Analysis of pupil dilation using fractal dimension and eye-head integrated tracking system in a multitasking environment
Metadata[+] Show full item record
[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] In this study, a laboratory experiment was designed to investigate the relationship between the pupil dilation and multitasking performance by using eye and head integrated human-in-the-loop (HITL) system. Thirty-six university students joined in this study as participants, and they were asked to do several different tasks simultaneously and give correct response in time according to the dynamic system change. The primary task domain is a process gauge monitoring task, which simulates operator responsibilities in the control room of oil and gas refinery plants. Two different operation interfaces were designed for the primary task to identify the interface influence on human multitasking performance. The secondary task was Multi-Attribute Task Battery (MATB), which is consisted of three tasks, system monitoring, target tracking, and dynamic resource management. During the experiment, participants need to keep moving their eye and head to switch among four tasks; an eye-head integrated tracking system was introduced to continuously record participants' pupil change and visual scan path during the entire experiment time. Also, participants experienced two levels of complexity (low and high) scenarios during the experiment. NASA-TLX was applied to measure participants' workload when facing to the different operation interface and complexity scenarios. For exploring the pattern of pupil changes, the fractal dimension analysis was used as a tool to measure if their pupil dilations have any self-similar features in different condition. Analysis of the data revealed overall multitasking performance was highest in low complexity by using Visual Thesaurus overview display. Moreover, a positive correlation was found between the performance and fractal dimension of pupil dilation. This finding suggested eye-tracking data could be a potential valid source to identify the performance level in the multitasking environment.
Access to files is limited to the University of Missouri--Columbia.