As the demand for precision in the manufacturing industry grows, Digital Twin (DT) technology is gaining attention for its potential to enhance equipment performance and process reliability. However, existing research has primarily focused on specific stages of design or operation, leaving a gap in the literature concerning DT models that can be utilized throughout the entire equipment lifecycle. To address this gap, this study proposes a method for developing a DT that employs a consistent Finite Element (FE) model across all phases of the equipment lifecycle. We utilized actual measurement data to ensure high fidelity in the FE model of previous-generation equipment, which we refer to as the Pre-DT. This Pre-DT was instrumental in improving design during the new equipment development phase. Additionally, the DT model was implemented to predict equipment status in real time using the Reduced-Order Model (ROM) method, functioning as a virtual sensor during operation. This approach was applied to the equipment development process, aligned with the asset lifecycle concept of RAMI 4.0, and was tested on an MLCC cutting equipment to validate its effectiveness.

This study proposes a systematic data preprocessing algorithm tailored for AI-based modeling of manufacturing data from a roll-to-roll (R2R) lithium iron phosphate (LFP) battery electrode coating process. The preprocessing strategy specifically addresses process characteristics and spatiotemporal inconsistencies in sensor data, significantly improving data quality for machine learning applications. Utilizing the refined dataset, machine learning models were created to predict coating-related characteristics, resulting in high explanatory power and low prediction errors. This framework effectively illustrates the potential of data-driven modeling for reliable predictions and quantitative analysis of coating uniformity in battery manufacturing.

This paper presents an advanced robotic automation framework that combines an impedance-based compliance controller with an imitation learning network for high-precision peg-in-hole assembly. The framework is characterized by three key features. First, it employs an impedance-based compliance controller to ensure stable contact. This approach enables the robot to adapt flexibly to external contact forces, functioning like a spring-damper system to prevent potential damage. Second, domain randomization is applied to both geometric and visual properties within a high-fidelity simulation environment. This strategy effectively narrows the reality gap, enhancing robustness against environmental uncertainties and visual disturbances. Third, the framework utilizes an action-chunking-transformer (ACT) network to predict precise action sequences based on multimodal data, reducing compounding errors in trajectory generation and improving assembly success rates. Each feature is supported by specific advancements, such as real-time force feedback integration, diverse simulation scenario generation, and multimodal sensor fusion. Extensive experiments conducted in various unseen environments demonstrate the framework's effectiveness, confirming its suitability for complex assembly tasks that require high adaptability and precision under diverse conditions.

The increasing adoption of industrial robot arms in advanced manufacturing has heightened the need for flexible trajectory planning methods that go beyond traditional offline programming (OLP) tools, which are often expensive, proprietary, and limiting. This study introduces an OLP-free pipeline designed to generate robot trajectory data and optimize paths for six-degree-of-freedom (6-DOF) robot arms using discrete reinforcement learning. Initially, five-axis NC code derived from CAD/CAM data is transformed into tool center point (TCP) trajectories through coordinate transformations. An analytical inverse kinematics solver then produces multiple joint solutions for each TCP pose, creating a discrete action space from which the learning agent can select feasible joint configurations along the trajectory. A reward function that considers variations in joint velocity and acceleration, as well as pose error, facilitates the simultaneous optimization of motion smoothness and tracking accuracy. The optimized trajectories are validated using an open-source physics simulator, showing enhanced motion stability, accuracy, and collision safety compared to conventional OLP-based paths. This proposed framework provides a flexible and cost-effective alternative to commercial OLP tools and lays a scalable foundation for future applications in automated and collaborative manufacturing systems.

Manufacturing systems are increasingly required to operate in high-mix, low-volume production environments, where process flexibility is crucial. One effective way to achieve this flexibility is through the use of multiple processing alternatives (MPA), allowing a product to be produced using different process plans or component structures. In MPA environments, scheduling decisions must address both the selection of processing alternatives for each product and the execution order of the resulting production tasks. Additionally, processing times often vary due to machine conditions and process variability, further complicating scheduling. This study introduces a dual-network-based deep reinforcement learning method for scheduling in manufacturing systems with multiple processing alternatives. The framework utilizes two Q-networks to learn both the selection of processing alternatives and the dispatching rules. Computational experiments demonstrate that the proposed method effectively reduces both the average makespan and its variability compared to a genetic algorithm-based approach, particularly as the problem size increases, showcasing its effectiveness in the face of processing time uncertainty.

