SFT 2026-27 - Lunar Sleep-Shift Scheduling Tool
SFT 2026-27 - Lunar Sleep-Shift Scheduling Tool
(Sleep Shift Recommendation System, SSRS)
Executive Summary
Build an onboard tool that generates personalized, evidence-based sleep/wake schedules and fatigue-mitigation advice for astronauts, running offline and improving through feedback. Objective: produce recommendations for sleep, wake, naps, caffeine, and sleep-aid timing autonomously. Deliverables: end-user requirements with access protocols; hardware/software technical specs; high-level mission requirements and challenges; a requirements document; a student presentation; a project video; and a posting on nasahunch.com.
Requested By
NASA Johnson Space Center - Space Medicine / Fatigue Management. Program: NASA HUNCH.
Sponsors: HRP Flight Surgeon (JSC-SD311); N. Hall (GRC).
Problem Statement
Astronaut sleep-shift plans are built on the ground by fatigue specialists and can take about a day to turn around. Moon and Mars missions add communication delays and frequent operations (for example, multiple EVAs per week), so crews must re-plan sleep without the ground. An onboard tool is needed that recommends sleep, nap, caffeine, and sleep-aid timing autonomously.
Requirements Overview
Daily sleep-onset / wake recommendations per crewmember
Nap, caffeine, and sleep-aid (e.g., melatonin) timing
Gradual shifting rules (<= 2 hr per sleep period; ~24 hr to adjust per 1 hr shift)
A fatigue score (e.g., 0-100) rating how fatiguing a schedule is
Light / exercise guidance; runs offline on a laptop or tablet
A feedback loop that learns from adherence and alertness
Major Constraints
Must run offline — no cloud access required onboard
Earth communication delay (~2.6 s to the Moon; minutes to Mars)
Operations always take priority over optimal sleep
Runs on a laptop (SSC) or tablet; lunar ~14-day light/dark cycle
Key Challenges
Encoding sleep-science rules (sleep debt, circadian shift limits)
Individual sleep need varies (~5-10 hr) and crews often shift together as a team
Producing a meaningful, trustworthy fatigue score
Building a feedback / learning loop from crew input
Reliable offline operation with conservative safe defaults on data loss
Cards
1 — Other Points / Comments
Two distinct roles: a biomedical lead (Westover) and a software lead (HUNCH / Hayes)
Rollout is crawl-walk-run: specialists use it first, then crew in-mission for autonomy
The fatigue team is to provide a prioritized (Pareto) list of sleep-shift guidance and student resources
Two user types are served: the fatigue-management team and the crew
Security / facial recognition is optional, not required for the MVP
2 — Examples of Excellence
A working MVP that outputs a clear daily plan: sleep/wake, naps, caffeine, and sleep aids
Honors operational priorities and lets users iterate or override recommendations
Reports a fatigue score (e.g., 0-100) so a flight director can judge whether a schedule is acceptable
Runs fully offline on a laptop or tablet
Emits copy-paste scheduling constraints for the Ops Planner (e.g., 'asleep by 20:00, wind-down by 18:00, no exercise after 17:00')
3 — Examples of Innovation
A feedback / learning loop that adapts to each crewmember's adherence and alertness
Wearable biometric input (e.g., a fitness tracker) for circadian phase and sleep quality
Vigilance testing or facial-recognition fatigue detection
An AI model trained on fatigue-specialist knowledge
Live ground-to-crew data sync, with future auto-integration into the Ops Planner timeline
Suggestions for High School Students
Scope to a Minimum Viable Product: a clear recommended schedule (sleep, wake, naps, caffeine, sleep aids) plus optional light/exercise tips
AI is not required — a rules-based scheduling tool is acceptable
Facial recognition / vigilance testing is optional (above-and-beyond only)
Ops Planner timeline integration can be deferred to a later phase
Delivery target is roughly a May timeframe; there is no specific mission deadline