Related: 00_index.md | 01_core_philosophy.md

Architecture Benefits

Feature	Flat Architecture Benefit
Refactoring	Moving a component from "Avionics" to "Payload" is just a string rename. No pointer surgery.
Testing	Unit test any component in isolation by mocking the Signal Backplane.
Parallelism	Easy to identify independent sub-graphs in the flat list for thread-pool execution.
Optimization	Trivially exposes the global f(x,u) function needed for trajectory optimization.

Comparison with Hierarchical Architectures

Aspect	Hierarchical (Scene Graph)	Flat (IDOA)
Data Flow	parent->child->doThing()	Backplane signal read/write
Ownership	Deep nesting, pointer chains	All components are peers
Refactoring	Pointer surgery required	String rename
Testing	Must mock parent chain	Mock backplane only
Symbolic Mode	Deep branching breaks tracing	Linear execution trace
Memory Layout	Scattered across heap	Contiguous state vector

Key Design Wins

1. Simulink/GNC Compatibility

Aerospace engineers think in block diagrams. IDOA maps directly to this mental model—components are blocks, signals are wires.

2. Janus Symbolic Compatibility

CasADi requires a linear execution trace to build symbolic graphs. Deep object hierarchies with virtual dispatch obscure this trace. Flat execution enables clean graph extraction.

3. Data Locality

Contiguous state vectors are cache-friendly and solver-compatible. No chasing pointers through a scene graph.

4. Monte Carlo Performance

Provision once, Stage many times. Memory allocation happens once; each run only resets state and re-wires inputs.

5. Deterministic Execution

Flat component list with topological sort guarantees execution order. No hidden dependencies or callback ordering issues.

When to Use This Architecture

Good fit:

6DOF aerospace simulation
Hardware-in-the-loop testing
Trajectory optimization
Monte Carlo analysis
Multi-vehicle simulations

Less suitable:

Game engines with deep scene hierarchies
GUI applications
Event-driven systems with sparse interactions