Chemical Mechanical Polishing (CMP) is a crucial process in advanced semiconductor manufacturing, essential for achieving global planarization of the wafer surface, which directly impacts device performance and yield. Uniform material removal across the wafer is vital; however, non-uniformity frequently occurs, even with nominally uniform applied pressure. A prevalent issue is the edge effect, where the removal rate at the wafer edge significantly differs from that at the center, resulting in reduced uniformity and compromised device reliability. To tackle this challenge, this study explores the effectiveness of a multi-zone pressure-controlled carrier in enhancing polishing uniformity. Conventional single-zone carriers can only influence a narrow region of approximately 5–7 mm at the wafer edge, leading to limited improvements in nonuniformity of about 3%. In contrast, the multi-zone carrier allows for precise pressure control over a broader range, extending from 3 mm to 20 mm from the wafer edge. Experimental results show that this approach reduces non-uniformity to below 3% while effectively addressing edge removal deficiencies. These findings underscore the significant potential of multi-zone carriers to improve CMP process precision. Consequently, the proposed method is anticipated to enhance both productivity and quality in semiconductor fabrication.

In this study, we proposed a methodology for predicting tool wear in the turning process using the SVR model. This model maintains stable performance even in small-scale data environments and demonstrates robust characteristics against outliers. We detected changes in machining performance caused by tool wear through an AE sensor and accelerometer. Features were extracted from the acquired sensor signals and utilized in the machine learning model. Prior to training, the extracted features underwent a preliminary screening process based on distance correlation. By optimizing the feature combination using the RFECV algorithm, we achieved a prediction accuracy of R² = 0.95. The analysis revealed that key features influencing the tool wear prediction model included several significant variables. Additionally, we found that evaluating feature importance allowed for more efficient model improvement. Overall, when developing a tool wear prediction model for cutting, it is crucial to utilize various sensor signals, extract features in both the time and frequency domains, and optimize the combination of those features.

We present an automated incasing process designed to replace traditional manual packaging of dried seaweed. This system consists of three key components: a cage mechanism that compresses and transfers six bundles, a handling device for stacking the bundles, and a collaborative robot that performs the box incasing operation based on sensor input. The handling device utilizes pneumatic actuators and a wire-linked folding plate to minimize interference within the confined box space, while also allowing for adjustable dimensions to accommodate seasonal variations in bundle size. Field validation was carried out under continuous input conditions using a conveyor. The collaborative robot followed a predefined sequence triggered by a presence sensor, effectively grasping, stacking, compressing, and transferring bundles without causing product damage. Experimental results indicated that the system successfully incased 72 bundles per box with stable performance and reliable placement. These findings demonstrate the feasibility of replacing labor-intensive operations with collaborative robotic automation in seafood packaging, highlighting opportunities for enhanced consistency, ergonomics, and productivity.

This study introduces an automated robotic system designed to replace manual maintenance in cold rolling mills, where hazardous confined spaces present significant safety risks to workers. To enhance safety and efficiency, we modified a commercial aerial work platform into a teleoperated mobile robot. The system includes a redesigned end-effector equipped with high-pressure cleaning nozzles and a wide-angle camera for visual inspection. Experimental validation in both laboratory and field settings demonstrated the system's maneuverability and effectiveness. The results indicate that this robotic solution can successfully reduce safety hazards by minimizing manual intervention while ensuring high-quality cleaning and inspection in industrial rolling mills.

This study examines the deformation behavior and microstructural evolution of 6061 aluminum alloy processed through severe plastic deformation (SPD) via biaxial alternate forging. The objective was to evaluate both the alloy's formability limit and mechanical properties. Finite element (FE) analysis was conducted to simulate the biaxial alternate forging process, incorporating the strain-hardening coefficient and the number of forging passes. When the strain-hardening coefficient was set to 0, an average effective strain of approximately 440% was observed in a 4 mm diameter region at the core of the workpiece after eight forging passes. In contrast, with a strain-hardening coefficient of 0.2, the average effective strain under the same conditions decreased to about 300%. The FE analysis of the 6061 aluminum alloy estimated an average effective strain of 326% after eight passes, indicating a level of severe plastic deformation well beyond the elongation capacity of the initial material. Tensile testing revealed that after two passes, the material showed a gradual increase in strength with only a minimal reduction in elongation. Even after accumulating a significant strain of 326% through eight passes, optical microscopy displayed deformed grains and twinning structures, with no signs of recrystallization across all examined forging conditions.

