Search

Quality control is being increasingly automatised in the context of Industry 4.0. Its automatisation reduces inspection times and ensures the same criteria are used to evaluate all products. One of the challenges when developing supervised machine learning models is the availability of labelled data. Few-shot learning promises to be able to learn from few samples and, therefore, reduce the labelling effort. In this work, we combine this approach with unsupervised methods that learn anomaly maps on unlabelled data, providing additional information to the model and enhancing the classification models' discriminative capability. Our results show that the few-shot learning models achieve competitive results compared to those trained in a classical supervised classification setting. Furthermore, we develop novel active learning data sampling strategies to label an initial support set. The results show that using sampling strategies to create and label the initial support set yields better results than selecting samples at random. We performed the experiments on four datasets considering real-world data provided by Philips Consumer Lifestyle BV and Iber-Oleff - Componentes Tecnicos Em Plástico, S.A. [ABSTRACT FROM AUTHOR]

We consider a stochastic capacitated lot sizing problem with in-line rework. Both defective and defect-free products can be the result of an imperfect production process in real-world production systems. However, the proportion of defective items in a production lot is subject to uncertainty. For economic and environmental reasons, the possibility to rework defective products appears reasonable since these products usually have considerable value. In this paper, a nonlinear model formulation for integrated lot sizing and rework planning is proposed. We use a sample average approach to approximate the generic nonlinear model formulation. To cope with the randomness of the proportion of defective products, we apply a flexible planning approach that allows the adjustment of production and rework quantities. We conduct extensive numerical investigations to evaluate the performance of the proposed planning approach. [ABSTRACT FROM AUTHOR]

Monitoring the number of defects in constant‐size units is a well‐defined problem in the industrial domain and usually, the c$c$ control chart is used for monitoring the total number of defects in a product or a sample of products. The c‐chart tracks the total number of defects in each case by assuming that the underlying number of defects (single or several different types of defects) follows approximately the Poisson distribution. An interesting class of problems where the c$c$‐chart is used is when the number of defects in a surface is of interest. Although the number of defects on the surface of products characterizes the quality of the products, it is especially important how concentrated the defects are in specific parts of the product. In this paper, we introduce a scan‐based monitoring procedure, which simultaneously combines control charts for monitoring the evolvement of the number of defects (in general, events) through time and scan statistics for exploring the spatial distribution of defects. The numerical illustration showed that the new procedure has excellent performance under different scenarios. [ABSTRACT FROM AUTHOR]

Defects in integrated circuit (IC) packaging are inevitable. Several factors can cause defects in IC packaging such as material quality, errors in machine and human handling operations, and non-optimized processes. An automated optical inspection (AOI) is a typical method to find defects in the IC manufacturing field. Nevertheless, AOI requires human assistance in the event of uncertain defect classification. Human inspection often misses very tiny defects and is inconsistent throughout the inspection. Therefore, this study proposed a low-cost IC packaging defect classification system using edge impulse and ESP32-CAM. The method involves training a deep learning model (i.e., convolutional neural network (CNN)) using a dataset of non-defective and defective ICs on Edge Impulse. For defective ICs, the top surface of the ICs is deliberately scratched to imitate the cosmetic defects. ICs with scratch-free on their top surfaces are considered non-defective ICs. A successfully trained model using Edge Impulse is subsequently deployed on ESP32-CAM. The model is optimized to fit the limited resources of the ESP32-CAM. By using the built-in camera in ESP32-CAM, the trained model can perform a real-time image classification of non-defective/defective ICs. The proposed system achieves 86.1% prediction accuracy by using a 1,571 image dataset of defective and non-defective ICs. [ABSTRACT FROM AUTHOR]

