Date of Award

2024

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Applied Cognitive Science and Human Factors (PhD)

Administrative Home Department

Department of Cognitive and Learning Sciences

Advisor 1

Kelly Steelman

Committee Member 1

Shane Mueller

Committee Member 2

Elizabeth Veinott

Committee Member 3

Joonbum Lee

Abstract

Advanced Driver Assist Systems (ADAS) are SAE level 2 features that require active driver control while engaged. Although drivers can have their feet off the pedals and hands off the steering wheel, they must maintain supervisory control of the vehicle. However, when these features are in use, drivers may become less aware of their surroundings, increasing the risk of accidents. To counter this problem, vehicle manufacturers use driver monitoring strategies to ensure drivers remain attentive while ADAS features are active. These monitoring strategies employ instructions to ensure drivers are engaged in the driving task. These instructions are broadly classified into hands-on-the-wheel and hands-free (eyes-on-the-road). The hands-on-the-wheel strategy measures driver engagement by examining steering wheel torque, while the hands-free strategy tracks the driver's eyes to ensure they remain on the road. Although both strategies are commonly used in vehicles with SAE level 2 automation, there is a lack of publicly available data on their effectiveness and impact on takeover performance.

In this dissertation, three studies were conducted to measure the effects of the hands-on-the-wheel and eyes-on-the-road driver monitoring strategies on situation awareness, change detection, mind-wandering, and gaze behavior. Study 1 was exploratory and utilized a low-fidelity semi-automated driving task to examine the effects of the two engagement strategies on driver attention during level 2 ADAS driving. Study 2 was an extension of Study 1 and moved to more naturalistic automation-related change detection in addition to a SAGAT freeze-probe protocol and comfort, fatigue, engagement, and takeover readiness measures in addition to the ones measured in Study 1. Study 3 extended Study 2 in a medium-fidelity driving simulator to investigate the effects of the two driver engagement strategies on driving performance variables and driver attention. Study 1 found that the hands-on-the-wheel strategy promoted less mind wandering during level 2 automated driving. Study 2 found that while the hands-on-the-wheel strategy also promoted less mind wandering, it promoted higher situation awareness, more perceived engagement with automated driving, less self-reported fatigue, and faster response to takeover requests. On the contrary, Study 3 found that the eyes-on-the-road strategy exhibited higher SA, faster responses to takeover requests, and less steering wheel variability but closer following distances post-takeover.

Although the three studies have mixed findings, the hands-on-the-wheel strategy appears more promising because it engages drivers physically with the driving task, potentially leading to safer driving behaviors. This work has broader implications for SAE level 2 and 3 ADAS features, reinforcing the need for an engagement strategy with driver monitoring systems. Even as level 3 and higher technologies are developed, the results here inform strategies for automation-level step-downs as the drivers are brought back into actively controlling the vehicle.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS