İnsansız Hava Aracı Telemetrisinde Otomatik Makine Öğrenmesi Tabanlı Anomali Tespiti

Öz

İnsansız Hava Araçlarının (İHA) telemetri verilerinin güvenilir şekilde analiz edilmesi, özellikle karmaşık ve riskli ortamlarda görev başarısı ve operasyonel güvenlik açısından kritik öneme sahiptir. Bu veri akışları, sensör arızaları, çevresel etkenler veya siber-fiziksel saldırılar nedeniyle anormallikler içerebilir. Bu nedenle, sağlam bir anomali tespit mekanizması gereklidir. Bu çalışma, yüksek frekanslı İHA telemetrisi için özel olarak tasarlanmış, gözetimsiz bir anomali tespit çerçevesi sunmaktadır. Yaklaşım, alan bilgisine dayalı özellik mühendisliğini, model seçimi ve hiperparametre ayarlarını otomatikleştiren bir AutoML tabanlı optimizasyon süreciyle birleştirir. Sistem; Local Outlier Factor, Isolation Forest, One-Class SVM ve Elliptic Envelope olmak üzere dört farklı gözetimsiz algoritmayı entegre ederek, İHA operasyonlarının dinamik doğasına uyum sağlar. 48 farklı İHA görevinden toplanan 127.000 örnek içeren gerçek dünya veri kümesi üzerinde yapılan değerlendirmelerde, uzmanlar tarafından etiketlenmiş anomali segmentleri yalnızca doğrulama amacıyla kullanılmıştır. En iyi performans, %92 doğruluk, %88 kesinlik, %85 duyarlılık ve %86 F1-skoru ile Local Outlier Factor algoritması tarafından elde edilmiştir. Gerçek zamanlı uygulamalar için ölçeklenebilir ve düşük gecikmeli olarak tasarlanan bu sistem, İHA’larda otomatik arıza izleme ve güvenli, dayanıklı hava araçları ekosistemlerinin gelişimine önemli katkılar sunmaktadır.

Anahtar Kelimeler

• Anomali tespiti
• insansız hava araçları
• otomatik makine öğrenmesi

Anomaly Detection in Unmanned Aerial Vehicle Telemetry Using Automated Machine Learning

Öz

Reliable analysis of UAV telemetry data is critical for mission safety, especially as drones are increasingly deployed in complex and high-risk environments. These data streams often include anomalies arising from sensor faults, environmental disruptions, or cyber-physical attacks, making robust anomaly detection essential. This study introduces an unsupervised anomaly detection framework designed specifically for high-frequency UAV telemetry. It combines domain-driven feature engineering with an AutoML-based optimization pipeline that enables automated model selection and hyperparameter tuning. The framework integrates four unsupervised algorithms—Local Outlier Factor, Isolation Forest, One-Class SVM, and Elliptic Envelope—ensuring adaptability to the dynamic nature of UAV operations. Evaluated on a real-world dataset of 127,000 samples from 48 UAV missions, the system uses expert-labeled anomaly segments solely for validation to preserve the integrity of unsupervised learning. Among all methods, Local Outlier Factor yielded the best results with 0.920 accuracy, 0.880 precision, 0.850 recall, and 0.860 F1-score. Scalable and low-latency, the proposed solution is well-suited for real-time deployment. By bridging theoretical advances with operational needs, this work contributes to safer and more resilient aerial robotic systems.

Anahtar Kelimeler

• Anomaly detection
• unmanned aerial vehicles
• automated machine learning

Kaynakça

1. Idrees R, Maiti A, Garg S. A clustering algorithm for detecting differential deviations in the multivariate time-series IoT data based on sensor relationship. Knowl Inf Syst 2024; 67: 2641-2690.
2. Chen Z, Li Z, Huang J, Liu S, Long H. An effective method for anomaly detection in industrial Internet of Things using XGBoost and LSTM. Sci Rep 2024; 14: 1-23.
3. Canonico R, Esposito G, Navarro A, Romano SP, Sperlí G, Vignali A. An anomaly-based approach for cyber-physical threat detection using network and sensor data. Comput Commun 2025; 234: 1-14.
4. Kuchar K, Fujdiak R. Analyzing anomalies in industrial networks: A data-driven approach to enhance security in manufacturing processes. Comput Secur 2025; 153: 1-15.
5. Kumar D, Agraharam PC, Liu Y, Namilae S. Anomaly detection for composite manufacturing using AI models. J Intell Manuf 2024; 1-17.
6. Chung J, Shen B, Kong ZJ. Anomaly detection in additive manufacturing processes using supervised classification with imbalanced sensor data based on generative adversarial network. J Intell Manuf 2024; 35: 2387-2406.
7. Engbers H, Freitag M. Automated model selection for multivariate anomaly detection in manufacturing systems. J Intell Manuf 2024; 1-19.
8. Sezgin A, Boyacı A. AID4I: An Intrusion Detection Framework for Industrial Internet of Things Using Automated Machine Learning. Comput Mater Continua 2023; 76(2): 2121-2143.