Carbon fiber reinforced plastics inevitably develop defects such as delamination, inclusions, and impacts during manufacturing and usage, which can adversely affect their performance. Ultrasonic phased array inspection is the most effective method for conducting nondestructive testing to ensure their quality. However, the diversity of defects within carbon fiber reinforced plastics makes it challenging for the current ultrasonic phased array inspection techniques to accurately identify these defects. Therefore, this paper presents a method for the ultrasonic phased array nondestructive testing and classification of various internal defects in carbon fiber reinforced plastics based on convolutional neural networks. We prepared an ultrasonic C-scan dataset containing multiple types of internal defects, analyzed the defect features in the ultrasonic C-scan images, and established an autoencoded classifier network to recognize manufacturing defects and impact defects of varying sizes. The experiments showed that the proposed method demonstrates superior defect feature extraction capabilities and can more accurately identify both impact and manufacturing defects. [ABSTRACT FROM AUTHOR]

Additive manufacturing (AM) of H13 tool steel using the powder bed fusion (PBF) method is often limited by issues such as porosity (including lack of fusion, gas, and keyhole porosity), balling formation, and elemental segregation. These factors significantly impact the mechanical properties of the final product. This research investigates how volumetric energy density (VED) influences these issues and their subsequent effects on mechanical properties. VED affects the mechanical properties in multifaceted ways. An optimized VED prevents a lack of fusion and minimizes gas and keyhole porosities, especially near edges, ultimately reducing the material's susceptibility to fracture under tensile load. Additionally, VED influences elemental segregation within a single laser track during the process; lower VED leads to less elemental segregation. This reduced segregation minimizes oxide formation, which are crack initiation sites after heat treatment, thereby enhancing mechanical strength. The study also identifies that an optimal VED minimizes balling formation, further reducing elemental segregation and improving mechanical properties. A VED range of 57–59 J/mm3 is found to be optimal for preventing a lack of fusion, minimizing segregation, and reducing near-surface defects. Furthermore, a relationship is established between these defects, the microhardness profile, and the mechanical properties, suggesting that microhardness can serve as a predictive tool for the mechanical properties of PBF-ed metal. [ABSTRACT FROM AUTHOR]

Calendering is a crucial process in manufacturing of lithium-ion batteries electrodes. Wrinkles and corrugations are the main manufacturing defects in the calendering process. This study aims at studying the formation mechanism of wrinkles and corrugations. Corrugations and wrinkles of the electrodes were revealed to be more dependent on the difference between the front and back tensions. The tortuosity of corrugations was reduced by 34.2% when the tension difference decreased from 20 to 5 N. The increase in the tension difference led to a linear increase of the shear displacement δ, causing severe wrinkles in the uncoated area of the electrodes. The analytical prediction model for wrinkles during calendering was established based on the shearing of a rectangular membrane. The optimal web tension conditions, at a tension difference of 17 N, was achieved to obtain the optimal calendered electrodes. [ABSTRACT FROM AUTHOR]

Rubber pad forming is one of the primary ways of manufacturing aviation sheet metal components. In addition to reducing forming die costs, using elastic rubber in rubber pad forming prevents scratching on sheet metal, which is crucial for flight safety. However, during the forming process, the rubber pad primarily applies a vertical force to the sheet metal and cannot effectively exert pressure on the sheet metal laterally. Consequently, the contours of the finished product could fall outside specifications. This can cause severe manufacturing defects such as wrinkling or springback and therefore require a significant amount of labor to remedy and prevent any defects. As such, improvements in production efficiency are necessary. This study presents a rubber pad forming die with a lateral forming force mechanism that provides a lateral compressive stroke to the forming process of sheet metal via wedge-shaped structures; in other words, a lateral forming force to rubber pad forming is applied. In our exploration of stacking rubber pads of varying hardnesses in different orders, the maximum force was a larger-the-better objective, and the thickness compression ratio of the rubber pad was a smaller-the-better objective. We then combined grey relational analysis and the entropy method to identify the rubber pad stacking order and thicknesses with the optimal forming results. Subsequently, we constructed an actual rubber pad forming die with a lateral forming force mechanism as well as performed experiments and verification using 6061-T6 sheet metal with a thickness of 1 mm for aircraft wing reinforcement. The results demonstrated that employing rubber pad forming with a lateral forming force mechanism could reduce the amount of springback and surface wrinkling, meet blueprint requirements, and provide a new method and application of rubber pad forming for aviation sheet metal. [ABSTRACT FROM AUTHOR]

