The rail fastening system forms an indispensable part of the rail tracks and needs to be periodically inspected to ensure safe, reliable and sustainable rail operations. Automated visual inspection has gained significant importance for fastener inspection in recent years. Position accuracy, robustness, and practical limitations due to the complex environment are some of the major concerns associated with this method. This study investigates the combined use of image processing and deep learning algorithms for detecting missing clamps within a rail fastening system. The images used for this study was acquired during field inspections carried out along the Borlänge-Avesta line in Sweden. The image processing techniques proposed in this study enabled the improvement of the fastener position and removal of redundant information from the fastener images. In addition, image augmentation was carried out to enhance the data set, ensure experimental reliability and replicate practical challenges associated with such visual inspection. Convolutional neural network and ResNet-50 algorithms are used for classification purposes, and both the algorithms achieved over 98% accuracy during training and validation and over 94% accuracy during the test stage. Both the algorithms also maintained a good balance between the precision and recall scores during the test stage. CNN and ResNet-50 algorithms were also tested to analyse their performances when the clamp areas were covered. CNN was able to accurately predict the fastener state up to 70% of clamp area occlusion, and ResNet-50 was able to achieve accurate predictions up to 75% of clamp area occlusion.

The rail fastening system forms an integral part of rail tracks, as it maintains the rail in a fixed position, upholding the track stability and track gauge. Hence, it becomes necessary to monitor their conditions periodically to ensure safe and reliable operation of the railway. Inspection is normally carried out manually by trained operators or by employing 2-D visual inspection methods. However, these methods have drawbacks when visibility is minimal and are found to be expensive and time consuming. In the previous study, the authors proposed a train-based differential eddy current sensor system that uses the principle of electromagnetic induction for inspecting the railway fastening system that can overcome the above-mentioned challenges. The sensor system includes two individual differential eddy current sensors with a driving field frequency of 18 kHz and 27 kHz respectively. This study analyses the performance of a machine learning algorithm for detecting and analysing missing clamps within the fastening system, measured using a train-based differential eddy current sensor. The data required for the study was collected from field measurements carried out along a heavy haul railway line in the north of Sweden, using the train-based differential eddy current sensor system. Six classification algorithms are tested in this study and the best performing model achieved a precision and recall of 96.64% and 95.52% respectively. The results from the study shows that the performance of the machine learning algorithms improved when features from both the driving channels were used simultaneously to represent the fasteners. The best performing algorithm also maintained a good balance between the precision and recall scores during the test stage.

The rail fastening system plays a crucial role in railway tracks as it ensures operational safety by fixing the rail on to the sleeper. Early detection of rail fastener system defects is crucial to ensure track safety and to enable maintenance optimization. Fastener inspections are normally conducted either manually by trained maintenance personnel or by using automated 2-D visual inspection methods. Such methods have drawbacks when visibility is limited, and they are also found to be expensive in terms of system maintenance cost and track possession time. In a previous study, the authors proposed a train-based differential eddy current sensor system based on the principle of electromagnetic induction for fastener inspection that could overcome the challenges mentioned above. The detection in the previous study was carried out with the aid of a supervised machine learning algorithm. This study reports the finding of a case study, along a heavy haul line in the north of Sweden, using the same eddy current sensor system mounted on an in-service freight train. In this study, unsupervised machine learning models for detecting and analyzing missing clamps in a fastener system were developed. The differential eddy current measurement system was set to use a driving field frequency of 27 kHz. An anomaly detection model combining isolation forest (IF) and connectivity-based outlier factor (COF) was implemented to detect anomalies from fastener inspection measurements. To group the anomalies into meaningful clusters and to detect missing clamps within the fastening system, an unsupervised clustering based on the DBSCAN algorithm was also implemented. The models were verified by measuring a section of the track for which the track conditions were known. The proposed anomaly detection model had a detection accuracy of 96.79% and also exhibited a high score of sensitivity and specificity. The DBSCAN model was successful in clustering missing clamps, both one and two missing clamps, from a fastening system separately.

Railway switches and crossings (S&Cs) are critical, high-value assets in railway networks. A single failure of such an asset could result in severe network disturbance and considerable economical losses. Squats are common rail surface defects of S&Cs and need to be detected and estimated at an early stage to minimise maintenance costs and increase the reliability of S&Cs. For practicality, installation of wired or wireless sensors along the S&C may not be reliable due to the risk of damages of power and signal cables or sensors. To cope with these issues, this study presents a method for collecting and processing vibration data from an accelerometer installed at the point machine to extract features related to the squat defects of the S&C. An unsupervised anomaly-detection method using the isolation forest algorithm is applied to generate anomaly scores from the features. Important features are ranked and selected. This paper describes the procedure of parameter tuning and presents the achieved anomaly scores. The results show that the proposed method is effective and that the generated anomaly scores indicate the health status of an S&C regarding squat defects.

Railway switches and crossings (S&C) are among the most important high-value components in a railway network and a failure of such an asset could result in severe network disturbance. Therefore, potential defects need to be detected at an early stage to prevent traffic-disturbing downtime or even severe accidents. A squat is a common defect of S&Cs that has to be monitored and repaired to reduce such risks. In this study, a testbed including a full-scale S&C and a bogie wagon was developed. Vibrations were measured for different squat sizes by an accelerometer mounted at the point machine. A method of processing the vibration data and the speed data is proposed to investigate the possibility of detecting and quantifying the severity of a squat. One key technology used is wavelet denoising. The study shows that it is possible to monitor the development of the squat size on the rail up to around 13 m from the point machine. The relationships between the normalised peak-to-peak amplitude of the vibration signal and the squat depth were also estimated.

A Bayesian hierarchical model (BHM) is developed to predict bearing life using envelope acceleration data in combination with a degradation model and prior knowledge of the bearing rating life. The BHM enables the inference of individual bearings, groups of bearings, or bearings operating under certain conditions. The key benefit of the BHM approach is that the relationships between the bearing model parameters and their prior distributions can be expressed at different hierarchical levels. We begin our analysis using a bearing rating life calculation L10h and an estimate of its associated failure time distribution. Realistic variations to constrain our prior distribution of the failure time are then applied before measurements are available. When data become available, estimates more representative of our specific batch and operating conditions are inferred, both on the individual bearing level and the bearing group level. The proposed prognostics methodology can be used in situations with varying amounts of data. The presented BHM approach can also be used to predict the remaining useful life (RUL) of bearings both in situations in which the bearing is considered to be in a healthy state and in situations after a defect has been detected.