Astronaut looking across Earth’s illuminated horizon from orbit

Human health beyond Earth

Healthy People.
Farther Together.

Open-source, offline-first astronaut health intelligence for long-duration human spaceflight.

Turning physiological, biological, cognitive, behavioral and spacecraft-environment data into personalized, explainable and actionable health insight — even when Earth is offline.

Built for NASA Space Apps Challenge 2026

Explore MITOCHONDRIA

Demo, team and documentation are public. Mission Console tasks may require sign-in and an authorized mission membership.

The mission health loop

From signal to supported action.

Not another fitness dashboard. The platform direction unifies personal baselines, laboratory context, cognition, behavior, environment, radiation, data quality and provenance.

  1. 01

    Monitor

    Capture connected physiological, biological and environmental signals.

  2. 02

    Personalize

    Interpret change against each crew member’s evolving baseline.

  3. 03

    Correlate

    Connect health, cognition, workload, habitat and radiation context.

  4. 04

    Detect

    Surface meaningful trends while accounting for data quality.

  5. 05

    Explain

    Make every signal traceable, understandable and reviewable.

  6. 06

    Act

    Support timely, protocol-aligned human decisions.

  7. 07

    Learn

    Preserve provenance and improve mission knowledge over time.

Mission capabilities

Intelligence designed around the human.

Monitoring

A unified view across physiological, environmental and self-reported signals, each shown with its source, freshness and quality — LIVE, STALE or DISCONNECTED is always visible.

Open Live Monitoring

Personal baseline

Each crew member is compared with their own versioned baseline, kept separate from clinical reference intervals and mission limits.

Learn about baselines

Trend & pattern detection

Governed trend policies look across days, not single readings, and report confidence, confounders and missing context.

Open Health Intelligence

Risk assessment

Decision support that states uncertainty openly. Research-only relationships can never become operational alerts or a diagnosis.

Read the safety boundaries

Explainable intelligence

Every finding carries its evidence path so crew and ground teams can see why it appeared — and challenge it.

Open Explainable Alerts

Action guidance

Versioned protocol guidance such as repeat measurement or workload review supports, never replaces, qualified judgment.

Open Action Guidance

Offline Mission Mode

Encrypted local storage and a store-and-forward queue keep the health loop running when Earth is out of contact.

Learn how offline works

Device-agnostic integration

One adapter contract for wearables, laboratory systems and habitat telemetry; unknown metrics are rejected, not guessed.

Learn about integration

The Mitochondria is conceived as human decision support—not an autonomous doctor. It does not diagnose and does not replace qualified medical judgment.

Mission principle

Astronaut +
spacecraft environment
= one health context.

Human health cannot be interpreted independently from mission conditions.

CO₂, radiation, sleep and circadian disruption, workload, habitat conditions and personal physiology form one connected operational picture.

Local mission nodeOPERATING
Earth linkDELAYED
Health loopCONTINUES

Offline Mission Mode

No Earth connection?
The mission health loop continues.

The architecture is designed for local operation: resilient capture, local protocols and rules, store-and-forward synchronization, and graceful degradation when connectivity disappears.

Architecture direction: these capabilities describe the intended system design and are not presented as completed flight functionality.

Technology

How the platform is built.

Eight in-depth topics explain each part of the architecture, why it matters beyond low Earth orbit, and exactly what is implemented today.

  • Layered, vendor-neutral design

    System Architecture

    Presentation, application, domain, integration and infrastructure boundaries keep health meaning independent from vendor APIs and screens.

    Learn more about System Architecture
  • Universal device adapter SDK

    Device Integration

    One adapter lifecycle — discovery, connection, capabilities, streaming, quality, normalization, cursors, health and disconnect — for every source.

    Learn more about Device Integration
  • Wearable source classes

    Wearables

    Web Bluetooth, Android Health Connect and Apple HealthKit bridge contracts and vendor-cloud imports converge on one canonical metric model.

    Learn more about Wearables
  • Periodic biomarker ingestion

    Medical & Laboratory Data

    CSV, explicitly mapped FHIR Observations, manual entry and medical-device readings keep specimen, method and reference-context provenance.

    Learn more about Medical & Laboratory Data
  • Habitat and radiation context

    Environmental Telemetry

    Habitat and dosimetry adapters keep compartment, detector, unit and quality context on the same mission timeline as physiology.

    Learn more about Environmental Telemetry
  • One reviewed meaning per metric

    Canonical Health Data Model

    Metrics are normalized through a reviewed registry with deterministic serialization, unit safety, provenance and research/operational classification.

    Learn more about Canonical Health Data Model
  • The health loop continues without Earth

    Offline-First Architecture

    Encrypted local storage, append-only events, a protocol cache and an idempotent sync queue keep the astronaut-side loop working during network loss.

    Learn more about Offline-First Architecture
  • Mission-scoped access, audited

    Security & Privacy

    Forced row-level security, mission-scoped roles, pseudonymized research export, audit trails and encrypted local storage form the current security boundary.

    Learn more about Security & Privacy
View all technology topics

Open-source mission

Built to be inspected.
Built to be extended.
Built beyond one vendor.

An open architecture direction for reproducible science, transparent decisions, adaptable device connectors, health-data interoperability, community contribution and local or self-hostable deployment.

View the repository

Extensible by design

Adapters and interfaces that can evolve with mission needs.

Reproducible foundations

Methods and provenance designed to withstand scrutiny.

Interoperable direction

A path across devices, data models and mission systems.

Challenge 2026

Built for the 2026 NASA International Space Apps Challenge

An independent open-source project exploring resilient, human-centered health intelligence for crews venturing farther from Earth.

Visit NASA Space Apps 2026