Carbon fiber-reinforced composite hydrogen storage cylinder is a key component used in hydrogen fuel cell electric vehicles. However, some micro defects such as voids and delamination are inevitable during the manufacturing process. An efficient detection method for manufacturing defects is still lacking at present. In this work, industrial computerized tomography (CT) scanning was carried out and a large number of micro delamination with scattered sizes and random locations were found in the filament winding layer. Shearography technique based on digital speckle pattern interferometry (DSPI) was used to measure the surface deformation of the cylinders. It was found that the "butterfly-shaped" interference fringes representing the anomalous responses from defects can be significantly observed at the pressure difference of 0.62%–0.69% working pressure. Also, the crack was found originated from the delamination defect with the most significant "butterfly-shaped" fringes, which leads to a large area of interlaminar destruction during the hydraulic bursting test. [Display omitted] • Plenty of delamination defects were found in composite hydrogen storage cylinders. • Shearography measurement captured "butterfly-shaped" fringes in cylinder surfaces. • Such fringes represent the anomalous responses from delamination defects. • The anomalies can be detected at 0.62–0.69% working pressure using shearography. • "Butterfly-shaped" fringes revealed crack origination in hydraulic bursting test. [ABSTRACT FROM AUTHOR]

Surface defects in industrial refrigerator manufacturing processes can cause significant production losses and compromise product quality. This area is underexplored and currently, visual quality inspection is a subjective process that requires expert intervention, which limits process efficiency and can lead to errors in defect detection. This paper presents a novel approach for automatic surface defect detection using a combination of convolutional neural networks (CNN) and deflectometry. The proposed method takes advantage of the high accuracy and robustness of CNNs in image classification tasks and the sensitivity of deflectometry to detect subtle surface variations. First, a prototype was built to get the images from the refrigerator. Second, using video recordings, we captured surface topographic data using deflectometry, which we then use to generate surface images. Next, we train a CNN to classify the surface images as defective or normal. The proposed method offers a promising solution for automatic detection and quality control of surface defects in refrigerator manufacturing processes. However, this method could also improve the production of vehicles, household appliances in general, and any product that can suffer scratches and dents. [ABSTRACT FROM AUTHOR]

In this study, the prospective application of lime as a desulfurization agent for the cast‐iron industry is technically examined. This investigation encompasses a series of laboratory experiments conducted under atmospheric conditions, mirroring industrial settings by exploring two distinct methods for introducing lime powder onto and into molten cast iron using surface addition and gas injection techniques. Deoxidation agents (FeSi, SiC, and Al) are also incorporated to enhance the lime‐based desulfurization results. Based on the findings of this study, it is indicated that lime can be a reliable cast‐iron desulfurization agent by reaching an end‐sulfur concentration of <0.015 wt%, thus providing an opportunity for a sustainable alternative for the foundry industry. In this study, it is also revealed that adding a small quantity of Al is more effective at enhancing desulfurization results than Si due to its role in increasing the proportion of liquid slag during desulfurization. However, caution is advised regarding the limit of aluminum concentration in cast iron (0.1 wt%), and treatment temperatures should be kept above 1400 °C to prevent counterproductive effects and undesirable defects in the product. [ABSTRACT FROM AUTHOR]

X‐ray computed tomography (XCT) is extremely useful for the non‐destructive analysis of additively manufactured (AM) components. AM components often show manufacturing defects such as lack‐of‐fusion (LoF), which are detrimental to the fatigue life of components. To better understand how cracks initiate and propagate from internal defects, we fabricated Ti‐6Al‐4V samples with an internal cavity using electron beam powder bed fusion. The samples were tested in high‐cycle and very high‐cycle fatigue regimes. XCT was used to locate crack initiation sites and to determine characteristic properties of cracks and defects with the aid of deep learning segmentation. LoF defects exposed to the outer surface of the samples after machining were found to be as detrimental to fatigue life as the internal artificial defects. This work can benefit industries that utilize the AM of high‐strength, lightweight alloys, in the design and manufacturing of components to improve part reliability and fatigue life. [ABSTRACT FROM AUTHOR]

