1. Introduction

2. Related Work

2.1. Artificial Intelligence (AI) in Modern Military Applications

• Applications in Safety Management

In the realm of safety management, Generative AI (GAI) plays a crucial role, especially in aviation safety prediction. By "learning and understanding" vast amounts of aviation safety reports, GAI can develop a "Civil Aviation Complex System Safety Model," deducing and simulating aviation safety laws to achieve early warnings and predictions of unsafe events influenced by factors such as "human, aircraft, environment, and management." Additionally, GAI can accurately simulate and reproduce safety incidents that are challenging to observe and test in real scenarios, aiding in accident investigation and review analysis.

Table 1. Application Data Example.

Table 1. Application Data Example.

Application Area	Specific Function	Practical Case
Safety Prediction	Establish complex safety models, deduce safety laws	GAI model successfully predicted potential safety hazards for a flight
Accident Simulation	Reproduce difficult-to-observe safety incidents, assist investigations	GAI simulated the entire process of an aviation accident, helping identify causes

2.

Applications in Flight Operations

In-flight operations, GAI is expected to be deeply integrated with the cockpit, becoming an ideal flight assistant in real flights or training. On one hand, GAI can help pilots and crew quickly obtain the aircraft's status, provide critical route, weather, and other information, and assist in making timely and accurate decisions. Liu et al. (2024) present a study focused on predicting flight accidents using a Back Propagation Neural Network (BPNN). Their research involves developing a predictive model that analyzes historical flight accident data to identify key factors contributing to accidents and predict potential risk scenarios.

The BPNN-based model is designed to learn from various input features, including flight parameters, operational conditions, and past incident records, to generate risk assessments and identify high-risk situations. This approach aims to enhance flight safety by providing actionable insights and early warnings based on predictive analytics. The study demonstrates the efficacy of neural network models in improving aviation safety management and highlights the potential for integrating advanced machine learning techniques into accident prediction and risk mitigation strategies. On the other hand, GAI's "generative" advantage can be utilized to create a rich and vivid flight training scene library in virtual reality, offering near-real or customized interactive scenes and feedback information to enhance training efficiency.

Table 2. Application Data Example.

Table 2. Application Data Example.

Application Area	Specific Function	Practical Case
Flight Assistance	Provide aircraft status, assist in decision-making	GAI helped pilots make quick decisions during simulated flights
Virtual Training	Create interactive training scenes to improve efficiency	GAI-based virtual training system significantly improved training outcomes

3.

Applications in Aircraft Maintenance, Repair, and Overhaul (MRO)

In MRO, GAI can act as both an "intelligent manual" and a "maintenance expert." Leveraging fleet big data, GAI can predict faults and provide corresponding maintenance measures, advancing the current maintenance concept from "preventive maintenance" to "predictive maintenance," thereby improving safety and efficiency. Leading predictive maintenance platforms like Lufthansa's Aviatar and Air France Engineering's PROGNOS are exploring GAI technology to further enhance their predictive capabilities and safety.

Table 3. Application Data Example.

Table 3. Application Data Example.

Application Area	Specific Function	Practical Case
Fault Prediction	Predict faults and provide maintenance advice	GAI predicted and preemptively addressed potential faults in an aircraft
Maintenance Optimization	Shift from preventive to predictive maintenance	PROGNOS platform used GAI to significantly enhance maintenance efficiency

4.

Applications in Production and Operations

In production and operations, GAI can enhance airline operational efficiency in various dimensions, such as improving crew scheduling efficiency and enhancing passenger service levels. Compared to flow-based operation planning software, GAI-based crew scheduling programs can intelligently analyze rules and operational manuals, achieving a better balance between airline operations, regulatory constraints, and human resource efficiency. Additionally, GAI helps passengers with ticket booking, simple boarding procedures, and more, reducing customer service resource investment for airlines.

Many large airlines both domestically and internationally have begun implementing relevant technologies.

Table 4. Application Data Example.

Table 4. Application Data Example.

Application Area	Specific Function	Practical Case
Crew Scheduling	Intelligent rule analysis for efficient scheduling	GAI technology optimized crew scheduling processes for an airline
Customer Service	Handle ticket booking and boarding, reduce service resource investment	GAI improved customer service efficiency in large airlines

The new generation of generative AI is set to profoundly enhance the level of intelligence in various fields of civil aviation, bringing new opportunities for its digital transformation. To fully leverage this new quality of productivity, especially in the civil aviation sector, attention should be given to three key issues: first, studying the algorithmic model mechanisms of GAI to improve model explainability and break through the "black box" of deep learning; second, addressing the balance between "artificial" and "intelligent" elements, building an ideal "human-machine interaction" paradigm to achieve the best balance between human intervention and assisted decision-making; and third, standardizing the development and application of GAI in terms of network security, data security, and ethical security, to build a reliable and controllable intelligent civil aviation system to ensure aviation safety.

