01 / ADAPTATION
Learning under altered physics
Generalization, predictive models and reliable adaptation when space conditions differ from Earth-trained assumptions.
Independent research initiative
Intelligence
beyond Earth.
What if artificial intelligence were designed for space from the beginning? We explore how learning, autonomy and scientific discovery change beyond Earth's familiar conditions.
Machine Learning × Space ScienceDarmstadt, Germany · Open internationallyIndependent · Emerging
Our premise
Space is not just another destination for AI. It is a new condition for intelligence.
Research agenda
Questions at the frontier of intelligent systems.
Our focus is methodological: what must change in machine learning when physics, experiments and mission resources cannot be taken for granted?
02 / DISCOVERY
Autonomous scientific discovery
Closed-loop experiments that select measurements, learn from observations and refine scientific hypotheses with limited intervention.
03 / CONSTRAINTS
Intelligence within mission limits
Learning and decision-making under finite energy, computation, communication windows and experimental opportunities.
2026 · Inaugural public activity
An open research agenda, not a closed club.
Our first community activity was designed for World Space Week 2026: an invitation to identify open questions across machine learning, autonomous systems and space science.
World Space Week 2026 · Cycle I
Space-Native AI 2026:
Intelligence Beyond Earth
Submit a research question or project concept, help identify emerging themes, and take part in the first Space-Native AI research cycle.
Community directions
One name. Many scientific frontiers.
Space-Native AI is the umbrella initiative. Specific research tracks can evolve without changing the community's identity.
01
Learning in Space
Machine learning, physical adaptation and world models.
02
Autonomy in Space
Intelligent systems, robotics and experimental agency.
03
Bio in Space
Learning-guided life-science investigations and future autonomous laboratories.
04
Living in Space
Emerging questions in sustainable space habitats, agriculture and resource cycles.
Join the conversation
What should an intelligent system learn when every experiment, watt and observation counts?
Cycle I runs through 15 November 2026. Contributions will be synthesized into themes, selected submissions will receive feedback, and Featured Contributions will help shape follow-up community activities.
Write your contribution ↗