Skip to content

CLAUDE.md

Latest commit

 

History

History
44 lines (30 loc) · 1.76 KB

CLAUDE.md

File metadata and controls

44 lines (30 loc) · 1.76 KB

ADCS Lifecycle Demo

What this is

A demonstration of bidirectional requirements traceability for a satellite Attitude Determination and Control System (ADCS). The demo traces the full lifecycle from SysMLv2 structural specification through symbolic analysis (SymPy), numerical simulation (scipy), and human expert attestation — with all traceability stored as RDF and all artifacts version-controlled in git.

Core principle

Evidence does not verify requirements; evidence supports a human judgment that requirements are satisfied. Models are imperfect representations of physical systems. The engineer judges model adequacy and evidence sufficiency. Only human attestation connects evidence to requirement satisfaction.

Architecture

Three-layer ontology:

  • Layer 1 (W3C): PROV-O for provenance, Dublin Core for metadata
  • Layer 2 (SysMLv2): Structural model, requirements, satisfy relationships
  • Layer 3 (RTM): Evidence, attestation, traceability (custom rtm: namespace)

Key directories

  • ontology/ — OWL TBox (rtm.ttl) and namespace constants (prefixes.py)
  • structural/ — SysMLv2 RDF model (satellite.ttl, parameters.ttl)
  • analysis/ — SymPy symbolic analysis + scipy numerical simulation
  • evidence/ — Content hashing and RDF evidence binding
  • traceability/ — RTM assembly, SPARQL queries, human attestation
  • pipeline/ — Stage orchestrator (stages 1-8)
  • interrogate/ — Visualization, explanation chains, reproducibility checks

Toolchain

  • Python 3.12+, managed by uv
  • rdflib for RDF/SPARQL, sympy for symbolic math, scipy for ODE integration
  • ProofScript/ProofBuilder pattern reimplemented from gds-proof (self-contained)

Running

uv run python -m pipeline.runner

Tests

uv run pytest