Physics-informed replicas of aircraft systems - fusing MATLAB Simscape multidomain physics with deep learning to model oleo-pneumatic dynamics, classify faults across 11–12 severity states, predict RUL, and generate ATA-compliant maintenance directives via embedded on-device LLM. No cloud required.
Hyperledger Fabric blockchain architecture securing ADS-B, ACARS, ADS-C, CPDLC and operational telemetry. Ed25519 cryptographic signing at ingestion layer with semantic attack detection - validated against ICAO Doc 10037 and RTCA DO-326A.
End-to-end pipelines: FlightGear UDP → Telegraf → InfluxDB → Grafana. Field experience at NCAA (airworthiness regulatory) and Executive Airlift (line maintenance, King Air 350). AI-driven MRO inventory forecasting - Cranfield University collaboration.
Systems traceable to CS-23.473/479, CS-25, and EASA AMC guidelines. Fault signatures mapped to ATA chapters, maintenance outputs formatted for direct MRO workflow use - not just academic prototypes. Nigeria Civil Aviation Regulations field experience.