Problem definition: The paucity of data on governmental regulatory agencies' product safety defect investigations has restricted our knowledge about (1) the determinants of a regulator's decisions to open or close an investigation, (2) the process it follows between opening and closing of an investigation, and (3) the outcomes of the investigation when it is closed. Methodology/results: The authors view a safety regulator's opening and closing of a product defect investigation as a decision of interest to the operations management discipline. This data paper describes a rich, novel, and hand-collected data set of all investigations that the National Highway Traffic Safety Administration—the U.S. regulator for automobile safety—opened and closed against 187 manufacturers between 2009 and 2021. The authors provide two Microsoft Excel data files, one capturing data for the investigations opened and the other for the investigations closed. The data files enable researchers to address three sets of research questions. First, researchers can use the "Data on Investigations Opened" file to model the determinants of a regulator's opening of a product defect investigation. Second, researchers can mine the textual variables from both files to identify the steps involved in the investigation process. They can also use the process variables included in the data to investigate the regulator's efficiency in opening and closing investigations. Third, researchers can use the "Data on Investigations Closed" file to better understand when and why a regulator closes an investigation and the outcomes of the closed investigations. Managerial implications: The data files can also be valuable to nonacademic stakeholders (e.g., governmental organizations and regulators, journalists, liability lawyers, politicians, and safety advocates). The authors provide an open-access website that simplifies the use of the data for a nonacademic audience and allows them to draw insights from the data via graphs and tables. Supplemental Material: The online supplement is available at https://doi.org/10.1287/msom.2023.0705. [ABSTRACT FROM AUTHOR]

With the Fourth Industrial Revolution (Industry 4.0) and advances in digital technology, zero-defect manufacturing (ZDM) has become a transformative and attractive concept that has the potential to reshape the manufacturing landscape. In this structured literature review, recent developments in ZDM in the textile industry from 2004 to 2023 are examined, with a focus on detection, repair, prediction, and prevention. Through bibliometrics analysis and evaluation of the current situation of ZDM technology, four main shortcomings are highlighted, to be specific, limitations in automated defect detection, incomplete artificial intelligence (AI)-based repair strategies, restricted predictive research, and focused prevention mechanism. Meanwhile, open challenges that require urgent attention are explored, that is systematic ZDM strategy integration, data management complexity, and demand for flexible ZDM frameworks. To address these shortcomings and challenges, three further prospectives are proposed, including addressing research imbalance, vision for an integrated ZDM system, and evolutionary predictive models. These prospectives aim to advance the field and drive the holistic development of ZDM technologies in the textile industry by promoting a more intelligent production strategy with higher quality and less waste. [ABSTRACT FROM AUTHOR]

