In this paper, we propose a deep convolutional neural network solution to the analysis of image data for the detection of rail surface defects. The images are obtained from many hours of automated video recordings. This huge amount of data makes it impossible to manually inspect the images and detect rail surface defects. Therefore, automated detection of rail defects can help to save time and costs, and to ensure rail transportation safety. However, one major challenge is that the extraction of suitable features for detection of rail surface defects is a non-trivial and difficult task. Therefore, we propose to use convolutional neural networks as a viable technique for feature learning. Deep convolutional neural networks have recently been applied to a number of similar domains with success. We compare the results of different network architectures characterized by different sizes and activation functions. In this way, we explore the efficiency of the proposed deep convolutional neural network for detection and classification. The experimental results are promising and demonstrate the capability of the proposed approach.Accepted Author Manuscript. Link to published article (IEEE): http://dx.

Railway infrastructure monitoring is a vital task to ensure rail transportation safety. A rail failure could result in not only a considerable impact on train delays and maintenance costs, but also on safety of passengers. In this article, the aim is to assess the risk of a rail failure by analyzing a type of rail surface defect called squats that are detected automatically among the huge number of records from video cameras. We propose an image processing approach for automatic detection of squats, especially severe types that are prone to rail breaks. We measure the visual length of the squats and use them to model the failure risk. For the assessment of the rail failure risk, we estimate the probability of rail failure based on the growth of squats. Moreover, we perform severity and crack growth analyses to consider the impact of rail traffic loads on defects in three different growth scenarios. The failure risk estimations are provided for several samples of squats with different crack growth lengths on a busy rail track of the Dutch railway network. The results illustrate the practicality and efficiency of the proposed approach.

In this paper, a decision support approach is proposed for condition-based maintenance of rails relying on expert-based systems. The methodology takes into account both the actual conditions of the rails (using axle box acceleration measurements and rail video images) and the prior knowledge of the railway track. The approach provides an integrated estimation of the rail health conditions to support the maintenance decisions for a given time period. An expert-based system is defined to analyse interdependency between the prior knowledge of the track (defined by influential factors) and the surface defect measurements over the rail. When the rail health conditions is computed, the different track segments are prioritized, in order to facilitate grinding planning of those segments of rail that are prone to critical conditions. In this paper, real-life rail conditions measurements from the track Amersfoort-Weert in the Dutch railway network are used to show the benefits of the proposed methodology. The results support infrastructure managers to analyse the problems in their rail infrastructure and to efficiently perform a condition-based maintenance decision making. IntroductionThe increase in train traffic and axle loads affect the health conditions of railway infrastructure. Hence, efficient infrastructure monitoring and maintenance is among the major concerns of infrastructure managers in order to improve the performance of railway operations (Åhrén and Parida, 2009). As such, infrastructure health conditions should be monitored and considered in the maintenance decision making process. Effective management of infrastructure health conditions is crucial to guarantee the desired asset quality level (Parida and Chattopadhyay, 2007;Gandomi and Haider, 2015;Zywiel and Oberlechner, 2001). It also plays an important role in meeting the demands for the whole system performance when the infrastructure is upgraded e.g. when increasing traffic capacity, the maintenance regime should be adapted to avoid compromising safety and infrastructure health requirements. To keep the infrastructure system working at an effective level, a conditions-based maintenance system is required not only to consider the actual heath conditions but also evolution during the maintenance decision horizon (Jamshidi et al., 2017b;Li et al., 2014).Condition-based monitoring is used in railway infrastructures to estimate the actual health conditions of the assets, so that degradation processes can be effectively controlled. It helps to keep the infrastructure manager continually informed of the estimated health of the railway infrastructure. Condition-based monitoring is supposed to collect information that will allow an effective https://doi.

Rail defect detection by video cameras has recently gained much attention in both academia and industry. Rail image data has two properties. It is highly imbalanced towards the non-defective class and it has a large number of unlabeled data samples available for semisupervised learning techniques. In this paper we investigate if positive defective candidates selected from the unlabeled data can help improve the balance between the two classes and gain performance on detecting a specific type of defects called Squats. We compare data sampling techniques as well and conclude that the semi-supervised techniques are a reasonable alternative for improving performance on applications such as rail track Squat detection from image data.

No abstract