This directory contains comprehensive examples and case studies demonstrating Morphogen's multi-domain capabilities.
Status: Design Document Domains: Physics (Racing), Neural Network, Genetic Algorithm, Rendering, Telemetry, Recording Complexity: Advanced Hardware: GPU-optimized (RTX 3060 12GB target)
A complete racing AI training pipeline demonstrating how Morphogen's unified operator model enables seamless integration of physics simulation, neural network inference, genetic algorithms, and real-time visualization — all in one composable graph.
What This Demonstrates:
- Multi-domain composition (Physics + NN + GA + Rendering + Telemetry + Recording)
- GPU-accelerated parallel evaluation (64+ agents simultaneously)
- Genetic algorithm operators for neural network evolution
- Real-time visualization and debugging
- Deterministic training with perfect reproducibility
- 10-100× performance vs traditional Unity + Python approaches
Key Insight: This is one of Morphogen's strongest use cases because it merges traditionally-separate technologies (game engine physics, ML frameworks, custom GA code) into a single clean pipeline.
Use Cases:
- Autonomous racing AI
- Self-driving car simulation
- Drone flight controllers
- Robot navigation
- Game AI (any physics-based agent control)
- Reinforcement learning benchmarks
Related Documentation:
- reference/genetic-algorithm-operators.md — GA operator catalog
- specifications/operator-registry.md — Operator metadata format
- ADR-002: Cross-Domain Architectural Patterns — Unified patterns
Status: Design Document Domains: Geometry, FluidNetwork, ThermalODE, FluidJet, CombustionLight Complexity: Advanced
A complete multi-physics modeling example showing how Morphogen's operator graph paradigm extends from audio/graphics into engineering physics.
What This Demonstrates:
- Multi-physics system modeling (draft pressure, flow networks, thermal ODEs, jets, combustion)
- Cross-domain integration (Geometry → Fluid → Thermal → Combustion → Visualization)
- Reference-based composition (anchors, frames) for physical systems
- Design optimization workflow (parameter sweeps, objective search)
- New domain requirements (FluidNetwork, ThermalODE, FluidJet, CombustionLight)
Key Insight: The J-tube fire pit is basically a little multi-physics engine in steel — and Morphogen is exactly the kind of thing that wants to model that.
Use Cases:
- Fire pits (secondary combustion optimization)
- Mufflers (exhaust flow and back-pressure)
- HVAC systems (air duct networks)
- Heat exchangers (thermal-fluid coupling)
- Burners (combustion quality estimation)
Related Documentation:
- specifications/physics-domains.md — Detailed operator specifications for new physics domains
- ADR-002: Cross-Domain Architectural Patterns — Reference systems and anchors
- architecture/domain-architecture.md — Complete domain vision
Status: Design Document Domains: OrbitalMechanics, Aerodynamics, RocketEquation, PartsAssembly, PhysicsIntegration, FailureMechanics Complexity: Advanced
Demonstrates how Morphogen can model KSP-style physics (orbital mechanics, aerodynamics, rocket staging, part assembly) and become a framework for aerospace simulations.
What This Demonstrates:
- Real-time multi-domain physics (orbits, aero, propulsion, structures)
- Part-based assembly system using operator graphs
- GPU-accelerated physics integration
- Level-of-detail physics switching (patched conics ↔ N-body)
- Cross-domain integration (Geometry/TiaCAD → Physics → Audio → Visualization)
- Educational platform for orbital mechanics and aerospace engineering
Key Insight: KSP's entire gameplay loop maps perfectly onto Morphogen's domain architecture — proving Morphogen can handle real-time game physics and aerospace simulation.
Use Cases:
- Spaceflight simulation games (KSP-like)
- Aerospace education (teaching orbital mechanics)
- Mission planning tools (trajectory optimization, launch windows)
- Satellite constellation design (Starlink, etc.)
- Research simulations (spacecraft dynamics)
Unique Features:
- Integration with J-tube combustion domain for realistic engine modeling
- AudioDomain integration for engine sounds and aerodynamic noise
- Procedural planet generation using NoiseDomain
- Part geometry from TiaCAD
Related Documentation:
- specifications/physics-domains.md — Physics operator specifications
- ADR-002: Cross-Domain Architectural Patterns — Reference systems and anchors
- specifications/geometry.md — TiaCAD integration for part geometry
- examples/j-tube-firepit-multiphysics.md — Combustion domain for engines
Coming soon...
- Real-time audio synthesis with field-driven modulation
- Fractal visualization with parameter animation
- Cross-domain: audio → visual (sonification)
- J-Tube Fire Pit (current) — Multi-physics thermal-fluid system
- Racing AI Pipeline (current) — Racing car dynamics with neural control
- Kerbal Space Program Physics (current) — Orbital mechanics, aerodynamics, rocket staging
Coming soon...
- N-body gravity simulation with Barnes-Hut optimization
- Fluid simulation (Navier-Stokes on GPU)
- Particle-field coupling (PIC/FLIP methods)
- J-Tube Fire Pit (current) — Parametric design with optimization
- Kerbal Space Program (current) — Rocket design and mission planning
Coming soon...
- Muffler design (exhaust flow and acoustics)
- Heat exchanger optimization
- HVAC system balancing
Coming soon...
- Option pricing with Monte Carlo
- Stochastic volatility models (Heston)
- PDE-based pricing (finite difference methods)
- Racing AI Pipeline (current) — Genetic algorithm + neural network evolution
Coming soon...
- Neural fields (NeRF-style)
- Fourier Neural Operators for PDE solving
- Differentiable physics simulation
Start with the J-Tube Fire Pit example to understand:
- How operators compose into pipelines
- How domains integrate via references/anchors
- How parameters flow through the graph
- How optimization wraps around pipelines
Use the physics domains (FluidNetwork, ThermalODE, etc.) as templates:
- Study operator specifications
- Note cross-domain coupling patterns
- Understand determinism requirements
- See MLIR lowering strategies
Examples show end-to-end workflows:
- Geometry setup
- Simulation pipeline
- Visualization
- Optimization
- Export/results
We welcome new examples! Guidelines:
- Focus on multi-domain integration
- Include complete operator specifications
- Show cross-domain flows clearly
- Provide validation (analytical, experimental, or CFD comparison)
- Document use cases and generalizations
Template structure:
# Example Title
## 1. Physical/Conceptual System
## 2. Morphogen Modeling Pipeline
## 3. Domain Requirements
## 4. Operator Specifications
## 5. Complete Code Example
## 6. Validation & Testing
## 7. Generalizations & Extensions| Level | Description | Example |
|---|---|---|
| Basic | Single domain, few operators | Sine wave oscillator |
| Intermediate | 2-3 domains, cross-domain flows | Audio → Visual sonification |
| Advanced | 4+ domains, multi-physics | J-Tube Fire Pit |
| Research | Novel domain combinations | Differentiable physics + ML |
- specifications/physics-domains.md — Physics domain operators
- specifications/operator-registry.md — Operator metadata format
- specifications/geometry.md — Geometry domain (TiaCAD patterns)
- specifications/coordinate-frames.md — Frames and anchors
- ADR-002: Cross-Domain Architectural Patterns — Unified patterns
- architecture/domain-architecture.md — Complete domain vision
- guides/domain-implementation.md — Implementing new domains
- Architecture questions: See ADR-002
- Domain design: See architecture/domain-architecture.md
- Operator specs: See specifications/operator-registry.md
Morphogen is not a library. Morphogen is a platform.