11 - Operations and Sustainment

ON THE FEASIBILITY OF MACHINE LEARNING-BASED DETECTION OF AIRCRAFT LIGHTNING STRIKE RISK AREAS

E. Yoshikawa¹, S. Horishita², Y. Hamada², T. Ushio¹, V. Chandrasekar, Colorado State University, United States; ¹The University of Osaka, Japan ;²Japan Aerospace Exploration Agency, Japan

Aircraft lightning can disrupt flight operations. We apply a support vector machine to radar-derived features to detect lightning strike risk areas. The proposed method achieves an 86% probability of detection and a 39% false alarm rate, demonstrating the feasibility of machine-learning-based risk detection for operational airline support.

View Paper

DOI: 10.71945/icas2026_0803