This article marks a significant advancement in the field of quality management, specifically focusing on the evolution from traditional Statistical Quality Control (SQC) and Statistical Process Control (SPC) methods to a more advanced Learning Quality Control (LQC) approach. The research introduces Quality 4.0 (Q4.0) as a novel paradigm that fuses the technologies of the fourth industrial revolution, Manufacturing Big Data (MBD), Industrial Internet of Things (IIoT), Cloud Storage and Computing (CSC) and Artificial Intelligence (AI), with traditional quality management practices. The central theme of this study is exploring the limitations inherent in conventional quality control methods when faced with the complexities of modern manufacturing environments. The authors propose LQC systems as a solution, employing binary classification algorithms to predict and detect defects in manufacturing processes. This represents a shift from reactive to proactive quality measures enabled by AI's real-time data processing capabilities. The document delves into the evolution of manufacturing data across industrial revolutions, highlighting the exponential growth of unstructured data and its challenges. Through case studies, the authors illustrate the practical applications of LQC systems, demonstrating their ability to learn complex patterns in hyperdimensional spaces and automate tasks traditionally performed visually. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) provides considerable opportunities to assist human work. However, one crucial challenge of human–AI collaboration is that many AI algorithms operate in a black-box manner where the way how the AI makes predictions remains opaque. This makes it difficult for humans to validate a prediction made by AI against their own domain knowledge. For this reason, we hypothesize that augmenting humans with explainable AI improves task performance in human–AI collaboration. To test this hypothesis, we implement explainable AI in the form of visual heatmaps in inspection tasks conducted by domain experts. Visual heatmaps have the advantage that they are easy to understand and help to localize relevant parts of an image. We then compare participants that were either supported by (a) black-box AI or (b) explainable AI, where the latter supports them to follow AI predictions when the AI is accurate or overrule the AI when the AI predictions are wrong. We conducted two preregistered experiments with representative, real-world visual inspection tasks from manufacturing and medicine. The first experiment was conducted with factory workers from an electronics factory, who performed assessments of whether electronic products have defects. The second experiment was conducted with radiologists, who performed assessments of chest X-ray images to identify lung lesions. The results of our experiments with domain experts performing real-world tasks show that task performance improves when participants are supported by explainable AI with heatmaps instead of black-box AI. We find that explainable AI as a decision aid improved the task performance by 7.7 percentage points (95% confidence interval [CI]: 3.3% to 12.0%, ) in the manufacturing experiment and by 4.7 percentage points (95% CI: 1.1% to 8.3%, ) in the medical experiment compared to black-box AI. These gains represent a significant improvement in task performance. [ABSTRACT FROM AUTHOR]

To ensure the quality, safety, and long-term durability of fiber-reinforced composite components, it is necessary to explore approaches for monitoring and detecting defects or damage. Nondestructive detection techniques based on dynamic properties and structural response, such as natural frequency, damping, and modal shape. This study presents a numerical-experimental methodology based on dynamic analysis of structures made of composite materials. The methodology utilizes a surrogate model to establish a design envelope through a Kriging metamodel and to assess a damage index for structures affected by impact damage. The Latin Hypercube method is employed to generate values for the input variables, while the finite element method is utilized to calculate the natural frequencies. The Kriging metamodel is then employed to generate a numerical model, which is optimized using the Efficient Global Optimization algorithm and the expected improvement metric to minimize computational costs. The methodology yields a frequency range and determines a design envelope to evaluate the manufacturing quality of the structure. A damage index is used to identify structures with defects or impact damage, allowing for the assessment of severity. Additionally, the study evaluates the impact of incorporating metrics into the Latin Hypercube method to further reduce computational costs. Finally, this proposed approach contributes to the development of monitoring systems for assessing the manufacturing quality of composite structures and detecting impact damage through dynamic analysis. By utilizing this methodology, it becomes possible to effectively identify and evaluate the severity of defects and damage in composite structures, thus enhancing quality evaluation processes. [ABSTRACT FROM AUTHOR]

Abstract\nHIGHLIGHTSThe effect of automated fiber placement periodical gap and overlap defects on the microstructure, load–displacement curves, and out-of-plane displacements of bistable laminates due to the higher ratio of gap width to laminate thickness as well as the high sensitivity to imperfections is studied in this article. Results show that applying caul plates during curing can accurately predict snap loads and maximum out-of-plane displacements of the laminates, as well as prevent significant variations in curvature at the bending boundary. Gaps and overlaps of various widths embedded in a single layer facilitate the customization of the mechanical properties of bistable laminates.This article combines continuous fiber thermoset resin prepreg tapes with automated fiber placement technology to realize the manufacture of bistable laminates, which provides an efficient and accurate method compared with manual layup.Bistable laminates embedded gaps and overlaps are cured with caul plates that promote resin flow, resulting in uniform thickness variation while avoiding stress concentration and fluctuating curvature variations near the bending boundary.Because manufacturing bistable composite laminates with automated fiber placement process is unable to eliminate gap and overlap defects, we can utilize this interesting phenomenon of periodic gaps and overlaps of different width embedding in the 0° layer to tailor the mechanical properties of bistable laminates. [ABSTRACT FROM AUTHOR]

