YOLOv11-based Detection of Wagon Brake Cylinder Conditions

Abstract

Railway transportation stands out as a safe and efficient mode of transport for both freight and passengers. However, failures in train braking systems pose financial and safety risks. In this study, it is proposed to use the recently introduced YOLOv11 (You Only Look Once) models to monitor the mechanical brakes used in wagons. This approach aims to prevent the locking of wheels due to stuck mechanical brakes while the train is in motion, thereby avoiding continuous metal friction and mitigating risks such as Flatted wheels, wheel fractures, rail damage, and fire hazards. Such failures not only cause material damage and operational disruptions but also lead to potential loss of life and costly accidents. Traditional methods of manually inspecting brake cylinders provide limited safety and are inefficient in terms of operational effectiveness. Therefore, the automatic monitoring and fault detection of brake cylinders have become crucial. To achieve this, a dataset consisting of three different classes—braked, empty, and evacuated—was used. Using this dataset, YOLOv11n, YOLOv11s, YOLOv11m, YOLOv11l, and YOLOv11x models were trained. The performance of these trained models was evaluated based on accuracy, precision, recall, and F1 scores. The results indicate that the YOLOv11X model is more suitable for cases where reducing false negatives (FN) is critical. However, when minimizing false positives (FP) is a priority, YOLOv11m or YOLOv11s models are more appropriate. For an overall balanced performance, the YOLOv11X model is preferable for the braked condition, while YOLOv11s or YOLOv11m models are more suitable for the evacuated condition. Ultimately, this study demonstrates that the detection of braking mechanisms in trains with high accuracy using YOLOv11 models can significantly contribute to reducing train accidents, thereby preventing loss of life and costly incidents.

Keywords

• YoloV11
• Railway Safety
• Brake Cylinder Detection
• Workplace Safety

YOLOv11-based Detection of Wagon Brake Cylinder Conditions

Abstract

Railway transportation stands out as a safe and efficient mode of transport for both freight and passengers. However, failures in train braking systems pose financial and safety risks. In this study, it is proposed to use the recently introduced YOLOv11 (You Only Look Once) models to monitor the mechanical brakes used in wagons. This approach aims to prevent the locking of wheels due to stuck mechanical brakes while the train is in motion, thereby avoiding continuous metal friction and mitigating risks such as Flatted wheels, wheel fractures, rail damage, and fire hazards. Such failures not only cause material damage and operational disruptions but also lead to potential loss of life and costly accidents. Traditional methods of manually inspecting brake cylinders provide limited safety and are inefficient in terms of operational effectiveness. Therefore, the automatic monitoring and fault detection of brake cylinders have become crucial. To achieve this, a dataset consisting of three different classes—braked, empty, and evacuated—was used. Using this dataset, YOLOv11n, YOLOv11s, YOLOv11m, YOLOv11l, and YOLOv11x models were trained. The performance of these trained models was evaluated based on accuracy, precision, recall, and F1 scores. The results indicate that the YOLOv11X model is more suitable for cases where reducing false negatives (FN) is critical. However, when minimizing false positives (FP) is a priority, YOLOv11m or YOLOv11s models are more appropriate. For an overall balanced performance, the YOLOv11X model is preferable for the braked condition, while YOLOv11s or YOLOv11m models are more suitable for the evacuated condition. Ultimately, this study demonstrates that the detection of braking mechanisms in trains with high accuracy using YOLOv11 models can significantly contribute to reducing train accidents, thereby preventing loss of life and costly incidents

