Stefan Sikora
Continued Airworthiness Engineer · Certification & Airworthiness
Warsaw, Poland
Continued Airworthiness Engineer responsible for managing the in-service technical health of the certified fleet. Leads Airworthiness Directive impact assessments, develops Service Bulletins for in-service issues, and monitors fleet operational trends using data from MRO reporting systems. Works at the interface between OEM engineering and airline operators to translate field findings into technical solutions.
Expertise
- Airworthiness Directive (AD) impact assessment
- Service Bulletin (SB) technical analysis
- fleet health monitoring and trend analysis
- continued airworthiness maintenance program (CAMP)
- MRO interface and operational experience feedback
Technologies
Work History
2025-02
Predictive maintenance AI pilot — worked with AI team to build an anomaly detection model for early warning of HPT blade deterioration from ACARS EGT margin trends. Provided labeled historical data and validated model outputs.
Challenge: EGT margin deterioration is a gradual process with high variability between flights due to ambient conditions. The AI model confused ambient temperature effects with blade deterioration, producing false positives at summer high-temperature airports. Required explicit ambient temperature normalization before model training.
Learned: Engine health monitoring AI models must normalize all inputs for ambient condition effects before training. EGT margin is a function of ambient temperature — a model trained on unnormalized EGT margin will learn ambient temperature effects as 'blade deterioration' and generate false positives at hot airports.
2024-04
Fleet health monitoring dashboard — built Python-based dashboard integrating ACARS operational data, MRO inspection findings, and AD compliance status for real-time fleet condition overview.
Challenge: ACARS data from 6 different airlines came in 4 different formats with different parameter naming conventions (e.g., N1 vs. Fan_Speed vs. core_speed_n1). Building the normalization layer required coordination with each airline's flight operations engineering team to verify parameter equivalences.
Learned: Fleet health monitoring data integration requires a formal data dictionary agreed with each contributing airline. Informal parameter name matching leads to silent misidentification of equivalent parameters — a normalization layer built without airline confirmation will contain errors that are difficult to detect downstream.
2023-09
Service Bulletin development for in-service combustor liner cracking — engineering investigation of 6 field reports of premature liner cracking, root cause analysis, and Service Bulletin for inspection and repair procedure.
Challenge: Fleet data analysis showed cracking was concentrated in aircraft with >85% flight cycles below 45 minutes — short-haul operators. Thermal fatigue from frequent thermal cycling was the root cause, but the original certification analysis used a mix of flight profiles. Required updating the life limit for short-haul flight profiles.
Learned: In-service distress driven by operational profile mismatch requires fleet segmentation in the data analysis. Do not aggregate all fleet cycles — segment by short-haul, medium-haul, and long-haul before looking for operational trends. The failure rate for short-haul may be completely different from the fleet average.
2023-03
Airworthiness Directive impact assessment for EASA AD 2023-0041 (HPT stage 1 blade life limit reduction) — assessed impact on 47 in-service aircraft, computed mandatory inspection intervals, and coordinated parts availability with supply chain.
Challenge: AD compliance timing for 47 aircraft depended on individual fleet cycle counts — 12 aircraft needed immediate inspection within 30 days. Supply chain had only 8 spare blade sets available. Coordinated a prioritized maintenance schedule with 6 airlines to sequence inspections within the parts availability constraint.
Learned: AD impact assessments for large fleets must be automated — manual fleet compliance tracking across 47 aircraft with individual cycle counts is error-prone. Python-based fleet compliance tracker using SAP PM data eliminates calculation errors and enables real-time reprioritization when conditions change.