ATA Chapter Onboarding Playbook

Version: 1.0.0
Last Updated: 2026-01-16
Applies To: AMPEL360 Q100 Program

Overview
Prerequisites
Step-by-Step Onboarding Checklist
Chapter-Specific Customization Guide
HITL Escalation Decision Tree
Source-Tracking Requirements
Effectivity Rules Configuration
Validation & Compliance
Common Pitfalls

Overview

This playbook provides a comprehensive guide for onboarding new ATA Chapters into the AMPEL360 Q100 OPT-IN Framework, implementing the full GENESIS → SSOT → LLM ENGINE → CSDB → PUBS → EXPORT pipeline.

End-to-End Process Flow

GENESIS (Uncertainty Space)
    ├── O-KNOT: Discovery Layer
    ├── Y-KNOT: Justification Layer
    └── KNOT: Framing Layer
         ↓
SSOT (Certainty Space)
    ├── LC01: Problem Statement (KNOTS, KNU_PLAN, TOKENOMICS, TBD_REGISTER)
    ├── LC02-LC14: Lifecycle-bound artifacts with _derivation.yaml
         ↓
LLM ENGINE METADATA LAYER
    ├── Enrichment with confidence scores
    ├── Source tracking (docId, span references)
    ├── HITL escalation for low-confidence inferences
         ↓
CSDB (S1000D Compliant Data Modules)
    ├── DM: Data Modules (YAML → XML)
    ├── BREX: Business Rules Validation
    ├── Transforms: XSLT for format conversion
         ↓
PUBS (Publications)
    ├── AMM: Aircraft Maintenance Manual
    ├── TRN: Training Materials
         ↓
EXPORT
    └── Multiple formats (PDF, HTML, IETP)

Prerequisites

Before starting chapter onboarding:

Framework Understanding: Read ATA-00-00 Reference Implementation
Axis Assignment: Determine which OPT-IN axis (O/N/T/P/I) the chapter belongs to
Stakeholder Identification: Identify chapter owner and key stakeholders
Domain Expertise: Ensure domain expert availability for HITL review
Tools Setup: Python 3.6+, XSLT processor, JSON Schema validator

Step-by-Step Onboarding Checklist

Phase 1: Scaffold Generation

1.1 Run Scaffold Script

cd OPT-IN_FRAMEWORK/_templates
python scaffold_chapter.py --chapter XX --title "Chapter Title" --axis AXIS_NAME

1.2 Verify Folder Structure

AXIS_NAME/
└── ATA_XX-TITLE/
    └── ATA-XX-title/
        └── XX-00-general/
            ├── GENESIS/
            ├── SSOT/
            └── PUB/

Phase 2: GENESIS Layer Population

Phase 3: SSOT Layer Configuration

Phase 4: LLM Engine Metadata Layer Integration

Phase 5: CSDB Integration

Phase 6: Publication & Export

Phase 7: Validation & Quality Assurance

Chapter-Specific Customization Guide

High-Priority Chapters

ATA 28 (Fuel) - T-TECHNOLOGIES

Domain Specifics:

LH₂ storage systems
Cryogenic handling procedures
H₂-specific terminology (mass flow kg/s, energy density kJ/kg)
Safety-critical considerations (flammability, embrittlement)

Customization Points:

TOKENOMICS: Higher TT allocation (500-600 TT) due to novel technology
TBD_REGISTER: High volume of TBDs expected for novel H₂ systems
LLM Confidence: Lower auto-accept threshold (0.80) for safety-critical content
HITL: Mandatory review for all H₂ safety procedures

ATA 71 (Power Plant) - T-TECHNOLOGIES

Domain Specifics:

Fuel cell stacks
Electric motors and power electronics
Thermal management
Hybrid propulsion architecture

Customization Points:

KNOT Dependencies: Heavy dependency on ATA 28 (Fuel)
SSOT/LC04: Detailed design interfaces with ATA 24 (Electrical Power)
CSDB: High illustration count (ICN) for complex assemblies
Effectivity: Variant-specific for different fuel cell configurations

