With the development of level 3 AVs, drivers can now disengage from the driving task for extended periods of time. However, drivers are still responsible for the overall safety of their drive. Moreover, when drivers are not engaged in their monitoring task, they lose situational awareness. This leaves drivers vulnerable when they have to retake control from the AV. This research looks to advance the development of camera-based driver monitoring systems that measure situational awareness. In addition, this research examines the effect of adaptable warning systems on driver situational awareness and takeover performance.
In this study, we use situational awareness as ground truth to compare adaptable warning systems that reengage drivers in the monitoring task. Camera-based driver monitoring systems that measure gaze behavior can be used to adapt warning systems. Twenty-four participants split into three groups were asked to drive for approximately 40 miles in a level 3 AV simulator while completing a visual-manual secondary task. During the drive, participants experienced four events in which they had to disengage from the secondary task and take back control from the AV. Two interface designs based on gaze behavior were compared to a baseline warning system. The Attentional Maintenance group was given an alert throughout the drive after a fixed amount of time in which their gaze was directed away from the road. The State-Contingent Takeover group was given an alert only before takeover events after a fixed amount of time in which their gaze was directed away from the road. Results show that attentional maintenance alerts can increase situational awareness and takeover response time during automation failure. Future research to increase situational awareness is discussed in terms of advancements in cognitive control and bilateral communication between the driver and the AV.
| Identifer | oai:union.ndltd.org:uiowa.edu/oai:ir.uiowa.edu:etd-8470 |
| Date | 01 August 2019 |
| Creators | Kashef, Omeed |
| Contributors | McGehee, Daniel V. |
| Publisher | University of Iowa |
| Source Sets | University of Iowa |
| Language | English |
| Detected Language | English |
| Type | thesis |
| Format | application/pdf |
| Source | Theses and Dissertations |
| Rights | Copyright © 2019 Omeed Kashef |
Page generated in 0.002 seconds