Feasibility of Machine-Learning Models for Coordinating Crew Health Status and Performance Projections in Long Distance Spaceflight Missions

This research seeks to understand how spaceflight crew health status and performance projections can be modeled and communicated during crew-forward spaceflight mission operations. Crew health for the International Space Station (ISS) is closely monitored by the “flight surgeon” console with onboard medical equipment supported by ground-based personnel to provide real-time assessment and health determinations. Future human spaceflight missions to the Moon, and one day Mars, will impose transmission delays that make real-time monitoring by the ground-based team infeasible. Emergent predictive information display technologies and models are needed to enable crew health, augmentation, medical, and performance support, referred to as CHAMPS. Important questions for this research are how and where intelligent agent systems for CHAMPS should be enabled. To address these questions, the authors evaluate the feasibility of using machine-learning (ML) with a human-systems integration approach to provide personalized medical data and performance projections to aid in problem-solving strategies such as event detection, isolation, and recovery (EDIR). As human spaceflight mission operations expand further from Earth, integrating EDIR activities with crew health status and projections of capabilities will be critical to enhance and sustain crew safety and performance. The concepts developed from this research for displaying crew health status and performance projections will contribute to the knowledge gaps in distributed teams performing dynamic real-time problem-solving in future long distance spaceflight missions.