ATA 95 (AI/ML) - T-TECHNOLOGIES

Domain Specifics:

Novel Q100 extension chapter
Learning assurance and certification challenges
AI model versioning and traceability
Explainability requirements

Customization Points:

GENESIS: Extensive O-KNOT discovery phase (regulatory uncertainty)
TBD_REGISTER: High TBD count for certification approach
LLM Metadata: Special handling for AI training data provenance
BREX: Custom validation rules for AI artifact documentation

ATA 96 (DPP/Traceability) - N-NEURAL_NETWORKS

Domain Specifics:

Digital Product Passport integration
Blockchain/DLT for traceability
Lifecycle data capture
Regulatory compliance (EU DPP requirements)

Customization Points:

SSOT/LC08: Enhanced configuration management
Source Tracking: Bidirectional traceability to physical components
CSDB: Extended metadata fields for DPP attributes
Export: Additional format for DPP data exchange (JSON-LD)

HITL Escalation Decision Tree

┌─────────────────────────────┐
│   Content Extracted/        │
│   Generated by LLM          │
└──────────┬──────────────────┘
           │
           ▼
    ┌──────────────┐
    │ Confidence   │
    │ Score?       │
    └──┬───────┬───┘
       │       │
   ≥0.85│      │<0.60
       │       │
       ▼       ▼
    ┌────┐  ┌─────────────────┐
    │AUTO│  │  MANDATORY HITL │
    │PASS│  │  - Human Review │
    └────┘  │  - Mark source  │
            │  - Escalate     │
            └─────────────────┘
       │
   0.60-0.84
       │
       ▼
    ┌─────────────────────┐
    │  Content Type?      │
    └──┬──────┬───────┬───┘
       │      │       │
   Safety  Part#  Other
       │      │       │
       ▼      ▼       ▼
    ┌────┐ ┌────┐ ┌────────┐
    │HITL│ │HITL│ │Optional│
    │REQ │ │REQ │ │Review  │
    └────┘ └────┘ └────────┘

HITL Triggers

Always Escalate:

Safety-critical procedures (any confidence level)
Part numbers or specifications (unless ≥0.90 confidence AND verified against parts DB)
Torque specifications (never auto-accept)
Novel terminology not in glossary
Conflicting information from multiple sources

Conditionally Escalate:

Medium confidence (0.60-0.84) AND first occurrence in chapter
Cross-chapter dependencies
Regulatory references
Revision/update to existing approved content

HITL Response Template:

hitl_review:
  artifact_id: "KNU-XX-YY-NNN-REQ-001"
  span_reference: "docId:ATA-XX-DOC-1234, span:567-890"
  confidence_score: 0.72
  issue: "Ambiguous terminology for H₂ flow rate unit"
  reviewer: "{{REVIEWER_NAME}}"
  review_date: "{{DATE}}"
  decision: "ACCEPT|REJECT|MODIFY"
  resolution: "Clarified to use kg/s per ATA-28 convention"
  confidence_after: 0.95

Source-Tracking Requirements

docId Schema

Format: ATA-{CHAPTER}-{DOCTYPE}-{SEQUENCE}

Examples:

ATA-28-REQ-0001 - Requirements document
ATA-71-DESIGN-0023 - Design document
ATA-95-CERT-0005 - Certification evidence

Implementation:

source_tracking:
  doc_id: "ATA-28-REQ-0001"
  doc_title: "H₂ Storage System Requirements"
  doc_version: "1.2"
  doc_date: "2026-01-15"
  doc_owner: "STK_SE"

Span Referencing

Format: docId:{DOC_ID}, span:{START}-{END}[, page:{PAGE}]

Examples:

extracted_content: "The minimum fuel cell operating temperature is -40°C."
source_span: "docId:ATA-71-DESIGN-0012, span:2345-2398, page:14"
confidence: 0.89

Span Types:

