Using Deep Learning-Enabled Computer Vision to Measure Team Dynamics During Simulated Medical Emergencies in Space: A Feasibility Study

Abstract

Long-duration spaceflight crews will rely on teamwork and behavioral skills to successfully operate in extreme environments. The development of automated metrics of team dynamics may play an important role on the assessment of team performance in a reliable and unobtrusive manner. In this study, we explored the feasibility of using a computer vision system to capture motion data of multiple crew members during simulated medical events in a spacecraft simulator, and generate multiple metrics of team dynamics.