LSTM-Autoencoder Based Anomaly Detection Using Vibration Data of Wind Turbines.

The problem of energy depletion has brought wind energy under consideration to replace oil- or chemical-based energy. However, the breakdown of wind turbines is a major concern. Accordingly, unsupervised learning was performed using the vibration signal of a wind power generator to achieve an outlier detection performance of 97%. We analyzed the vibration data through wavelet packet conversion and identified a specific frequency band that showed a large difference between the normal and abnormal data. To emphasize these specific frequency bands, high-pass filters were applied to maximize the difference. Subsequently, the dimensions of the data were reduced through principal component analysis, giving unique characteristics to the data preprocessing process. Normal data collected from a wind farm located in northern Sweden was first preprocessed and trained using a long short-term memory (LSTM) autoencoder to perform outlier detection. The LSTM Autoencoder is a model specialized for time-series data that learns the patterns of normal data and detects other data as outliers. Therefore, we propose a method for outlier detection through data preprocessing and unsupervised learning, utilizing the vibration signals from wind generators. This will facilitate the quick and accurate detection of wind power generator failures and provide alternatives to the problem of energy depletion.

YOLOv5 based object detection in reel package X-ray images of semiconductor component.

The industrial manufacturing landscape is currently shifting toward the incorporation of technologies based on artificial intelligence (AI). This transition includes an evolution toward smart factory infrastructure, with a specific focus on AI-driven strategies in production and quality control. Specifically, AI-empowered computer vision has emerged as a potent tool that offers a departure from extant rule-based systems and provides enhanced operational efficiency at manufacturing sites. As the manufacturing sector embraces this new paradigm, the impetus to integrate AI-integrated manufacturing is evident. Within this framework, one salient application is AI deep learning-facilitated small-object detection, which is poised to have extensive implications for diverse industrial applications. This study describes an optimized iteration of the YOLOv5 model, which is known for its efficacious single-stage object-detection abilities underpinned by PyTorch. Our proposed "improved model" incorporates an additional layer to the model's canonical three-layer architecture, augmenting accuracy and computational expediency. Empirical evaluations using semiconductor X-ray imagery reveal the model's superior performance metrics. Given the intricate specifications of surface-mount technologies, which are characterized by a plethora of micro-scale components, our model makes a seminal contribution to real-time, in-line production assessments. Quantitative analyses show that our improved model attained a mean average precision of 0.622, surpassing YOLOv5's 0.349, and a marked accuracy enhancement of 0.865, which is a significant improvement on YOLOv5's 0.552. These findings bolster the model's robustness and potential applicability, particularly in discerning objects at reel granularities during real-time inferencing.

Federated Learning for Predictive Maintenance and Anomaly Detection Using Time Series Data Distribution Shifts in Manufacturing Processes.

In the manufacturing process, equipment failure is directly related to productivity, so predictive maintenance plays a very important role. Industrial parks are distributed, and data heterogeneity exists among heterogeneous equipment, which makes predictive maintenance of equipment challenging. In this paper, we propose two main techniques to enable effective predictive maintenance in this environment. We propose a 1DCNN-Bilstm model for time series anomaly detection and predictive maintenance of manufacturing processes. The model combines a 1D convolutional neural network (1DCNN) and a bidirectional LSTM (Bilstm), which is effective in extracting features from time series data and detecting anomalies. In this paper, we combine a federated learning framework with these models to consider the distributional shifts of time series data and perform anomaly detection and predictive maintenance based on them. In this paper, we utilize the pump dataset to evaluate the performance of the combination of several federated learning frameworks and time series anomaly detection models. Experimental results show that the proposed framework achieves a test accuracy of 97.2%, which shows its potential to be utilized for real-world predictive maintenance in the future.

Velocity range-based reward shaping technique for effective map-less navigation with LiDAR sensor and deep reinforcement learning.

In recent years, sensor components similar to human sensory functions have been rapidly developed in the hardware field, enabling the acquisition of information at a level beyond that of humans, and in the software field, artificial intelligence technology has been utilized to enable cognitive abilities and decision-making such as prediction, analysis, and judgment. These changes are being utilized in various industries and fields. In particular, new hardware and software technologies are being rapidly applied to robotics products, showing a level of performance and completeness that was previously unimaginable. In this paper, we researched the topic of establishing an optimal path plan for autonomous driving using LiDAR sensors and deep reinforcement learning in a workplace without map and grid coordinates for mobile robots, which are widely used in logistics and manufacturing sites. For this purpose, we reviewed the hardware configuration of mobile robots capable of autonomous driving, checked the characteristics of the main core sensors, and investigated the core technologies of autonomous driving. In addition, we reviewed the appropriate deep reinforcement learning algorithm to realize the autonomous driving of mobile robots, defined a deep neural network for autonomous driving data conversion, and defined a reward function for path planning. The contents investigated in this paper were built into a simulation environment to verify the autonomous path planning through experiment, and an additional reward technique "Velocity Range-based Evaluation Method" was proposed for further improvement of performance indicators required in the real field, and the effectiveness was verified. The simulation environment and detailed results of experiments are described in this paper, and it is expected as guidance and reference research for applying these technologies in the field.