Character Offset: span:2345-2398 (byte or character position)
Line Range: lines:45-47 (line numbers)
Section: section:3.2.1 (document structure)
Page: page:14 (PDF page number)

Bidirectional Traceability

All KNUs must maintain:

Upstream Trace: Link to source documents via docId/span
Derivation Trace: Link to parent KNOT via _derivation.yaml
Downstream Trace: Link to CSDB DMs and publications

Example:

knu_id: "KNU-28-10-001-REQ-001"
derivation_lineage:
  parent_knot: "KNOT-28-10-001"
  source_documents:
    - doc_id: "ATA-28-REQ-0001"
      spans: ["span:100-250", "span:500-750"]
      confidence: 0.92
downstream_artifacts:
  - "DMC-AMPEL360-28-10-00-00A-001A-A"
  - "KNU-28-10-001-ICD-001"

Effectivity Rules Configuration

Effectivity Dimensions

Aircraft Serial Number Range: Q100-001 to Q100-999
Configuration: Baseline, Option A (extended range), Option B (cargo)
Modification State: Pre-MOD-XXX, Post-MOD-XXX
Certification Basis: FAA, EASA, CAAC

Effectivity Expression Syntax

applicability:
  aircraft_serial_numbers:
    from: "Q100-001"
    to: "Q100-999"
  
  configurations:
    include: ["BASELINE", "OPTION_A"]
    exclude: ["OPTION_B"]
  
  modifications:
    required: ["MOD-28-001"]  # H₂ tank upgrade
    incompatible: ["MOD-71-005"]  # Legacy fuel system
  
  certification:
    authorities: ["FAA", "EASA"]
    special_conditions: ["SC-28-H2-001"]

Chapter-Specific Effectivity

ATA 28 (Fuel):

effectivity:
  fuel_type:
    - configuration: "BASELINE"
      fuel: "LH2"
      applicability: "ALL"
    - configuration: "HYBRID"
      fuel: "LH2+BATTERY"
      applicability: "Q100-501 onwards"

ATA 71 (Power Plant):

effectivity:
  propulsion_type:
    - type: "FUEL_CELL_PRIMARY"
      serial_range: "Q100-001 to Q100-500"
    - type: "FUEL_CELL_HYBRID"
      serial_range: "Q100-501 onwards"

CSDB Effectivity Integration

Link effectivity to S1000D applicability:

<applic>
  <assert applicPropertyIdent="Q100_CONFIG" applicPropertyType="prodattr" applicPropertyValues="BASELINE OPTION_A"/>
  <assert applicPropertyIdent="Q100_SERIAL" applicPropertyType="prodattr" applicPropertyValues="Q100-001:Q100-999"/>
</applic>

Validation & Compliance

Pre-Deployment Checklist

Note: The following validation scripts are planned for future implementation. For now, perform manual validation of schema compliance, BREX rules, and traceability.

Schema Validation (Future tool: validate_schema.py)

# Manual validation: Use JSON Schema validators for YAML files
# Example: yamllint or python jsonschema library

BREX Validation (Future tool: brex_validator.py)

# Manual validation: Use S1000D BREX checker if available
# s1000d-brex-check --brex DMC-AMPEL360-00-00-00-00A-022A-A --dm DMC-*.XML

Traceability Audit (Future tool: audit_traceability.py)

# Manual validation: Check for broken links in _derivation.yaml files
# Verify parent_knot references exist

Confidence Score Analysis (Future tool: confidence_report.py)

# Manual validation: Review confidence scores in metadata
# Ensure low-confidence items are flagged for HITL review

Continuous Integration Pipeline

# .github/workflows/ata-chapter-validation.yml
name: ATA Chapter Validation
on: [push, pull_request]

jobs:
  validate:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
        uses: actions/checkout@v2
      
      - name: Validate YAML Schemas
        run: python scripts/validate_all_schemas.py
      
      - name: BREX Validation
        run: |
          python scripts/brex_validate.py --chapter ${{ matrix.chapter }}
      
