0164 Sleep, Circadian, and Homeostatic Influences on Visual Attention Performance During Team-Based Simulated Ground Vehicle Missions

Abstract Introduction Sleep insufficiencies and fatigue can impair cognitive performance and operational effectiveness, compromising warfighter readiness and mission success. Visual attention in particular is sensitive to sleep loss—especially with increasing time-on-task—an issue for long-term missions. However, it remains unclear which subjective or objective sleep-related indicators best predict visual attention in dynamic, visually demanding operations. This study investigated the extent to which various sleep factors predicted visual attention during a collaborative, team-based simulated ground vehicle mission. Methods N=240 participants (40 teams of n=6) completed four simulated drives within a shared environment under different, randomly ordered assistance conditions: No Assistance (manual operation of vehicles), Object Detection System (ODS; targets identified with video/audio), Vehicle Automation (self-driving vehicle), and Vehicle Automation + ODS (i.e., Combined). During each drive, their primary mission was to search for and locate target objects of interest as a team. Sleep outcomes were assessed at baseline (appointment time, sleep quantity, sleep quality, KSS, time awake, nap, caffeine consumption) and in-task (PERCLOS =30%, KSS). Roadway Environment Attention was calculated as the percentage of time eye-gaze intersected the simulated roadway environment during each drive, while Mean Look Reaction Time (RT) was the average time for eye-gaze to intersect with target objects once in view. A series of mixed linear growth models were run for each attention outcome using the above predictors. Results Roadway Environment Attention during the Combined Assistance condition was significantly lower than during the No Assistance (Tukey-adjusted p.001), Vehicle Automation (Tukey-adjusted p.001), and ODS conditions (Tukey-adjusted p=.016); no other predictors were significant (all ps≥.163). There were no significant predictors of Mean Look RT (all ps≥.240), though higher KSS showed a marginal trend toward slower RTs (p=.067). Conclusion Contrary to our expectation, few sleep-related factors predicted visual attention during this short-duration, team-based simulated mission. Assistance systems had a stronger influence on gaze behavior than sleepiness, suggesting workload offloading drove attention reductions. Longer or more fatigue-inducing paradigms may be needed to capture sleep-related declines in visual attention. Support (if any) Automotive Research Center in accordance with Cooperative Agreement W56HZV-24-2-0001 U.S. Army DEVCOM Ground Vehicle Systems Center. DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited. OPSEC # 10191.