Inspiration

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

AstroTEA: Space Agriculture AI

Inspiration

Sustaining life in space is fundamentally a resource management problem. Missions operate under strict constraints such as limited water, energy, and space, and traditional planning approaches are often static and inflexible.

This project was inspired by the question: How can we design a system that not only selects the best biological setup for a mission, but also simulates how it evolves over time?

AstroTEA was created as a decision intelligence system to explore this idea in a structured and practical way.


How We Built It

The system is built around two core components:

Deterministic Decision Engine

We implemented a rule-based scoring system that evaluates:

  • Environment (Mars, Moon, ISS)
  • Mission duration
  • Resource constraints
  • Optimization goals

Each biological component (crops, algae, microbes) is scored, and the system selects the optimal configuration. This ensures that outputs are consistent, explainable, and reproducible.


Simulation Engine

The selected system is then simulated over time using a fully deterministic model.

  • Time progresses in discrete steps
  • Resource consumption and production are tracked
  • Risk evolves dynamically

A simplified representation:

[ Risk(t) = Risk_{base} + \sum_{i=1}^{t} f(resources_i, system_stability_i) ]

The simulation includes:

  • Water recovery loops with delayed effects
  • Energy production and consumption balance
  • Multi-factor life-support metrics

AI Integration

AI is used strictly as a post-processing layer.

It is responsible for:

  • Explaining system decisions
  • Summarizing results
  • Suggesting improvements

All core decisions remain deterministic to maintain reliability.


Challenges

Balancing AI and Reliability

Initial attempts to use AI in decision-making led to inconsistent and non-reproducible results. This required a redesign where AI was limited to explanation only.


Simulation Realism

Early versions of the simulation were too linear and unrealistic. Improvements were made by introducing delayed resource recovery, duration-aware risk scaling, and multi-factor dependencies.


System Integration

Ensuring smooth interaction between decision logic, simulation, and AI outputs required careful separation of responsibilities and clear data flow design.


What We Learned

  • Deterministic systems provide strong control and transparency
  • AI is most effective when used as a support layer rather than a core engine
  • Simulation design requires balancing realism with stability
  • Building a working, testable system is more valuable than adding unnecessary complexity

Conclusion

AstroTEA demonstrates a hybrid approach where deterministic decision-making and simulation are combined with AI-based explanation.

It serves as a functional prototype for exploring sustainable biological systems in space missions and highlights the importance of explainable, reliable decision systems in complex environments.

What we learned

What's next for AstroTEA: Space Agriculture AI

Share this project:

Updates