Date of Award
Doctor of Philosophy (PhD)
Dr. David M. Neyens, Committee Chair
Dr. B. Rae Cho
Dr. William G. Ferrell
Dr. Mary Elizabeth Kurz-Edsall
Dr. Sara Lu Riggs
The number of vehicles on the road with advanced and automated driving support systems (DSSs) is increasing. However, there may be some issues related to the implementation of DSSs in vehicles. One of those issues caused by the automated DSSs relates to the drivers' being out-of-the-loop. As drivers' roles are transitioned from system operators to systems supervisors (as in autonomous vehicles), drivers' situation awareness of the driving surroundings may decrease which could negatively affect their responses when they need to take control of the vehicle from the malfunctioned (or failed) DSSs. Additionally, with both the adaptive cruise control (ACC) and lane keeping (LK) systems engaged, the longitudinal and lateral positions of the vehicle are under the control of automation and the vehicle becomes a semi-autonomous vehicle (i.e., the vehicles are now at level 2 automation based on the definitions of the National Highway Traffic Safety Administration taxonomy for automation). In semi-autonomous vehicles, drivers are more likely to interact with non-driving tasks and engage in risky behaviors (e.g., long glances away from the forward road way), as the demand of the driving tasks is much lower than manually driving and driving with only ACC engaged. This may worsen drivers' responses to the failures of semi-autonomous vehicle components, when drivers are engaged in non-driving tasks. The objectives of this dissertation were to assess how drivers respond to the failures of the LK system with different levels of vehicle automation and to assess the effects of drivers' engagements in non-driving tasks on their behaviors associated with a failure of the LK system. This dissertation also investigates if a lane departure warning would mitigate the negative effects of out-of-the-loop problem brought on by automation and improve drivers' responses to the LK system fails especially when drivers are engaging both the ACC and LK systems. Additionally, the relationships between drivers' personalities and attitudes toward automation and their responses during the failure of the LK system were evaluated. Three experiments were used to address the dissertation research objectives. The results demonstrate that drivers in semi-autonomous vehicles (level 2 automation vehicles) have less safe behaviors (e.g., more engagement in non-driving tasks and longer glances away from the roadway) than their peers who were manually driving the vehicles. During the failures of the LK systems, drivers in semi-autonomous vehicles have worse driving behaviors compared to their counterparts driving manually or driving with the LK system engaged. Non-driving tasks also increase drivers' reaction time to safety critical events in semi-autonomous vehicles. However, the effects of audible lane departure warnings on drivers' responses to potential lane departure events were not consistent between the level 0 automation condition (i.e., the manual driving condition) and level 2 automation condition (i.e., the automated driving condition). Overall, audible warnings with 1.48 s prediction time assist drivers' in responding to the lane departure events following the failure of the LK system in semi-autonomous vehicles. However, the effects of audible warnings on drivers' responses to the potential lane departure events are divergent when drivers are manually operating the vehicles. Though audible warnings as one type of discrete feedback of automation activities help drivers improve their responses to safety critical events in semi-autonomous vehicles, they cannot solve the out-of-control loop problem caused by automation. Future work should evaluate if continuous feedback could address the out-of-control loop problem brought on by automation and keep drivers in the vehicle control loop in semi- or fully- autonomous vehicles.
Shen, Sijun, "Quantifying Drivers' Responses to Failures of Semi-autonomous Vehicle Systems" (2016). All Dissertations. 1664.