of safety; a handover with large delay time may lead to a fatal accident. We investigated the current human driver models in the literature, and combined these models to get a deeper understanding of the phenomenon of the handover process. We demonstrated the possible ways of improving the drivers' performance during the handover processes, previously shown in the literature separately, and developed a unified way to explain how the performance can be improved. It is concluded that the most important factors that can improve the handover performance are the timely prediction of the handover action, an informative human-machine interface and proper training of the driver.
Paper
Full text
Handover Process Models of Autonomous Cars Up to Level 3 Autonomy
Semantic Scholar · Engineering · 2018
Abstract
Vehicles with lower level of autonomy (L2-3) require constant awareness of the human driver during automated cruising, since they are not able to solve all the possible situations that may arise during operation. When the car faces an unresolvable traffic situation, it hands over the control to the human driver. The speed and quality of the handover process is crucial in terms of safety; a handover with large delay time may lead to a fatal accident. We investigated the current human driver models in the literature, and combined these models to get a deeper understanding of the phenomenon of the handover process. We demonstrated the possible ways of improving the drivers' performance during the handover processes, previously shown in the literature separately, and developed a unified way to explain how the performance can be improved. It is concluded that the most important factors that can improve the handover performance are the timely prediction of the handover action, an informative human-machine interface and proper training of the driver.