9. Singh A, Rathore H. Advancing connected vehicle security through real-time sensor anomaly detection and recovery. Veh Commun 2025; 52: 1-11.
10. Lu Y, Yang T, Zhao C, Chen W, Zeng R. A swarm anomaly detection model for IoT UAVs based on a multi-modal denoising autoencoder and federated learning. Comput Ind Eng 2024; 196: 1-22.
11. Li T, Lin W, Ma R, Ma Z, Shen Y, Ma J. CoDetect: Cooperative Anomaly Detection with Privacy Protection Towards UAV Swarm. Sci China Inf Sci 2024; 67: 1-2.
12. Alzahrani MY. Enhancing Drone Security Through Multi-Sensor Anomaly Detection and Machine Learning. SN Comput Sci 2024; 5: 1-10.
13. Yang L, Li S, Li C, Zhu C, Zhang A, Liang G. Data-driven unsupervised anomaly detection and recovery of unmanned aerial vehicle flight data based on spatiotemporal correlation. Sci China Technol Sci 2023; 66: 1304-1316.
14. Ozkat EC. Vibration data-driven anomaly detection in UAVs: A deep learning approach. Eng Sci Technol Int J 2024; 54: 1-11.
15. Ahn H, Chung S. Deep learning-based anomaly detection for individual drone vehicles performing swarm missions. Expert Syst Appl 2024; 244: 1-14.
16. Sezgin A. Scenario-Driven Evaluation of Autonomous Agents: Integrating Large Language Model for UAV Mission Reliability. Drones 2025; 9(3): 1-21.
17. Malviya VK, Minn W, Shar LK, Jiang L. Fuzzing drones for anomaly detection: A systematic literature review. Comput Secur 2025; 148: 1-14.
18. Sezgin A, Boyacı A. Securing the Skies: Exploring Privacy and Security Challenges in Internet of Drones. In: 10th Int Conf on Recent Advances in Air and Space Technologies (RAST); 2023; Istanbul, Türkiye.
19. Liu D, Wang N, Guo K, Wang B. Ensemble Transfer Learning Based Cross-Domain UAV Actuator Fault Detection. IEEE Sens J 2023; 23(4): 16363-16372.
20. Yoo JD, Kim GM, Song MG, Kim HK. MeNU: Memorizing normality for UAV anomaly detection with a few sensor values. Comput Secur 2025; 150: 1-15.
21. Wu Y, Liu L, Yu Y, Chen G, Hu J. Online ensemble learning-based anomaly detection for IoT systems. Appl Soft Comput 2025; 173: 1-12.
22. Schaller M, Kruse M, Ortega A, Lindauer M, Rosenhahn B. AutoML for multi-class anomaly compensation of sensor drift. Meas 2025; 250: 1-14.
23. Vajda DL, Do TV, Bérczes T, Farkas K. Machine learning-based real-time anomaly detection using data pre-processing in the telemetry of server farms. Sci Rep 2024; 14: 1-22.
24. Abdullah RY, Posonia AM, Nisha UB. An Enhanced Anomaly Forecasting in Distributed Wireless Sensor Network Using Fuzzy Model. Int J Fuzzy Syst 2022; 24: 3327-3347.
25. Wei X, Xu Y, Zhang H, Sun C, Li X, Huang F, Ma J. Sensor attack online classification for UAVs using machine learning. Comput Secur 2025; 150: 1-18.
26. Sharma T, Balyan A, Singh AK. Machine Learning-Based Energy Optimization and Anomaly Detection for Heterogeneous Wireless Sensor Network. SN Comput Sci 2024; 5: 1-16.