Applying Adaptive Gain Theory to Pupillary Dynamics During Hands-Off L2 Driving Under High Cognitive Load

Arousal plays a vital role in facilitating the human ability to respond flexibly in a goal-directed manner. Adaptive Gain Theory (AGT) posits that the Locus Coeruleus-Norepinephrine (LC-NE) system is implicated in this regulation and that pupillometry is a non-invasive indicator of LC activity. Most tests of AGT have focused on highly controlled cognitive and perceptual tasks, however it has been proposed that the LC-NE system plays a more general role in regulating the cortical based mechanisms involved in arousal. Transitions of control from intermediate levels of automation to manual driving are a potentially useful testbed for complex processes that are mediated by this neuromodulatory arousal system. To date, pupillometry has focused on the detection of cognitive load in the context of automated and manual driving. Therefore, this driving simulator experiment had two general aims: to establish whether pupillometry during hands-off Level 2 (L2) driving was a reliable and valid indicator of cognitive load, and to test the predictions of AGT in an applied context. The size and reactivity of drivers’ (N = 38) pupils were measured during hands-off L2 driving with and without a cognitive load task, followed by critical and non-critical transitions of control. Analysis revealed that mean, not standard deviation, of pupil diameter was a reliable indicator of cognitive load. Furthermore, partial support for AGT predictions was found; higher baseline pupil diameter was associated with smaller task-evoked pupillary responses (TEPRs), even after correcting for regression to the mean artifacts. Finally, more critical events were associated with larger tonic pupil diameter, indicating that drivers were exerting greater effort to manage the transition. These results indicate that pupillometry is a useful measure of cognitive load as well as an indication of the physiological arousal needed to facilitate cognition during transitions of control. We discuss the need for neuroscientific theories and methodologies in the pursuit of valid and reliable Driver Monitoring Systems (DMS).