2.2. AI-Assisted Fatigue Monitoring-Inertial Measurement Unit (IMU)

The Inertial Measurement Unit (IMU) is an advanced sensor technology that tracks and records an athlete's movement in real time, including data such as acceleration, rotation, and direction. When this rich data is met with AI, it can reveal deep patterns and trends in the athlete's training process.

Figure 1. Inertial Measurement Unit (IMU) sensor monitoring framework.

Figure 1. Inertial Measurement Unit (IMU) sensor monitoring framework.

In the Biro team's study, they used IMUs to collect dynamic data on 19 athletes during multiple training sessions. Then, by applying machine learning algorithms - including random forests, gradient hoists, and Long short-term memory networks (LSTM) - they developed models capable of predicting an athlete's fatigue level and endurance. Researchers have cleverly used the advantages of IMU devices in data acquisition: high spatio-temporal resolution: millisecond high frequency sampling, which can accurately reflect instantaneous motion changes; Multi-structure: synchronously record a variety of data such as triaxial acceleration and triaxial angular velocity, presenting a complete motion picture; No constraint: small size, wireless, wearable, does not affect the natural movement of athletes.

Figure 2. Sensor monitoring formula away.

Figure 2. Sensor monitoring formula away.

The Inertial Measurement Unit (IMU) sensor technology offers valuable contributions to pilot fatigue risk assessment by providing precise, real-time data on movement dynamics. By integrating IMUs with facial recognition and physiological signal analysis, the combined approach enhances the understanding of pilot fatigue. The IMU’s high spatio-temporal resolution allows for the detailed tracking of minute motion changes, which, when analyzed alongside facial expressions and physiological signals, can deliver a comprehensive picture of fatigue levels and movement patterns. This synergy aids in creating more accurate predictive models for fatigue, as it captures both the physiological and behavioral aspects of pilot performance. In study Liu et al. (2024) explore the application of machine learning techniques for predicting dangerous flight weather conditions.

The researchers propose a novel approach that leverages various machine learning algorithms to enhance the accuracy of weather forecasts critical for flight safety. Their model integrates historical weather data, real-time atmospheric measurements, and advanced learning algorithms to identify patterns and predict adverse weather events that could impact flight operations. This research highlights the potential of machine learning to improve flight safety by providing timely and precise weather predictions, thereby enabling better pre-flight planning and in-flight decision-making. The findings underscore the importance of incorporating cutting-edge technology into aviation safety protocols, offering a promising direction for future advancements in meteorological forecasting for aviation.

Additionally, the IMU’s ability to record multi-structural data, such as triaxial acceleration and angular velocity, complements facial recognition systems that monitor expressions and physiological signals indicating fatigue. The sensor’s small, wireless, and wearable design ensures that it does not interfere with natural movement, making it ideal for real-world applications where pilots' movements and conditions need to be continuously monitored without disrupting their duties. This integration allows for a more holistic and unobtrusive assessment of fatigue, leading to better preventive measures and improved flight safety.

3. Methodology

3.2. Study and Its Implications

This study introduces a novel training method for athletes that leverages AI and IMU technology to monitor fatigue and endurance levels in real time and adjust training plans accordingly. By comparing various machine learning models, the researchers identified a highly accurate prediction method crucial for preventing overtraining and enhancing performance.

Figure 3. Sports safety approach on specific use case.

Figure 3. Sports safety approach on specific use case.

The research not only highlights the potential of AI in sports science but also outlines future technological advancements and underscores the need for ethical and privacy considerations in technology applications. This article opens new perspectives on AI’s role in sports training and offers profound insights into how technology can elevate human performance. We look forward to seeing how AI continues to transform the sports world!

3.4. Model Performance Insights

The K-Nearest Neighbor Classifier and Gradient Boosting Classifier showed lower overall performance with accuracies of 48.65% and 47.13%, respectively, highlighting potential issues in fitting the data complexity. Decision Tree Classifier performed competitively with an accuracy of 50.66% and an F1 score of 51.15%, suggesting robustness in this task but not necessarily the most effective for fatigue prediction. Logistic Regression and Linear Discriminant Analysis** both showed high recall (over 63%) but lower precision, reflecting their capability to identify true positives effectively, albeit at the cost of increased false positives.

