LunCo is an open-source, high-fidelity robotics co-simulation platform built for System-Level Engineering and Concept of Operations (CONOPS). It bridges the gap between systems architecture, behavioral modeling, and real-time operations, providing the digital substrate for the next generation of space exploration.
Website | Documentation Hub | Join Discord
Most simulators focus on isolated physics. LunCo focuses on the System-of-Systems. We simulate not just how a rover drives, but how it interacts with the power grid, obeys the flight software, adheres to the SysML blueprint, and contributes to the overall mission timeline.
| Pillar | Technology | The Value Proposition |
|---|---|---|
| System Identity | SysML v2 | Definitive blueprints serving as the "Source of Truth" for system structure and requirements. |
| Native Collaboration | WebTransport | Built-in, multi-user engineering. Design, test, and operate in the same scene simultaneously. |
| Scene Composition | OpenUSD | Industrial-grade 3D interop with NVIDIA Omniverse. USD is our world format, not a sim engine. |
| Behavioral Rigor | Modelica / ROM | Multi-domain behavioral simulation (power, thermal, robotics) with high-fidelity dynamics. |
| Mission Autonomy | rhai Scenarios | Hot-reloadable per-entity flight software & declarative mission timelines — sense and command the world through the same API as the UI. |
| Robot Control | HIL / SIL / ROS2 | Native Hardware/Software-in-the-Loop support. Bridging logical intent to physical actuators. |
| Mission Control | XTCE / MAVLink | Standardized telemetry compatible with NASA OpenMCT, YAMCS, and professional ground stations. |
- System-Level Co-Simulation: Orchestrate multiple specialized engines (Modelica, Avian3D, GMAT) into a single cohesive mission scenario.
- Planetary Scale Precision: Built on a specialized f64 (double precision) spatial math foundation, ensuring absolute stability from millimetre-scale parts to lunar orbits.
- Native Multi-User: Architecture built from the ground up for collaboration. Every edit, command, and telemetry stream is replicated across the network.
- Headless-First & AI-Ready: Designed for automation. Scalable for massive parallel Monte Carlo analysis and end-to-end AI agent training.
- Scriptable Autonomy: Attach hot-reloadable rhai scenarios to any entity — lifecycle hooks, sensing, and declarative mission timelines drive behavior with no recompile, through the same command/query API the UI and AI agents use. See the Scripting Guide.
- Composition over Simulation: We use OpenUSD to compose complex scenes from modular parts, then attach simulation behaviors via our multi-engine backend.
Both windowed apps also run in your browser. These are early preview builds — expect rough edges and missing features:
- lunica.lunco.space — the Modelica engineering workbench
- sandbox.lunco.space — the physics sandbox
git clone https://github.com/LunCoSim/lunco-sim.git
cd lunco-simThen launch the entry point that fits your goal (each also builds for the browser via scripts/build_web.sh):
The flagship: celestial bodies, ephemeris, solar-system-scale precision, and the complete flight-software / robotics / avatar stack.
cargo run --release -p luncosimValidate robotics, suspension, and environment interactions in a collaborative 3D scene (windowed, or headless with --no-ui).
cargo run --release -p lunco-sandbox --bin sandboxFocus on Modelica modeling, schematic diagramming, and subsystem analysis.
cargo run --bin lunicaDriving it from code or an AI agent? Launch any app with
--apiand drive it over HTTP/MCP — see the AI Agent Guide and the task-oriented skills.
- Documentation Hub — Usage guides and architectural deep-dives.
- Scripting Guide — Write hot-reloadable rhai scenarios & mission timelines.
- AI Agent Guide & Skills — Drive and extend LunCoSim from code or an AI agent.
- Crates Index — A map of our 60+ specialized crates.
- Principles — Our non-negotiable mandates: TDD-First, Headless-First, and Tunability.
| Milestone | Status | Description |
|---|---|---|
| System-Level Core | ✅ Foundation | Multi-domain co-simulation (USD + Modelica + Avian3D) with f64 precision. |
| Real-world Validation | 📝 Planned | HIL/SIL Integration (Spec 027) for Hardware-in-the-loop validation. |
| Industrial Interop | 📝 Planned | NASA GMAT (Spec 022) for orbital mechanics and ROS2 for robotics control. |
| Advanced Physics | 📝 Planned | PINN-based Terramechanics (Spec 025) for high-fidelity regolith interaction. |
| Autonomous Missions | 📝 Planned | Agent-Driven Sim (Spec 033) and Mission Replay/Audit (Spec 020). |
LunCo is built by a global community of engineers and researchers making professional space engineering tools accessible to everyone.
Want to join the mission? Apply to the core team.