Digital Transformation of Aviation Safety and Strategies for Constructing a Safety Intelligence Framework

Abstract

The global aviation system is undergoing a data-centric digital transformation under ICAO’s GANP and ASBU framework. In aviation safety management, ICAO Annex 19 and Doc 9859 provide the foundation for SMS, while Doc 10159 emphasizes data-driven Safety Intelligence (SI). However, essential data related to flight operations, maintenance, training, procedures, and safety events often remain fragmented across organizational silos. This fragmentation limits the transition from reactive safety management to predictive and proactive safety practices. Accordingly, this study proposes an integrated Safety Intelligence framework for the digital transformation of airline safety management. The framework consists of a data processing and integration layer, a Knowledge Graph-based integration structure, a Safety Intelligence layer, and a decision-support layer. Key functions include anomaly detection, risk prediction, NLP and generative AI, system dynamics simulation, and conversational safety dashboards. It also supports personalized safety briefings, predictive maintenance alerts, and LLM-based executive report generation through natural language queries. Sustainable implementation requires data governance, Just Culture, role-based access control, de-identification, explainability, and auditability. This study redefines aviation safety data as a strategic asset for predictive safety management and operational efficiency, while empirical validation using actual airline data remains necessary.