Carbon/carbon (C/C) composites have strict requirements for their thermal properties in extreme environments, especially at high temperature, where their properties are crucial for long-term life. The structure of C/C composites is exceptionally complex and exhibits multi-scale characteristics, and their thermophysical properties are closely related to temperature. Therefore, exploring their thermal response and heat transfer mechanisms through experimental method is relatively costly. This paper constructs a multi-scale finite element model to investigate the influence of structure and manufacturing defects on performance, and analyzes the thermophysical properties of the composites at High-temperature. By constructing a micro-representative volume element (Micro-RVE) that includes fibers, matrix, and pores, and using a steady-state heat transfer analysis method, the influence of material phase distribution on performance is studied. At the same time, a Meso-RVE reflecting the layering form, needle punching effect and manufacturing defects is established, and the transient heat transfer analysis method is used to investigate the influence of the structural form on performance. Finite element analysis shows that, the simulation values of the three C/C composites of unidirectional fiber bundles, short-chopped fiber felts, and needle punched C/C composites show good consistency with the experimental results in the published literature. This paper uses numerical methods to calculate the thermophysical properties of C/C composites from room temperature to 900°C and explores their variation with temperature. The constructed multi-scale model provides an accurate and effective method for predicting the thermophysical properties of C/C composites and also provides new perspectives and insights for thermal-mechanical coupling analysis and structural design. [ABSTRACT FROM AUTHOR]

The manufacturing industry environment is rapidly evolving into smart manufacturing. It prioritizes digital innovations such as AI and digital transformation (DX) to increase productivity and create value through automation and intelligence. Vision systems for defect detection and quality control are being implemented across industries, including electronics, semiconductors, printing, metal, food, and packaging. Small and medium-sized manufacturing companies are increasingly demanding smart factory solutions for quality control to create added value and enhance competitiveness. In this paper, we design and develop a high-speed defect detection system based on a line-scan camera using deep learning. The camera is positioned for side-view imaging, allowing for detailed inspection of the component mounting and soldering quality on PCBs. To detect defects on PCBs, the system gathers extensive images of both flawless and defective products to train a deep learning model. An AI engine generated through this deep learning process is then applied to conduct defect inspections. The developed high-speed defect detection system was evaluated to have an accuracy of 99.5% in the experiment. This will be highly beneficial for precision quality management in small- and medium-sized enterprises [ABSTRACT FROM AUTHOR]

In order to realize real‐time detection of aluminum defects during aluminum production, the target detection algorithm needs to be able to run on locally deployed hardware. Convolutional neural networks can effectively extract representative features from high‐dimensional data such as images and videos, and capture spatial information in the data, making it easier to locate aluminum defects. Moreover, running CNN model inference on local hardware has high real‐time performance. Due to the advantages of convolutional neural network in anomaly detection, an improved CenterNet aluminum surface defect detection method was proposed. The algorithm combines common convolution and depthwise separable convolution to design a lightweight convolution module. Then, the Convolutional Block Attention Module is added to the feature extraction network to make the network better capture the rich input feature information of the image. Ultimately, the α‐DIoU loss function is implemented to enhance the precision of bounding box predictions. The experimental findings demonstrate that the proposed algorithm achieves an average detection accuracy (mAP) of 86.02%, which is 5.95% higher than the average detection accuracy of the traditional algorithm, and has a good detection effect on aluminum surface defects. Furthermore, there is an 11.9% reduction in model parameters and a 15.2% decrease in floating‐point computations, which helps to promote the deployment of embedded device platforms. © 2024 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

A leakage accident occurred in an elbow of a natural gas processing station in western China. In order to analyze the cause of the accident, non-destructive testing, chemical composition analysis, hardness testing, metallographic examination, scanning electron microscopy analysis and other experimental analyses were carried out on the elbow. The obtained results showed that the present leakage is mainly attributed to the cracks. In fact, the microstructure around the cracks was decarburized which fully indicates that the cracks have experienced a high temperature environment. In other words, the elbow had formed initial cracks before the heat treatment process of manufacturing. Then, the cracks grew until they penetrated the wall thickness under the action of working pressure. In addition, considering that the material properties test results of the elbow meet the standard requirements, and there are no H2S, H2 and other media that cause SSC and HIC in the service environment, it is once again confirmed that the failure of the elbow is caused by defects in the original manufacturing stage. It is recommended that the same batch of elbows in service to carry out internal and external defects, and strengthen the quality inspection of the new elbows, to avoid defective elbows in the field to put into use. [ABSTRACT FROM AUTHOR]

