WCSE

A Multi-level Heart Rate Fatigue Recognition Framework Based on CNN-LSTM for Air Traffic Controllers

Abstract— Fatigue is an important factor of air accidents caused by air traffic controllers. The conventional fatigue features extraction and recognition algorithms focus on individual behavioral and perceptual data, while ignoring the physiological conditions. Furthermore, the state-of-the-art heart rate detection method didn’t consider the issue of timing correlation. Therefore, this paper proposes a multi-level heart rate fatigue recognition model based on CNN-LSTM, which realizes fatigue recognition from static and dynamic features respectively. It achieved a high accuracy in recognizing air traffic controllers’ fatigue in complex environmental backgrounds, which are up to 95.5%. The experimental results demonstrated that the feature of heart rate is closely related to fatigue.

Index Terms— heart rate, fatigue feature extraction, LSTM, CNN, air traffic controller

Kai Wang, Junqi Bai
State Key Laboratory of Air Traffic Management System, CHINA
Xueyan Li, Yifan Xu, Zhiyuan Shen
College of Civil Aviation, Nanjing University of Aeronautics and Astronautics, CHINA

[Download]

Cite: Kai Wang, Junqi Bai, Xueyan Li, Yifan Xu, Zhiyuan Shen, "A Multi-level Heart Rate Fatigue Recognition Framework Based on CNN-LSTM for Air Traffic Controllers", 2025 the 15th International Workshop on Computer Science and Engineering (WCSE 2025), pp. 10-14, Jeju Island, South Korea, June 277-282, 2025.