Keywords

• yolov11
• Railway Safety
• Brake Cylinder Detection
• Workplace Safety

References

1. A. Ghosh. (2024). Yolov11 overview. Https:// Learnopencv.Com/Yolo11/.
2. Akhmedov, F. , ., Nasimov, R. , & Abdusalomov, A. (2024). Dehazing Algorithm Integration with YOLO-v10 for Ship Fire Detection. Fire, 7(9), 332.
3. Alif, M. A. R., & Hussain, M. (2024). YOLOv1 to YOLOv10: A comprehensive review of YOLO variants and their application in the agricultural domain. ArXiv Preprint ArXiv:2406.10139.
4. Brintha, K., & Joseph Jawhar, S. (2024). FOD-YOLO NET: Fasteners fault and object detection in railway tracks using deep yolo network . Journal of Intelligent & Fuzzy Systems, 46(4), 8123–8137.
5. Çak, R. , A. S. , & Çelebi, M. (2002). Demiryollari İle Yolcu Taşlmacillği Ve Yolcu Vagonu Onarimi. Sakarya University Journal of Science, 6–1.
6. Chen, R., Lin, Y., & Jin, T. (2022). High-speed railway pantograph-catenary anomaly detection method based on depth vision neural network. IEEE Transactions on Instrumentation and Measurement, 71, 1–10.
7. Cimen, M., Boyraz, O., Yildiz, M., & Boz, A. (2021). A new dorsal hand vein authentication system based on fractal dimension box counting method. Optik, 226.
8. Çimen, M. E. (2024). Comparison of Deep Learning and Yolov8 Models for Fox Detection Around the Henhouse. Journal of Smart Systems Research, 5(2), 76–90.

