AstroSim

Open-source, modular simulation framework for modeling, analyzing, and optimizing sustainable human and robotic presence in space — lunar bases, Mars habitats, orbital stations, and deep-space missions.

Current release: v1.2.0 (public · MIT)

Features

• Modular subsystems: Power, ECLSS, Thermal, Structure, ISRU, Compute/AI, Greenhouse
• Time-stepped engine with Monte Carlo uncertainty analysis
• Budgeting: energy, mass, and reliability tracking
• Interpretation: implications, verdict, references, and recommended actions in study reports
• Reference validation: compare results to NASA BVAD/OCHMO baselines (--validate)
• AI hooks: LLM integration points for optimization and insights (offline mode included)
• Visualization & export: matplotlib plots, HTML web dashboard, JSON/CSV output
• Plugin system: register custom subsystems via @register_subsystem
• Analysis: sensitivity sweeps, trade studies, scenario compare/suite, NL scenario editor
• Example scenarios: lunar base, Mars habitat, greenhouse, closed-loop Mars, orbital station

Quick Start

# Install from GitHub
pip install git+https://github.com/aadriantech/astrosim.git

# Or clone for development
git clone https://github.com/aadriantech/astrosim.git
cd astrosim
pip install -e ".[dev]"

# Run a scenario
astrosim scenarios/greenhouse_lunar.yaml --report --validate --no-plot

# Python API
python examples/run_lunar_base.py
python examples/run_mars_habitat.py

Example output: energy/mass budgets, NASA benchmark PASS/WARN table, engineering verdict, and parameter suggestions (e.g. increase solar or water recovery).

CLI highlights

Flag	What it does
--web	Interactive HTML dashboard
--report	study_report.md + JSON (implications, verdict, actions)
--validate	Compare to reference/benchmarks.yaml; writes validation_report.json
--compare	Multi-scenario metric table
--suite	Run canonical scenario suite
--monte-carlo N	Uncertainty summary
--trade-study	Solar/battery Pareto grid

Documentation

Doc	Description
IO.md	Inputs & outputs — scenarios in, reports/dashboards out
ARCHITECTURE.md	Engine, subsystems, budgets, events, plugins
SCENARIOS.md	YAML/JSON schema, parameters, examples
API.md	Public interfaces: Simulator, Subsystem, AIHooks
TUTORIAL.md	15-minute getting started
PLUGIN_COOKBOOK.md	Custom subsystem patterns
CONTRIBUTING.md	Human contributor guide
CHANGELOG.md	Release history
ROADMAP.md	Phase completion status (v1.2.0)
Example studies	Reproducible lunar/Mars trade reports
AGENTS.md	Agent methodology (PDD→TDD→CDD→AYSU)

Build static docs: bash scripts/build_docs.sh

Project Layout

src/astrosim/
  engine/        # Simulation loop and Monte Carlo runner
  subsystems/    # Pluggable habitat subsystem models
  budgeting/     # Energy, mass, reliability accounting
  validation/    # NASA benchmark checks
  export/        # JSON, CSV, study reports, interpretation
  ai/            # LLM hook interfaces
  analysis/      # Sensitivity, compare, suite, trade studies
  visualization/ # Matplotlib + HTML dashboards
reference/       # Official benchmark values (BVAD/OCHMO/ISS)
scenarios/       # YAML scenario definitions
examples/        # Runnable demo scripts
docs/            # Architecture, API, scenarios, contributing

Agent workflow (how this was built)

AstroSim was built with PDD → TDD → CDD → AYSU — open-sourced as ai-coding-scaffold.

Share / promote: copy-paste posts for X, Reddit, HN

Contributing

See CONTRIBUTING.md.

pip install -e ".[dev]"
pytest

Citation

See CITATION.cff for academic use.

License

MIT