      - name: Traceability Check
        run: python scripts/check_traceability.py
      
      - name: Confidence Report
        run: |
          python scripts/confidence_report.py > confidence_summary.txt
          cat confidence_summary.txt
      
      - name: Upload Results
        uses: actions/upload-artifact@v2
        with:
          name: validation-results
          path: confidence_summary.txt

Common Pitfalls

1. Inventing Data

❌ Wrong: LLM generates part number P/N: FCS-28-1234 when not in source
✅ Right: Mark as null with confidence < 0.60, escalate to HITL

2. Skipping _derivation.yaml

❌ Wrong: Create KNU without derivation metadata
✅ Right: Every KNU must have _derivation.yaml linking to parent KNOT

3. Inconsistent ATA Addressing

❌ Wrong: Mix of ATA-28-10 and ATA-2810 formats, or using formats inappropriately
✅ Right: Use ATA-{CC}-{SS} for chapter-section references (e.g., ATA-28-00). Use full ATA-{CC}-{SS}-{SU}-{SB}-{SX} format only when source documents specify sub-unit/subject granularity. Never use unhyphenated forms.

4. Missing Source Tracking

❌ Wrong: Content without docId/span reference
✅ Right: Every extracted fact has source_span with docId and position

5. Over-Ambitious Auto-Accept

❌ Wrong: Auto-accept all content with confidence > 0.70
✅ Right: Safety-critical content requires HITL regardless of confidence

6. Ignoring BREX Failures

❌ Wrong: "BREX validation failed but content looks OK"
✅ Right: Fix BREX errors before merging; BREX compliance is mandatory

7. Circular Dependencies

❌ Wrong: KNOT-28-10-001 depends on KNOT-28-10-002 which depends on KNOT-28-10-001
✅ Right: Use topological sort to verify dependency DAG

8. Effectivity Conflicts

❌ Wrong: DM applies to both "BASELINE" and "excludes: BASELINE"
✅ Right: Validate effectivity expressions for logical consistency

Support & Resources

Framework Documentation: OPT-IN_FRAMEWORK/README.md
ATA-00-00 Reference: 00-00-general/README.md
Template Library: _templates/
Scaffold Script: scaffold_chapter.py

Revision History

Date	Version	Author	Change
2026-01-16	1.0.0	STK_CM	Initial onboarding playbook

This playbook is a living document. Submit updates via pull request with changelog entry.

FilesExpand file tree

ATA_CHAPTER_ONBOARDING.md

Latest commit

History

ATA_CHAPTER_ONBOARDING.md

File metadata and controls

ATA Chapter Onboarding Playbook

Table of Contents

Overview

End-to-End Process Flow

Prerequisites

Step-by-Step Onboarding Checklist

Phase 1: Scaffold Generation

Phase 2: GENESIS Layer Population

Phase 3: SSOT Layer Configuration

Phase 4: LLM Engine Metadata Layer Integration

Phase 5: CSDB Integration

Phase 6: Publication & Export

Phase 7: Validation & Quality Assurance

Chapter-Specific Customization Guide

High-Priority Chapters

ATA 28 (Fuel) - T-TECHNOLOGIES

ATA 71 (Power Plant) - T-TECHNOLOGIES

ATA 95 (AI/ML) - T-TECHNOLOGIES

ATA 96 (DPP/Traceability) - N-NEURAL_NETWORKS

HITL Escalation Decision Tree

HITL Triggers

Source-Tracking Requirements

docId Schema

Span Referencing

Bidirectional Traceability

Effectivity Rules Configuration

Effectivity Dimensions

Effectivity Expression Syntax

Chapter-Specific Effectivity

CSDB Effectivity Integration

Validation & Compliance

Pre-Deployment Checklist

Continuous Integration Pipeline

Common Pitfalls

1. Inventing Data

2. Skipping _derivation.yaml

3. Inconsistent ATA Addressing

4. Missing Source Tracking

5. Over-Ambitious Auto-Accept

6. Ignoring BREX Failures

7. Circular Dependencies

8. Effectivity Conflicts

Support & Resources