Few-Shot Learning-Based Light-Weight WDCNN Model for Bearing Fault Diagnosis in Siamese Network.

In this study, bearing fault diagnosis is performed with a small amount of data through few-shot learning. Recently, a fault diagnosis method based on deep learning has achieved promising results. Most studies required numerous training samples for fault diagnosis. However, at manufacturing sites, it is impossible to have enough training samples to represent all fault types under all operating conditions. In addition, most studies consider only accuracy, and models are complex and computationally expensive. Research that only considers accuracy is inefficient since manufacturing sites change rapidly. Therefore, in this study, we propose a few-shot learning model that can effectively learn with small data. In addition, a Depthwise Separable Convolution layer that can effectively reduce parameters is used together. In order to find an efficient model, the optimal hyperparameters were found by adjusting the number of blocks and hyperparameters, and by using a Depthwise Separable Convolution layer for the optimal hyperparameters, it showed higher accuracy and fewer parameters than the existing model.

A Novel Approach to the Job Shop Scheduling Problem Based on the Deep Q-Network in a Cooperative Multi-Access Edge Computing Ecosystem.

In this study, based on multi-access edge computing (MEC), we provided the possibility of cooperating manufacturing processes. We tried to solve the job shop scheduling problem by applying DQN (deep Q-network), a reinforcement learning model, to this method. Here, to alleviate the overload of computing resources, an efficient DQN was used for the experiments using transfer learning data. Additionally, we conducted scheduling studies in the edge computing ecosystem of our manufacturing processes without the help of cloud centers. Cloud computing, an environment in which scheduling processing is performed, has issues sensitive to the manufacturing process in general, such as security issues and communication delay time, and research is being conducted in various fields, such as the introduction of an edge computing system that can replace them. We proposed a method of independently performing scheduling at the edge of the network through cooperative scheduling between edge devices within a multi-access edge computing structure. The proposed framework was evaluated, analyzed, and compared with existing frameworks in terms of providing solutions and services.

Real-Time Monitoring for Hydraulic States Based on Convolutional Bidirectional LSTM with Attention Mechanism.

By monitoring a hydraulic system using artificial intelligence, we can detect anomalous data in a manufacturing workshop. In addition, by analyzing the anomalous data, we can diagnose faults and prevent failures. However, artificial intelligence, especially deep learning, needs to learn much data, and it is often difficult to get enough data at the real manufacturing site. In this paper, we apply augmentation to increase the amount of data. In addition, we propose real-time monitoring based on a deep-learning model that uses convergence of a convolutional neural network (CNN), a bidirectional long short-term memory network (BiLSTM), and an attention mechanism. CNN extracts features from input data, and BiLSTM learns feature information. The learned information is then fed to the sigmoid classifier to find out if it is normal or abnormal. Experimental results show that the proposed model works better than other deep-learning models, such as CNN or long short-term memory (LSTM).

VODCA: Verification of Diagnosis Using CAM-Based Approach for Explainable Process Monitoring.

Process monitoring at industrial sites contributes to system stability by detecting and diagnosing unexpected changes in a system. Today, as the infrastructure of industrial sites is advancing because of the development of communication technology, vast amounts of data are generated, and the importance of a way to effectively monitor such data in order to diagnose a system is increasing daily. Because a method based on a deep neural network can effectively extract information from a large amount of data, methods have been proposed to monitor processes using such networks to detect system faults and abnormalities. Neural-network-based process monitoring is effective in detecting faults, but has difficulty in diagnosing because of the limitations of the black-box model. Therefore, in this paper we propose a process-monitoring framework that can detect and diagnose faults. The proposed method uses a class activation map that results from diagnosis of faults and abnormalities, and verifies the diagnosis by post-processing the class activation map. This improves the detection of faults and abnormalities and generates a class activation map that provides a more verified diagnosis to the end user. In order to evaluate the performance of the proposed method, we did a simulation using publicly available industrial motor datasets. In addition, after establishing a system that can apply the proposed method to actual manufacturing companies that produce sapphire nozzles, we carried out a case study on whether fault detection and diagnosis were possible.