9. Çimen, M. E. , Boyraz, Ö. F. , Garip, Z. , Pehlivan, İ. , Yıldız, M. Z. , & Boz, A. F. (2021). Görüntü işleme tabanlı kutu sayma yöntemi ile fraktal boyut hesabı için arayüz tasarımı. Politeknik Dergisi, 24(3), 867–878. https://doi.org/10.2339/politeknik.
10. Çimen, M. E., Garip, Z. B., Boyraz, Ö. F., Pehlivan, İ., Yıldız, M. Y., & Boz, A. F. (2020). An Interface Design For Calculation Of Fractal Dimension. 2019, 3–9.
11. Çimen, M. E., Garip, Z., Pala, M. A., Boz, A. F., & Akgül, A. (2019). Modelling Of A Chaotic System Motion In Video With Artificial Neural Networks. Chaos Theory And Applications, 1(1).
12. C.-Y.Wang, A., Bochkovskiy, & H.-Y. M. Liao. (2021). Scaled-Yolov4: Scaling Cross Stage Partial Network. In Proceedings Of The Ieee/Cvf Conference On Computer Vision And Pattern Recognition, Pages, 13029–13038. Demiryolu Sektör Raporu. (2023).
13. Ding, L., Tong, Y., Cui, Y., & Sun, Z. (2025). Improved Yolov10 Lightweight Bearing Surface Defect Detection Algorithm. Preprints.Org.
14. Ghahremani, A. , Adams, S. D. , Norton, M. , Khoo, S. Y., & Kouzani, A. Z. (2025). Detecting Defects In Solar Panels Using The Yolo V10 And V11 Algorithms. Electronics, 14(2), 344.
15. Ghahremani, A., Adams, S. D., Norton, M., Khoo, S. Y., & Kouzani, A. Z. (2025). Detecting Defects In Solar Panels Using The Yolo V10 And V11 Algorithms. Electronics, 14(2), 344.
16. Guarnido-Lopez, P., Ramirez-Agudelo, J. F., Denimal, E., & Benaouda, M. (2024). Programming And Setting Up The Object Detection Algorithm Yolo To Determine Feeding Activities Of Beef Cattle: A Comparison Between Yolov8m And Yolov10m. Animals, 14(19), 2821.
17. K. He, X. Zhang, S. Ren, & J. Sun. (2015). Spatial Pyramid Pooling In Deep Convolutional Networks For Visual Recognition. Ieee Transactions On Pattern Analysis And Machine Intelligence, 37(9), 1904–1916.
18. Khanam, R., Asghar, T., & Hussain, M. (2025). Comparative Performance Evaluation Of Yolov5, Yolov8, And Yolov11 For Solar Panel Defect Detection. Solar, 5(1).
19. Kishore, P. V. V., & Prasad, C. R. (2017). Computer vision based train rolling stock examination. Optik, 132, 427–444.
20. Krummenacher, G., Ong, C. S., Koller, S., Kobayashi, S., & Buhmann, J. M. (2017). Wheel defect detection with machine learning. IEEE Transactions on Intelligent Transportation Systems, 19(4), 1176–1187.
21. Lisanti, G., Karaman, S., Pezzatini, D., & Bimbo, A. D. (2018). A multi-camera image processing and visualization system for train safety assessment. Multimedia Tools and Applications, 77, 1583–1604.
22. Liu, X., Wang, X., Quan, W., Gu, G., Xu, X., & Gao, S. (2024). A pantograph-catenary arcing detection model for high-speed railway based on semantic segmentation and generative adversarial network. International Journal of Rail Transportation, 1–22.
23. Liu, Y., Lu, B., Peng, J., & Zhang, Z. (2020). Research on the use of YOLOv5 object detection algorithm in mask wearing recognition. World Sci. Res. J, 6(11), 276–284.
24. Mehta, P. , Vaghela, R., Pansuriya, N., Sarda, J., Bhatt, N., Bhoi, A. K., & Srinivasu, P. N. (2025). Benchmarking YOLO Variants for Enhanced Blood Cell Detection. International Journal of Imaging Systems and Technology, 35(1), 70037.
25. Minguell, M. G., & Pandit, R. (2023). TrackSafe: A comparative study of data-driven techniques for automated railway track fault detection using image datasets. Engineering Applications of Artificial Intelligence, 125, 106622.
26. N. Jegham, C. Y. Koh, M. Abdelatti, & A. Hendawi. (2024). Evaluating the evolution of yolo (you only look once) models: A comprehensive benchmark study of yolo11 and its predecessors. ArXiv Preprint ArXiv:2411.00201.
27. Olorunshola, O. E., Irhebhude, M. E., & Evwiekpaefe, A. E. (2023). A comparative study of YOLOv5 and YOLOv7 object detection algorithms. Journal of Computing and Social Informatics, 2(1), 1–12.
28. Öztürk, G., & Eldoğan, O. (2024). Prediction of Multivariate Chaotic Time Series using GRU, LSTM and RNN. Sakarya University Journal of Computer and Information Sciences, 7(2), 156–172.
29. Öztürk, G., Eldoğan, O., & Köker, R. (2024). Computer Vision-Based Lane Detection and Detection of Vehicle, Traffic Sign, Pedestrian Using YOLOv5. Sakarya University Journal of Science, 28(2), 418–430.
