Two significant constraints shape the research methodology. First, there is currently no in-mission crew time available to complete measures. In-mission data will need to be collected unobtrusively from available data streams (e.g., audiovisual, existing records such as schedules, and actigraphy). Second, Artemis II is anticipated to be a crew of 4 astronauts in flight for 10 days followed by additional crewed Artemis missions. This creates a scenario where there is important data for understanding risk characterization; however, there will be a small sample size from Artemis II, and risk characterization will need to be updated over time as additional crews fly in later Artemis missions.

Behavioral medicine risk

The research team conducted a literature review of spaceflight and spaceflight analog evidence to identify key indicators, or constructs, of behavioral health and performance. The team narrowed these to 4 constructs that appear to be most impacted by these environments. Contextual factors that impact behavioral health were also identified and linked to possible critical periods likely to be observed during the Artemis II mission. The team also identified natural language processing (NLP) tools to analyze the audio recording (LIWC and STRESSnet) and is currently drafting a codebook that links the identified BMed constructs to the dictionaries provided by these tools. Finally, the research team is conducting a literature search to determine behavioral coding schemes to characterize BMed constructs using video data.

Team risk

The research team drew from the Team Risk deliverable presented to the Human Subjects Review Board (HSRB) in 2021, which identified key indicators, or constructs, of team functioning and performance with recommended measures. The research team considered each indicator in the context of the Artemis II mission, selecting those that were most applicable. Next, the team identified unobtrusive measurement methods (e.g., lexical analysis of transcripts) and/or behaviors related to those constructs, which will allow researchers to code and analyze team interactions captured in crew audio/video data and space-to-ground loops.

HSIA risk

HSIA is a risk that is based in part on vehicle system capabilities and how those capabilities enable crew-ground team performance in responding to unanticipated major vehicle anomalies and executing safety-critical procedures. The research team draws from analyses of responses to actual International Space Station (ISS) anomalies, related human-systems engineering disciplines (cognitive engineering, safety engineering, reliability engineering, resilience engineering) as well as organization research (high-reliability organization) to identify key indicators, or constructs, of human-systems team performance. The team is in the process of identifying candidate constructs along with their associated crew behavior indicators as well as system capabilities needed to support resilient crew-ground team performance. In parallel, the team is conducting exploratory analysis using transcripts and videos from past missions (Apollo, ISS) to identify mission-critical communication patterns and derive resilience-indicating behaviors for the codebook.

Sleep risk

The research team drew from foundational sleep literature to determine key sleep metrics (sleep duration, sleep efficiency, wake after sleep onset, and sleep latency). The state of sleep as observed in video recordings will be correlated against a known tool used during spaceflight (actigraphy) to benchmark passive tools for measuring sleep. Using both actigraphy and video recordings, the team will examine the cognitive consequences of sleep loss and circadian misalignment on the constructs defined by the BMed, Teams, and HSIA groups. The team plans to specifically analyze patterns of behavior and language during critical periods when the crew might be less alert (for example, just after waking when humans experience sleep inertia) or when the crew might need to perform high-workload tasks.

Two significant constraints shape the research methodology. First, there is currently no in-mission crew time available to complete measures. In-mission data will need to be collected unobtrusively from available data streams (e.g., audiovisual, existing records such as schedules, and actigraphy). Second, Artemis II is anticipated to be a crew of 4 astronauts in flight for 10 days followed by additional crewed Artemis missions. This creates a scenario where there is important data for understanding risk characterization; however, there will be a small sample size from Artemis II, and risk characterization will need to be updated over time as additional crews fly in later Artemis missions.