Anna Kowalska

2025-02

Piloting AI-assisted FMEA tool in collaboration with AI team. Evaluated AI-generated failure mode suggestions for turbine disk and HPT blade use cases over 6-week pilot.

Challenge: AI-generated failure modes are predominantly generic (vibration, corrosion, fatigue) and miss aerospace-specific modes related to thermal cycling, tip rub clearance, and foreign object damage. Providing structured feedback to katarzyna-zielinska on domain gaps.

Learned: AI FMEA tools need to be trained on aerospace-specific failure mode databases, not generic manufacturing data. The tool is most useful for the brainstorming phase — reducing blank page problem — not for final FMEA sign-off.

2024-11

Led PPAP Level 3 submission for a new LPT (Low Pressure Turbine) blade supplier. Managed 300-hour endurance test documentation, dimensional measurement reports, and material certifications.

Challenge: Supplier's Cpk values for 2 critical dimensions were 1.32 — below our minimum requirement of 1.67. Negotiated an interim deviation with supplier improvement plan, monthly Cpk reporting, and a 6-month re-submission milestone.

Learned: Interim deviations must include specific, measurable improvement milestones and an automatic escalation trigger if milestones are missed. Vague 'supplier will improve' commitments are unenforceable.

2024-07

Created control plan templates for additive manufacturing QC — CT scan inspection sampling strategy, surface roughness acceptance criteria, and dimensional inspection sequence for complex internal geometries.

Challenge: CT scan defect interpretation is still highly operator-dependent — two operators reviewing the same scan would classify porosity differently. Working on a standardized defect atlas with photographic examples for each defect category.

Learned: For new inspection methods like CT scanning in aerospace, operator qualification and standardized reference materials (defect atlases) are as important as the measurement equipment itself.

2024-02

Conducted Process FMEA for LPBF (Laser Powder Bed Fusion) 3D printing of structural aluminum brackets. New failure modes not covered in standard templates: layer delamination, residual stress cracking, support structure adhesion failure, and porosity distribution.

Challenge: No AS13100 guidance exists specifically for additive manufacturing process FMEA. Developed internal guidelines by combining AM-specific literature with AS13100 process control requirements. Document approved by quality director after 3 review cycles.

Learned: When applying established quality frameworks to emerging manufacturing processes, the framework structure is valid but failure modes, detection methods, and control parameters require complete redefinition.

2023-10

Collaborated with AI team on an automated FMEA suggestion tool. Provided 350 labeled historical failure modes as training data and reviewed AI-generated FMEA candidates for aerospace accuracy.

Challenge: Significant vocabulary mismatch between AI developers and quality engineers. Terms like 'failure mode', 'effect', and 'cause' have precise FMEA meanings that differ from everyday usage. Spent first two weeks on domain knowledge transfer.

Learned: Cross-disciplinary AI projects need a shared glossary agreed upfront. What seems like communication overhead is actually risk reduction — the AI team's early prototypes were targeting the wrong field of the FMEA structure.

2023-06

Coordinated APQP Phase 2 for a new HPT (High Pressure Turbine) component. Synchronized design, manufacturing, and supplier activities to meet the design freeze milestone.

Challenge: Supplier drawings were in 4 inconsistent revision states — different teams were working from different drawing revisions without realizing it. Required a full drawing reconciliation exercise that delayed Phase 2 by 3 weeks.

Learned: A structured drawing change log must be a mandatory APQP Phase 1 deliverable. Revision control issues at the beginning cause cascading quality problems downstream.

2023-02

Led FMEA review for turbine blade casting process redesign. Facilitated cross-functional team review of 47 existing failure modes and identification of 12 new potential failures related to revised cooling channel geometry.

Challenge: Cross-functional alignment was difficult — manufacturing and design teams disagreed on severity ratings for 3 failure modes. One disagreement was escalated to chief engineer level. Structured a risk review meeting with explicit rating criteria that resolved all open items.

Learned: FMEA facilitation is as much conflict resolution as engineering analysis. Pre-agreeing on severity rating criteria before the review session prevents most disagreements from occurring.