Polymeric materials are inexpensive, lightweight, and easy to process. With injection molding technology, they can be used to mass-produce complex geometric products. Although virtual quality measurement systems can prevent product defects, they may not be effective for recycled polymers. Recycling is necessary to achieve sustainable production, but the recycling process can greatly affect material properties. Accordingly, we propose using transfer learning (TL) to apply quality models for virgin materials to recycled materials. Key features from cavity pressure sensor information were extracted as quality indices, and a neural network was trained to predict the part weight and geometric dimensions. TL was then used to fine-tune this pretrained model on data for recycled materials. This method requires little data collection and could greatly reduce the cost of improving quality when using recycled materials. For all quality metrics, the TL model achieved superior performance to a conventional model, even when applied to a 40% smaller dataset; it also converged more quickly and in fewer iterations. A comprehensive analysis indicated that the model had a favorable fit. These results confirm that TL is effective in predicting the quality of injection molding products made with recycled materials; it has superior predictive accuracy and requires less training time than conventional methods. The described process could reduce the cost of implementing process quality inspection systems, incentivizing the use of recycled materials. [ABSTRACT FROM AUTHOR]

X-rays can provide images when an object is visibly obstructed, allowing for motion measurements via x-ray digital image correlation (DIC). However, x-ray images are path-integrated and contain data for all objects between the source and detector. If multiple objects are present in the x-ray path, conventional DIC algorithms may fail to correlate the x-ray images. A new DIC algorithm called path-integrated (PI)-DIC addresses this issue by reformulating the matching criterion for DIC to account for multiple, independently-moving objects. PI-DIC requires a set of reference x-ray images of each independent object. However, due to experimental constraints, such reference images might not be obtainable from the experiment. This work focuses on the reliability of synthetically-generated reference images, in such cases. A simplified exemplar is used for demonstration purposes, consisting of two aluminum plates with tantalum x-ray DIC patterns undergoing independent rigid translations. Synthetic reference images based on the "as-designed" DIC patterns were generated. However, PI-DIC with the synthetic images suffered some biases due to manufacturing defects of the patterns. A systematic study of seven identified defect types found that an incorrect feature diameter was the most influential defect. Synthetic images were re-generated with the corrected feature diameter, and PI-DIC errors were improved by a factor of 3-4. Final biases ranged from 0.00-0.04 px, and standard uncertainties ranged from 0.06-0.11 px. In conclusion, PI-DIC accurately measured the independent displacement of two plates from a single series of path-integrated x-ray images using synthetically-generated reference images, and the methods and conclusions derived here can be extended to more generalized cases involving stereo PI-DIC for arbitrary specimen geometry and motion. This work thus extends the application space of x-ray imaging for full-field DIC measurements of multiple surfaces or objects in extreme environments where optical DIC is not possible. [ABSTRACT FROM AUTHOR]

The reliability growth of electromechanical products is a continuous process of addressing reliability defects, which is very important for manufacturing enterprises. At present, research on the reliability growth of electromechanical products mostly focuses on the reliability defects of the products themselves, ignoring the fact that manufacturing enterprises are the executors of product reliability related work. Improving the organizational reliability capability of manufacturing enterprises can enhance the reliability of electromechanical products. In order to understand the current situation of organizational reliability capability (ORC) in electromechanical product manufacturing enterprises and make improvements, this paper establishes an ORC evaluation indicator framework for electromechanical product manufacturing enterprises and evaluates it using the grey evaluation method. Firstly, an evaluation indicator framework for ORC is established based on enterprise research. Secondly, the ORC of electromechanical product manufacturing enterprises is evaluated by combining the three-parameter interval grey number and projection index function. Then, the evaluation results are analyzed from multiple perspectives to understand the current situation and shortcomings of ORC and guide its improvement. Finally, the evaluation indicator framework and method are explained through practical application in CNC machine tool manufacturing enterprises, and the effectiveness of the framework and method is demonstrated through the MTBF growth of CNC machine tools. [ABSTRACT FROM AUTHOR]