HACCP-based Cooperative Model for Smart Factory in South Korea.

Smart Factory has already become an irresistible entity in manufacturing, which has led to a much wider scope of existing manufacturing innovations and the achievement of qualitative improvements. Unsurpassed progress has been made in the manufacturing sector through the combination of professional know-how and advanced IT, and HACCP has already thoroughly managed all steps from raw materials for agricultural and livestock products to processing, packaging and distribution before digitalization. Sanitary issues in each country have now been raised to the level of national security under the influence of COVID-19. Although HACCP is already well-managed by country, we have looked at this discussion and detailed technological transformation from a digital data perspective, as it is now possible to collect, store, record and report data to government offices more smartly in line with technological advances in smart factories. The intersection of smart factories and HACCPs is consistent in terms of data collection, storage and utilization. Furthermore, the addition of Blockchain technology to strictly prevent data forgery is more interesting. Although the use of Blockchain in general factories has been relatively insignificant, it is expected that the use of Blockchain technology will be expanded through smart HACCP, which forms the interface between smart factory technology and HACCP, and that agricultural, livestock and household factories will introduce more smart factories.

Performance Analysis of Inter-Domain Handoff Scheme Based on Virtual Layer in PMIPv6 Networks for IP-Based Internet of Things.

Lately, we see that Internet of things (IoT) is introduced in medical services for global connection among patients, sensors, and all nearby things. The principal purpose of this global connection is to provide context awareness for the purpose of bringing convenience to a patient's life and more effectively implementing clinical processes. In health care, monitoring of biosignals of a patient has to be continuously performed while the patient moves inside and outside the hospital. Also, to monitor the accurate location and biosignals of the patient, appropriate mobility management is necessary to maintain connection between the patient and the hospital network. In this paper, a binding update scheme on PMIPv6, which reduces signal traffic during location updates by Virtual LMA (VLMA) on the top original Local Mobility Anchor (LMA) Domain, is proposed to reduce the total cost. If a Mobile Node (MN) moves to a Mobile Access Gateway (MAG)-located boundary of an adjacent LMA domain, the MN changes itself into a virtual mode, and this movement will be assumed to be a part of the VLMA domain. In the proposed scheme, MAGs eliminate global binding updates for MNs between LMA domains and significantly reduce the packet loss and latency by eliminating the handoff between LMAs. In conclusion, the performance analysis results show that the proposed scheme improves performance significantly versus PMIPv6 and HMIPv6 in terms of the binding update rate per user and average handoff latency.

Cost-effective handoff scheme based on mobility-aware dual pointer forwarding in proxy mobile IPv6 networks.

In this paper, a mobility-aware Dual Pointer Forwarding scheme (mDPF) is applied in Proxy Mobile IPv6 (PMIPv6) networks. The movement of a Mobile Node (MN) is classified as intra-domain and inter-domain handoff. When the MN moves, this scheme can reduce the high signaling overhead for intra-handoff/inter-handoff, because the Local Mobility Anchor (LMA) and Mobile Access Gateway (MAG) are connected by pointer chains. In other words, a handoff is aware of low mobility between the previously attached MAG (pMAG) and newly attached MAG (nMAG), and another handoff between the previously attached LMA (pLMA) and newly attached LMA (nLMA) is aware of high mobility. Based on these mobility-aware binding updates, the overhead of the packet delivery can be reduced. Also, we analyse the binding update cost and packet delivery cost for route optimization, based on the mathematical analytic model. Analytical results show that our mDPF outperforms the PMIPv6 and the other pointer forwarding schemes, in terms of reducing the total cost of signaling.

Direct photofabrication of focal-length-controlled microlens array using photoinduced migration mechanisms of photosensitive sol-gel hybrid materials.

Photosensitive sol-gel hybrid (SGH) materials exhibited the peculiar photoinduced migration behavior of unreacted molecules from unexposed areas to exposed areas by selective UV exposure. Using the photoinduced migration mechanism of the photosensitive SGH materials, the microlens array (MLA) with a smooth surface was directly photofabricated, and the focal length was controlled by changing the photoinduced migration parameters. The higher photoactive monomer content and the thicker film creating a higher curvature produced a smaller focal length of the MLA. Thus, a simple fabrication and easy control of the focal length can be applicable to a fabrication of an efficient MLA.