- Akıllı Ulaşım Sistemleri ve Uygulamaları Dergisi
- Volume:7 Issue:2
- Critical Connections: Enhancing Aviation Safety Through Network Analysis
Critical Connections: Enhancing Aviation Safety Through Network Analysis
Authors : Ayşe Aslı Yılmaz
Pages : 263-285
Doi:10.51513/jitsa.1524747
View : 71 | Download : 68
Publication Date : 2024-10-22
Article Type : Research Paper
Abstract :The objective of this study is to explore the influence of human factors on aviation accidents across different aircraft types using Network Analysis. This research specifically analyses centrality metrics to pinpoint the most influential errors, providing a more detailed understanding of their impact on aviation safety, while prior studies have acknowledged the role of human factors. The study employed Network Analysis using Python\'s NetworkX library. A dataset derived from the NTSB aviation accident database is used. Bipartite networks were constructed for each aircraft type, with nodes representing accidents and human error categories and edges indicating their relationships. The analysis revealed that pilot error consistently showed the highest degree centrality across all aircraft types, indicating its frequent involvement in accidents. Closeness centrality further highlighted the central role of pilot error, showing its significant influence on the network. The findings emphasized the critical importance of addressing pilot errors to improve aviation safety. While crew errors and other human errors were less frequent, their presence in the data suggested that a comprehensive safety strategy must also consider these factors. The results of this study demonstrate that pilot error is the most influential human factor in aviation accidents across various aircraft types. By focusing on targeted interventions such as enhanced pilot training and stricter safety protocols, the aviation industry can significantly reduce accident rates.Keywords : Havacılık Emniyeti, İnsan Faktörleri, Network Analizi, Merkezilik Ölçüleri
ORIGINAL ARTICLE URL
