What is it about?
This study presents an Artificial Intelligence tool that automatically analyzes aviation safety reports using Natural Language Processing and a Bidirectional Encoder Representations from Transformers (BERT) model. The developed framework classifies aviation occurrences into the standardized safety categories defined by the European Union Aviation Safety Agency (EASA), enabling fast and consistent safety assessment of abnormal aviation events. The model achieves a beyond-state-of-the-art F1 score of 94% while reducing analysis time from months to mere seconds. By enabling near real-time identification of safety risks, the proposed framework supports proactive aviation safety management and helps prevent future accidents and incidents.
Featured Image
Photo by Bing Hui Yau on Unsplash
Why is it important?
The developed framework expedites the aviation safety risk mitigation process from months (or even years) to seconds, thereby proactively avoiding aviation accidents.
Perspectives
Artificial Inteligence as a tool to enhance aviation safety!
Charilaos Kakoulidis
Aristotle University of Thessaloniki
Read the Original
This page is a summary of: Enhancing Aviation Safety: Natural Language Processing for Automated Classification of Aviation Occurrences, Journal of Air Transportation, September 2026, American Institute of Aeronautics and Astronautics (AIAA),
DOI: 10.2514/1.d0636.
You can read the full text:
Contributors
The following have contributed to this page
