Lithium-ion (Li-ion) batteries are used in various applications as the rechargeable power sources. The batteries undergo capacity fade during the repeated charge-discharge cycles, which eventually leads to the end of life (EOL). For the purpose of timely replacement before reaching the EOL, reliable prediction of the remaining useful life (RUL) during the cycles is of great importance. However, there may exist unhealthy batteries exhibiting the change of state at some cycles from those of normal degradation, which leads to their EOL sooner than expected. In this article, we propose a novel prognostic method using the particle filter (PF) that is capable of detecting the point of state change and adapting its algorithm to the new battery degradation pattern. The performance of the proposed method is demonstrated by the case study of Li-ion battery degradation data, comparing with the original PF algorithm. As a result, the proposed method shows better performance in terms of anomaly detection of degradation and adaptability to the new degradation process, which leads to more accurate and reliable RUL prediction.

Metal additive manufacturing (AM) has several similarities to conventional metal manufacturing, such as welding and cladding. During the manufacturing process, both metal AM and welding experience repeated partial melting and cooling, referred to as deposition. Owing to deposition, metal AM and welded products often share common product quality issues, such as layer misalignment, dimensional errors, and residual stress generation. This paper comprehensively reviews the similarities in quality monitoring methods between metal AM and conventional metal manufacturing. It was observed that a number of quality monitoring methods applied to metal AM and welding are interrelated; therefore, they can be used complementarily with each other.

Gears and bearings are one of the major components of many machines, which can result in operation downtime or even catastrophic failure of a whole system. This paper addresses a tutorial for the features extraction and selection of the gears and bearings, which is known as feature engineering, a prerequisite step for the prognostics and health management (PHM) of these components. While there have been many new developments in this field, no studies have addressed the tutorial aspects of features engineering to aid engineers in solving problems by their own effort, which is of practical importance for successful PHM. The paper aims at helping beginners learn the basic concepts, and implement the algorithms using the public datasets as well as those made by the authors. Matlab codes are provided for them to implement the process by their own hands.

A fault diagnosis of a train door system is carried out using the motor current signal that operates the door. A test rig is prepared, in which various fault modes are examined by applying extreme conditions, as well as the natural and artificial wears of critical components. Two approaches are undertaken toward the fault classification for comparative purposes: one is the traditional feature-based method that requires several steps for the processing features such as signal segmentation, the extraction of time-domain features, selection by Fisher’s discrimination, and K-nearest neighbor. The other is the deep learning approach by employing the convolutional neural network (CNN) to skip the hand-crafted features extraction process. In the traditional approach, good accuracy is found only after the current signal is segmented into the three velocity regimes, which enhances the discrimination capability. In the CNN, superior accuracy is obtained even by the original raw signal, which is more convenient in terms of implementation. However, in view of practical applications, the traditional approach is more useful in that the features processing can be easily applied to assess the health state of each fault and monitor the progression over time in the real operation, which is not enabled by the deep learning approach.