The Support Vector Machine (Linear Kernel) achieved the highest F1 score of 54.39% due to its high recall of 77.91%, though with lower precision, indicating a preference for identifying true positives over minimizing false positives. The Naive Bayes classifier also performed well in terms of F1 score (52.09%), suggesting it is effective at detecting true positives but may generate more false positives. Notably, the Dummy Classifier showed an impressive F1 score of 67.78% due to its unrealistic perfect recall, which is not useful for practical predictions. These results underscore the need for careful model selection and tuning to balance precision, recall, and overall performance in fatigue risk assessment, ensuring accurate and actionable insights for pilot safety management.

Figure 4. Participant 23 non-fatigue (NF) and fatigiie (1y) states. Details of more participants: Séedata availability statement.

Figure 4. Participant 23 non-fatigue (NF) and fatigiie (1y) states. Details of more participants: Séedata availability statement.

4. Conclusions

References

1. Chen, Y., Yan, S., Liu, S., Li, Y., & Xiao, Y. (2024, August). EmotionQueen: A Benchmark for Evaluating Empathy of Large Language Models. In Findings of the Association for Computational Linguistics ACL 2024 (pp. 2149-2176).
2. Chen, Z., Ge, J., Zhan, H., Huang, S., & Wang, D. (2021). Pareto self-supervised training for few-shot learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 13663-13672).
3. Feng, Z., Ge, M., & Meng, Q. (2024). Enhancing Energy Efficiency in Green Buildings Through Artificial Intelligence. [CrossRef]
4. Feng, Z., Ge, M., Meng, Q., & Chen, Y. (2024). Research on Old Building Renovation Strategies by using Green Building Technologies. [CrossRef]
5. Ge, M., Feng, Z., & Meng, Q. (2024). Urban Planning and Green Building Technologies Based on Artificial Intelligence: Principles, Applications, and Global Case Study Analysis. [CrossRef]
6. Huang, D., Liu, Z., & Li, Y. (2024). Research on Tumors Segmentation based on Image Enhancement Method. arXiv preprint arXiv:2406.05170. [CrossRef]
7. Huang, D., Xu, L., Tao, W., & Li, Y. (2024). Research on Genome Data Recognition and Analysis based on Louvain Algorithm. [CrossRef]
8. Jiang, G., Zhao, S., Yang, H., & Zhang, K. (2024). Research on Finance Risk Management based on Combination Optimization and Reinforcement Learning. [CrossRef]
9. Jin, Y., Shimizu, S., Li, Y., Yao, Y., Liu, X., Si, H., ... & Xiao, W. (2023). Proton therapy (PT) combined with concurrent chemotherapy for locally advanced non-small cell lung cancer with negative driver genes. Radiation Oncology, 18(1), 189. [CrossRef]
10. Kumada, H., Li, Y., Yasuoka, K., Naito, F., Kurihara, T., Sugimura, T., ... & Sakae, T. (2022). Current development status of iBNCT001, demonstrator of a LINAC-based neutron source for BNCT. Journal of Neutron Research, 24(3-4), 347-358. [CrossRef]
11. Lai, S., Feng, N., Sui, H., Ma, Z., Wang, H., Song, Z., ... & Yue, Y. (2024). FTS: A Framework to Find a Faithful TimeSieve. arXiv preprint arXiv:2405.19647. [CrossRef]
12. Li, B., Jiang, G., Li, N., & Song, C. (2024). Research on Large-scale Structured and Unstructured Data Processing based on Large Language Model. [CrossRef]
13. Li, B., Zhang, K., Sun, Y., & Zou, J. (2024). Research on Travel Route Planning Optimization based on Large Language Model. [CrossRef]
14. Li, B., Zhang, X., Wang, X. A., Yong, S., Zhang, J., & Huang, J. (2019, April). A Feature Extraction Method for Daily-periodic Time Series Based on AETA Electromagnetic Disturbance Data. In Proceedings of the 2019 4th International Conference on Mathematics and Artificial Intelligence (pp. 215-219).
15. Li, S., & Tajbakhsh, N. (2023). Scigraphqa: A large-scale synthetic multi-turn question-answering dataset for scientific graphs. arXiv preprint arXiv:2308.03349. [CrossRef]
