Feasibility of Real-Time Astronaut Eye-Tracking During Analog Extravehicular Activities.

INTRODUCTION: As lunar missions increase in frequency and duration, astronauts will conduct more extravehicular activities (EVAs) under demanding conditions that elevate workload, reduce situational awareness, and heighten the risk of operational errors. Despite extensive use in other high-risk fields, eye-tracking technology has not yet been applied to assess gaze behavior during EVAs. We evaluated the feasibility of using mobile eye-tracking within analog EVA suits, focusing on implementation, data quality, livestreaming capability, and the ability to compare gaze behavior across tasks and suits. METHODS: A trained subject performed locomotion, tool-cart navigation, and regolith sampling tasks under lunar-analog conditions while wearing different EVA training suits and a mobile eye-tracker fitted prior to donning the helmet. Data quality was evaluated in terms of robustness, reliability, and accuracy, while data quantity reflected the availability of sufficient data for analysis of fixations, saccades, and head movement. RESULTS: In total, 4.5 h of stable and accurate eye-tracking data were recorded, with no data loss and successful livestreaming between the subject and mission control. Post hoc descriptive analysis indicated task-dependent gaze behavior, with mean fixation durations across suits ranging from 231-367 ms during locomotion, 316-380 ms during tool-cart navigation, and 417-2260 ms during regolith sampling. DISCUSSION: Conducted in a representative EVA environment, this study demonstrates the feasibility of mobile eye-tracking within spacesuit helmets and its potential as an objective measure of cognitive workload and as a live monitoring tool between astronauts and mission control, enabling new possibilities for astronaut training and EVA operations. Maltese F, Steinberg J, Schultz J, Hodkinson P, Runswick OR. Feasibility of real-time astronaut eye-tracking during analog extravehicular activities. Aerosp Med Hum Perform. 2026; 97(7):543-549.