Published Paper Details:

Aviation safety is a critical concern due to the complex and high-risk nature of flight operations. Although aviation technology has advanced significantly, accident prevention systems are still largely reactive, relying on post-incident investigations. With the availability of large volumes of aviation and environmental data, artificial intelligence techniques can be utilized to predict potential risks before flight operations. This paper presents an AI-based flight accident risk prediction system using a One-Dimensional Convolutional Neural Network (1D-CNN). The proposed system analyzes multiple factors such as weather conditions, aircraft characteristics, flight duration, and pilot experience to estimate accident risk levels. Data preprocessing techniques including normalization and categorical encoding are applied to improve model performance. The trained CNN model effectively captures complex non-linear relationships among features and classifies flights into high-risk and low-risk categories with associated probability scores. The system is deployed through a Flask-based web application that allows real-time and manual risk prediction. Experimental results indicate that the proposed approach provides reliable predictive performance and supports proactive aviation safety management.