30. Pala, M., Cimen, M., Yildız, M., Cetinel, G., Avcıoglu, E., & Alaca, Y. (2022). CNN-Based Approach for Overlapping Erythrocyte Counting and Cell Type Classification in Peripheral Blood Images. Chaos Theory and Applications, 4(2).
31. Pala, M., Cimen, M., Yıldız, M., Eskiler, G., & Özkan, A. (2021). Holografik görüntülerde kenar tabanlı fraktal özniteliklerin hücre canlılık analizlerinde başarısı. Journal of Smart Systems Research, 2(2), 89–94.
32. Rakshit, U., Malakar, B., & Roy, B. K. (2018). Study on longitudinal forces of a freight train for different types of wagon connectors. IFAC-PapersOnLine, 283–288.
33. Rasheed, A. F., & Zarkoosh, M. (2024). YOLOv11 Optimization for Efficient Resource Utilization. ArXiv Preprint ArXiv:2412.14790. https://doi.org/10.48550/arXiv.2412.14790 Focus to learn more
34. Saini, A., Singh, D., & Alvarez, M. (2024). FishTwoMask R-CNN: Two-stage Mask R-CNN approach for detection of fishplates in high-altitude railroad track drone images. Multimedia Tools and Applications, 83(4), 10367–10392.
35. Sapkota, R. , Meng, Z. , Churuvija, M. , Du, X. , Ma, Z., & Karkee, M. (2024). Comprehensive performance evaluation of yolo11, yolov10, yolov9 and yolov8 on detecting and counting fruitlet in complex orchard environments. ArXiv Preprint ArXiv:2407.12040.
36. Sapkota, R., & Karkee, M. (2025). Improved YOLOv12 with LLM-Generated Synthetic Data for Enhanced Apple Detection and Benchmarking Against YOLOv11 and YOLOv10. ArXiv Preprint ArXiv:2503.00057.
37. Sasikala, N., Kishore, P. V. V., Kumar, D. A., & Prasad, C. R. (2019). Localized region based active contours with a weakly supervised shape image for inhomogeneous video segmentation of train bogie parts in building an automated train rolling examination. Multimedia Tools and Applications, 78, 14917–14946.
38. Savran, M., & Bulut, H. (2024). Real-Time Error Detection In Digital Games Based on the YOLO V10 Model. 2024 8th International Artificial Intelligence and Data Processing Symposium (IDAP).
39. Shang, L., Yang, Q., Wang, J., Li, S., & Lei, W. (2018). Detection of rail surface defects based on CNN image recognition and classification. 2018 20th International Conference on Advanced Communication Technology (ICACT), 45–51.
40. Sun, X., Gu, J., Huang, R., Zou, R., & Giron Palomares, B. (2019). Surface Defects Recognition of Wheel Hub Based on Improved Faster R-CNN. Electronics, 8(5), 481.
41. Tian, S., Lu, Y. , Jiang, F. , Zhan, C. , & Huang, C. (202 C.E.). Improved Campus Vehicle Detection Method Based on YOLOv11 and Grayscale Projection-Based Electronic Image Stabilization Algorithm. Traitement Du Signal, 41(6), 3335.
42. Ultralytics YOLO Dokümanlar. (2025, March 3). Https://Docs.Ultralytics.Com/Tr/Models/Yolo11/.
43. Wei, W., Huang, Y., Zheng, J., Rao, Y., Wei, Y., Tan, X., & OuYang, H. (2025). YOLOv11-based multi-task learning for enhanced bone fracture detection and classification in X-ray images. Journal of Radiation Research and Applied Sciences, 18(1), 101309.
44. Wei, X., Yang, Z., Liu, Y., Wei, D., Jia, L., & Li, Y. (2019). Railway track fastener defect detection based on image processing and deep learning techniques: A comparative study. Engineering Applications of Artificial Intelligence, 80, 66–81.
45. Yan, Y., Liu, H., Gan, L., & Zhu, R. (2025). A novel arc detection and identification method in pantograph-catenary system based on deep learning. Scientific Reports, 15(1), 3511.
46. Yang, Y., Liu, Z. ., Chen, J., Gao, H., & Wang, T. (2025). Railway foreign object intrusion detection Using UAV images and YOLO-UAT. IEEE Access.
47. Yang, Z., Lan, X., & Wang, H. (2025). Comparative Analysis of YOLO Series Algorithms for UAV-Based Highway Distress Inspection: Performance and Application Insights. Sensors, 25(5), 1475.
48. Yıldırım, B., & Cagıl, G. (2020). Bir Montaj Parçasının Derin Öğrenme ve Görüntü İşleme ile Tespiti. Journal of Intelligent Systems: Theory and Applications, 3(2), 31–37.
49. Yorgun, H. (1989). Demiryolu taşıtlarında fren sistemlerinin incelenmesi. Anadolu Universitesi.
50. Yu, C., & Lu, Z. (2024). YOLO-VSI: An Improved YOLOv8 Model for Detecting Railway Turnouts Defects in Complex Environments. Computers, Materials & Continua, 81(2).
51. Zhang, Z., Chen, P., Huang, Y., Dai, L., Xu, F., & Hu, H. (2024). Railway obstacle intrusion warning mechanism integrating YOLO-based detection and risk assessment. Journal of Industrial Information Integration, 38, 100571.