16. Li, S., Lin, R., & Pei, S. (2024). Multi-modal preference alignment remedies regression of visual instruction tuning on language model. arXiv preprint arXiv:2402.10884. [CrossRef]
17. Li, Y., Matsumoto, Y., Chen, L., Sugawara, Y., Oe, E., Fujisawa, N., ... & Sakurai, H. (2023). Smart Nanofiber Mesh with Locally Sustained Drug Release Enabled Synergistic Combination Therapy for Glioblastoma. Nanomaterials, 13(3), 414. [CrossRef]
18. Li, Y., Mizumoto, M., Oshiro, Y., Nitta, H., Saito, T., Iizumi, T., ... & Sakurai, H. (2023). A retrospective study of renal growth changes after proton beam therapy for Pediatric malignant tumor. Current Oncology, 30(2), 1560-1570. [CrossRef]
19. Liu, H., Shen, F., Qin, H., & Gao, F. (2024). Research on Flight Accidents Prediction based Back Propagation Neural Network. arXiv preprint arXiv:2406.13954. [CrossRef]
20. Liu, H., Xie, R., Qin, H., & Li, Y. (2024). Research on Dangerous Flight Weather Prediction based on Machine Learning. arXiv preprint arXiv:2406.12298.
21. Luo, H., Li, Y., Song, H., Zhao, K., Li, W., Hong, H., ... & Zhang, Y. (2024). Role of EZH2-mediated epigenetic modification on vascular smooth muscle in cardiovascular diseases: A mini-review. Frontiers in Pharmacology, 15, 1416992. [CrossRef]
22. Meng, Q., Ge, M., & Feng, Z. (2024). The Integration of Artificial Intelligence in Architectural Visualization Enhances Augmented Realism and Interactivity. [CrossRef]
23. Nakamura, M., Mizumoto, M., Saito, T., Shimizu, S., Li, Y., Oshiro, Y., ... & Sakurai, H. (2024). A systematic review and meta-analysis of radiotherapy and particle beam therapy for skull base chondrosarcoma: TRP-chondrosarcoma 2024. Frontiers in Oncology, 14, 1380716. [CrossRef]
24. Nitta, H., Mizumoto, M., Li, Y., Oshiro, Y., Fukushima, H., Suzuki, R., ... & Sakurai, H. (2024). An analysis of muscle growth after proton beam therapy for pediatric cancer. Journal of Radiation Research, 65(2), 251-255. [CrossRef]
25. Peng, Z. H. A. N. G., Bin-MeiZi, Z. H. A. N. G., Jian-Ke, Z. O. U., & Xiang-Beng, L. I. U. (2018). The Methods, Effects and Mechanism of Priming Attachment Security towards Social Behaviors. Journal of Psychological Science, (3), 615.
26. Qu, M. (2024). High Precision Measurement Technology of Geometric Parameters Based on Binocular Stereo Vision Application and Development Prospect of The System in Metrology and Detection. Journal of Computer Technology and Applied Mathematics, 1(3), 23-29.
27. Restrepo, D., Wu, C., Cajas, S. A., Nakayama, L. F., Celi, L. A. G., & Lopez, D. M. (2024). Multimodal Deep Learning for Low-Resource Settings: A Vector Embedding Alignment Approach for Healthcare Applications. medRxiv, 2024-06. [CrossRef]
28. Shimin, L. E. I., Ke, X. U., Huang, Y., & Xinye, S. H. A. (2020). An Xgboost based system for financial fraud detection. In E3S Web of Conferences (Vol. 214, p. 02042). EDP Sciences.
29. Shimizu, S., Mizumoto, M., Okumura, T., Li, Y., Baba, K., Murakami, M., ... & Sakurai, H. (2021). Proton beam therapy for a giant hepatic hemangioma: A case report and literature review. Clinical and Translational Radiation Oncology, 27, 152-156. [CrossRef]
30. Shimizu, S., Nakai, K., Li, Y., Mizumoto, M., Kumada, H., Ishikawa, E., ... & Sakurai, H. (2023). Boron neutron capture therapy for recurrent glioblastoma multiforme: imaging evaluation of a case with long-term local control and survival. Cureus, 15(1). [CrossRef]
31. Shui, H., Sha, X., Chen, B., & Wu, J. (2024, May). Stock weighted average price prediction based on feature engineering and Lightgbm model. In Proceedings of the 2024 International Conference on Digital Society and Artificial Intelligence (pp. 336-340).
32. Wang, C., Chen, J., Xie, Z., & Zou, J. (2024). Research on Education Big Data for Students Academic Performance Analysis based on Machine Learning. arXiv preprint arXiv:2407.16907. arXiv:2407.16907. [CrossRef]