In this study, we developed a deep learning-based real-time fault diagnosis system to automate the weaving preparation process in textile manufacturing. By analyzing typical faults such as shaft eccentricity and rotational imbalance, we designed a data-driven fault diagnosis algorithm. We utilized tension data from both normal and faulty states to implement AI-based diagnostic models, including 1D CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), and LSTM-AE (Long Short-Term Memory Autoencoder). These models enable real-time fault classification, followed by a comparative performance analysis. The LSTM-AE model achieved the best performance, with a classification accuracy of 99-100% for severe faults, such as 1.5 mm eccentricity and 100 or 150 g rotation imbalance, and 92.2% for minor faults like 1 mm eccentricity. This accuracy was optimized through threshold adjustments based on ROC curve analysis to select an optimal threshold. Building on these findings, we developed a GUI (Graphical User Interface) system capable of real- time fault diagnosis using TCP/IP (Transmission Control Protocol/Internet Protocol) communication between Python and LabVIEW. The results of this study are expected to accelerate the smartization of the weaving preparation process, contributing to improved textile quality and reduced defect rates, while also serving as a model for automation in other sectors.

Microfluidics allows for precise manipulation of small volumes of analytical solutions in diverse applications, including disease diagnostics, drug efficacy testing, chemical analysis, and water quality monitoring. Among these diverse applications, one of the most critical aspects is the precise and programmable control of flow within microfluidic control devices. However, microfluidic experiments that employ pressure control via a gas tank may encounter restricted mobility. To address these challenges, we developed an air pump feedback control system utilizing artificial intelligence image analysis and devised a method to enhance portability. In this paper, we utilized a commercially available portable pump to achieve the desired pressure and subsequently cease operation. In addressing the challenge of sustaining prolonged pressure, we implemented a strategy wherein the dimensions of the pressure vessel were modified, accompanied by iterative pump activations, thereby ensuring the sustained maintenance of pressure over time. The evaluation of the flow controller developed in this study involves conducting a comparative flow analysis with established pneumatic flow controllers. Furthermore, we employed artificial intelligence image analysis methods to automate the operation of iterative pumps. In conclusion, we anticipate that the developed portable microfluidic control device will lead to innovative advancements in modern technology and healthcare through its potential applications.

In the semiconductor manufacturing industry, efficient operation of wafer transfer robots has a direct impact on productivity and product quality. Ball screw misalignment anomalies are a critical factor affecting precision transport of robots. Early diagnosis of these anomalies is essential to maintaining system efficiency. This study proposed a method to effectively diagnose ball screw misalignment anomalies using 1D-CNN and 2D-CNN models. This method mainly uses binary classification to distinguish between normal and abnormal states. Additionally, explainable artificial intelligence (XAI) technology was applied to interpret diagnostic decisions of the two deep learning models, allowing users to convince prediction results of the AI model. This study was based on data collected through acceleration sensors and torque sensors. It compared accuracies of 1D-CNN and 2D-CNN models. It presents a method to explain the model"s predictions through XAI. Experimental results showed that the proposed method could diagnose ball screw misalignment anomalies with high accuracy. This is expected to contribute to the establishment of reliable abnormality diagnosis and preventive maintenance strategies in industrial sites.

Deep reinforcement learning (RL) has attracted research interest in the manufacturing area in recent years, but real implemented applications are rarely found. This is because agents have to explore the given environments many times until they learn how to maximize the rewards for actions, which they provide to the environments. While training, random actions or exploration from agents may be disastrous in many real-world applications, and thus, people usually use computer generated simulation environments to train agents. In this paper, we present a RL experiment applied to temperature control of a chamber for ultra-precision machines. The RL agent was built in Python and PyTorch framework using a Deep Q-Network (DQN) algorithm and its action commands were sent to National Instruments (NI) hardware, which ran C codes with a sampling rate of 1 Hz. For communication between the agent and the NI data acquisition unit, a data pipeline was constructed from the subprocess module and Popen class. The agent was forced to learn temperature control while reducing the energy consumption through a reward function, which considers both temperature bounds and energy savings. Effectiveness of the RL approach to a multi-objective temperature control problem was demonstrated in this research.

The tower crane is widely used in construction and transportation engineering. To improve working efficiency and safety, input shaping methods have been applied. Input shaping is a method of reducing residual vibration of flexible systems by convolving a sequence of impulses with unit step command. However, input shaping is based on the linear system theory in which its control performances are degraded, in case of nonlinearity and unmatched dynamics of the control systems. In this paper, a new optimal reference input shape design method based on minimizing cost function is suggested and applied, to a simple cart-pendulum system which is a simplified model of tower cranes. Since pendulum dynamics is nonlinear, analytic solution does not exist. To overcome this problem, in this paper, a machine learning approach is suggested to find optimal reference input shape for the cart position control. The feasibility of the proposed design method is verified through some simulation examples by using MatLab.