This study focuses on developing Spur Gear parts for electric hedge trimmers using precision cold forging technology. The existing product faced lifespan and quality issues, leading to frequent replacements and increased costs. To address these problems, we implemented a three-stage cold forging process, which enhances product durability and reduces expenses. Previously, Spur Gear parts were produced using traditional machining methods, including CNC machining, gear hobbing, and MCT machining. However, these methods resulted in frequent damage, contributing to higher costs. By transitioning to a three-stage cold forging process, we aim to significantly improve the lifespan of the product.

Magnetic gears transmit torque via non-contact electro-magnetic coupling, which eliminates mechanical contact and significantly reduces wear, backlash, and noise compared to traditional mechanical gears. These benefits make magnetic gears particularly appealing for high-precision, high-reliability applications. However, achieving both high torque density and high gear ratios necessitates an optimized structural design that promotes efficient magnetic flux distribution while minimizing leakage and saturation. This study focuses on a hollow-type magnetic gear for collaborative robots that offers a high gear ratio. It employs topology optimization in conjunction with finite element analysis (FEA) to enhance torque density and efficiency. Key design variables, such as the geometry of the ferromagnetic core and the arrangement of permanent magnets, were optimized to increase average torque and reduce torque ripple and electro-magnetic losses. A prototype based on the optimized model was fabricated, and its performance was validated using a conventional direct torque measurement system. Experimental results were compared with simulation predictions to evaluate accuracy and analyze loss characteristics. The findings demonstrate the effectiveness of the proposed optimization approach and provide practical guidelines for designing high-efficiency magnetic gears suitable for advanced drive systems, including electric mobility and renewable energy applications.

Ethanol poses a significant threat to driver safety, as its effects vary with blood alcohol concentration (BAC). Common methods for estimating BAC include breath alcohol analysis, which calculates BAC from the alcohol concentration in exhaled breath, and direct blood sampling. However, these methods have notable limitations. This study aims to classify alcohol concentration using non-invasive optical signal data obtained from biomimetic samples with varying alcohol levels. To replicate the high scattering characteristics of biological tissue, scattering effects were induced in the samples, and absorbance was measured using near-infrared (NIR) wavelengths, which penetrate biological tissue more deeply. A Random Forest (RF) model was trained using the measured absorbance values to classify alcohol concentration levels. The Area Under the ROC Curve (AUC) for each concentration level indicated effective model learning, and the classification results on the test set demonstrated statistically significant accuracy. These findings suggest that the RF model can classify alcohol concentrations non-invasively and without the loss of samples. Furthermore, incorporating additional optical properties beyond absorbance may improve the accuracy of future non-invasive alcohol concentration classification models.

The increasing use of computational modeling and simulation (CM&S) in the medical device sector has heightened the need for ensuring simulation credibility. The ASME V&V 40 standard offers a structured framework for assessing credibility, consisting of 23 factors divided into three main categories: Verification, Validation, and Applicability. However, practical guidance for implementing these factors is still scarce. This study systematically reviewed and analyzed ten CM&S-related publications in the medical device field that utilized the ASME V&V 40 framework. It examined how each publication addressed the credibility factors and compared their implementation methods, evaluation criteria, and credibility levels. From this comparative analysis, we developed implementation strategies focused on credibility factors, field-specific characteristics, and model risk levels in real-world regulatory and development contexts. Key considerations for the practical application of each factor were identified, and recommendations for effective implementation were proposed. These findings offer practical guidance for ensuring credibility in CM&S-based medical device development, performance evaluation, and regulatory processes. By clearly demonstrating the applicability of the ASME V&V 40 framework, this work provides valuable direction for related industries and research institutions, aiming to improve CM&S credibility and promote its broader adoption in healthcare.

Among 3D printing techniques, fused deposition modeling (FDM) is known for its design flexibility, rapid fabrication, and the ability to produce complex geometries without molds. However, weak interlayer adhesion often results in poor mechanical strength along the build (Z) direction, limiting its use in structural applications. Instead of altering printing parameters or switching technologies, we propose a simple microwave-irradiation post-treatment to enhance interlayer bonding in FDM-printed parts. By optimizing microwave power and exposure time, we significantly improved interlayer fusion while maintaining the original geometry. Cross-sectional microscopy before and after treatment confirmed markedly improved interlayer bonding (Unbonded interfacial area fraction: 56.82% → 15.51%; -41.31 percentage points, -72.7%). Correspondingly, the Z-direction tensile strength increased from 42.38 to 49.11 MPa (+6.73 MPa, +15.9%). This straightforward post-processing method effectively addresses a key limitation of FDM, thereby expanding its potential for structural and industrial applications.