SKYLIUM
EASA Certification Strategy

EASA Certification Roadmap

Our strategic pathway to achieve EASA homologation for AI-powered flight training — DO-178C, Part-FCL, and EU AI Act compliance.

DO-178CPart-FCLARA/ORAEU AI ActEPAS 2026

Regulatory Framework

DO-178C

Software Considerations

Primary standard for airborne software certification. Defines objectives for each DAL level. SKYLIUM targets DAL C/D for training software.

Part-FCL

Flight Crew Licensing

EASA regulation governing pilot training and licensing. Subpart J covers type ratings. SKYLIUM maps to CBTA competency framework.

ARA/ORA

Organisation Requirements

Authority Requirements for Aircrew (ARA) and Organisation Requirements for Aircrew (ORA). Governs ATO approval and oversight.

EU AI Act

Artificial Intelligence Regulation

SKYLIUM is designed to comply with EU AI Act requirements for high-risk AI systems in education and safety-critical domains.

EPAS 2026

European Plan for Aviation Safety

EASA's strategic priorities include digitalisation and AI. SKYLIUM directly addresses EPAS priorities for innovative training solutions.

EASA AI Roadmap

AI/ML Concept Paper

EASA's guidance on certification of AI/ML-based systems. SKYLIUM follows the recommended assurance framework for learning-enabled systems.

Certification Timeline

24-month phased approach from gap analysis to certification decision.

Phase 1

Foundation & Gap Analysis

Q2 – Q3 2026

● In Progress

Key Activities

EASA Innovation Hub application & engagement
DO-178C gap analysis for AI/ML software components
Part-FCL mapping: identify applicable regulations for AI-assisted training
Appoint DER (Designated Engineering Representative) for software
Establish Software Quality Assurance (SQA) processes
Begin PSAC (Plan for Software Aspects of Certification) document

Key Deliverables

PSAC — Plan for Software Aspects of Certification
SDP — Software Development Plan
Gap Analysis Report
Phase 2

DO-178C Compliance Build

Q4 2026 – Q1 2027

○ Upcoming

Key Activities

Define DAL (Design Assurance Level) for each software component
Implement DO-178C lifecycle processes (requirements, design, coding, testing)
AI/ML Supplement: Address DO-178C + ML considerations (EASA Concept Paper)
Develop SVP (Software Verification Plan) with test coverage matrix
Establish traceability: Requirements → Design → Code → Tests
Configuration Management (CM) system with full audit trail

Key Deliverables

SRS — Software Requirements Specification
SVP — Software Verification Plan
SCI — Software Configuration Index
Phase 3

Part-FCL Integration & ATO Framework

Q2 – Q3 2027

○ Upcoming

Key Activities

Map SKYLIUM modules to Part-FCL Subpart J (Type Rating)
Develop competency framework aligned with CBTA (Competency-Based Training)
Prepare ATO (Approved Training Organisation) partnership applications
ARA/ORA compliance assessment for training organization requirements
Submit Part-FCL deviation/exemption request if needed for AI instructor
Pilot programme with 2-3 partner ATOs (FTEJerez, CAE, etc.)

Key Deliverables

CBTA Mapping Document
ATO Partnership Agreements
Part-FCL Compliance Matrix
Phase 4

Certification Audit & Validation

Q4 2027 – Q1 2028

○ Upcoming

Key Activities

Stage 1 audit: Documentation review with NAA (AESA for Spain)
Stage 2 audit: On-site validation of processes and platform
Demonstrate AI explainability & bias mitigation (EU AI Act alignment)
Complete SAS (Software Accomplishment Summary)
Address audit findings and implement corrective actions
Final certification decision by competent authority

Key Deliverables

SAS — Software Accomplishment Summary
Audit Response Documents
Certification Application

Risk Assessment & Mitigation

Proactive risk management for the certification journey.

AI Explainability requirements exceed current capabilities

Prob: MediumImpact: High

→ Implement LIME/SHAP interpretability layers. Maintain human-in-the-loop for all assessment decisions.

DO-178C compliance costs exceed budget

Prob: MediumImpact: High

→ Target DAL D initially (lowest cost). Leverage EASA Innovation Hub for guidance. Phase certification incrementally.

NAA rejects AI as valid training supplement

Prob: LowImpact: Critical

→ Build evidence base with ATO pilot programmes. Demonstrate measurable learning outcomes. Align with CBTA framework.

EU AI Act classification as High-Risk blocks deployment

Prob: LowImpact: High

→ Proactively implement conformity assessment. Register in EU AI database. Ensure full transparency and auditability.

EASA Certification is Our Competitive Moat

Once certified, SKYLIUM becomes the first AI training platform with EASA endorsement — a barrier that competitors cannot easily replicate.

View DocumentationMeet the Team