A carbon brush/collector ring set will have phenomena such as firing and ablation during operation, which is due to the existence of various abnormal contact modes of the brush/ring during operation, thus changing the carbon brush/collector ring interface state. To analyze the effects of different contact modes on the performance of the brush/ring, in this paper, we construct the contact modes of the air gap (loss of contact leads to the existence of a small gap between the two surfaces), direct contact (contact with abrasive particulate media), and surface porosity contact (contact when there is a large pit on the surface of the collector ring due to manufacturing quality defects and abnormal abrasion), and analyze the effects of the various states on the core parameters such as current conduction, ring surface damage, and carbon brush abrasion, which provide a basis for the active suppression of the damage. [ABSTRACT FROM AUTHOR]

In the age of smart technology, the widespread use of small LCD (Liquid Crystal Display) necessitates pre-market defect detection to ensure quality and reduce the incidence of defective products. Manual inspection is both time-consuming and labor-intensive. Existing methods struggle with accurately detecting small targets, such as point defects, and handling defects with significant scale variations, such as line defects, especially in complex background conditions. To address these challenges, this paper presents the YOLO-DEI (Deep Enhancement Information) model, which integrates DCNv2 (Deformable convolution) into the backbone network to enhance feature extraction under geometric transformations. The model also includes the CEG (Contextual Enhancement Group) module to optimize feature aggregation during extraction, improving performance without increasing computational load. Furthermore, our proposed IGF (Information Guide Fusion) module refines feature fusion in the neck network, preserving both spatial and channel information. Experimental results indicate that the YOLO-DEI model increases precision by 2.9%, recall by 13.3%, and mean Average Precision (mAP50) by 12.9%, all while maintaining comparable parameter counts and computational costs. These significant improvements in defect detection performance highlight the model's potential for practical applications in ensuring the quality of LCD. [ABSTRACT FROM AUTHOR]

The manufacturing industry must maintain high-quality standards while meeting customer demands for customization, reduced carbon footprint, and competitive pricing. To address these challenges, companies are constantly improving their production processes using quality management tools. A crucial aspect of this improvement is the root cause analysis of manufacturing defects. In recent years, there has been a shift from traditional knowledge-driven approaches to data-driven approaches. However, there is a gap in the literature regarding a systematic overview of both methodological types, their overlaps, and the challenges they pose. To fill this gap, this study conducts a scoping literature review of root cause analysis in manufacturing, focusing on both data-driven and knowledge-driven approaches. For this, articles from IEEE Xplore, Scopus, and Web of Science are examined. This review finds that data-driven approaches have become dominant in recent years, with explainable artificial intelligence emerging as a particularly strong approach. Additionally, hybrid variants of root cause analysis, which combine expert knowledge and data-driven approaches, are also prevalent, leveraging the strengths of both worlds. Major challenges identified include dependence on expert knowledge, data availability, and management issues, as well as methodological difficulties. This article also evaluates the potential of artificial intelligence and hybrid approaches for the future, highlighting their promises in advancing root cause analysis in manufacturing. [ABSTRACT FROM AUTHOR]

This study examines the residual stress induced by manufacturing and its effect on failure in thermosetting unidirectional composites under quasi-static loading, using Finite Element-based computational models. During the curing process, the composite material develops residual stress fields due to various phenomena. These stress fields are predicted using a constitutive viscoelastic model and subsequently initialized within a damage-driven Phase Field model. Structural tensors are used to modify the stress-based failure criteria to account for inherent transverse isotropy. This influence is incorporated into the crack phase field evolution equation, enabling a modular framework that retains all residual stress information through a heat-transfer analogy. The proposed coupled computational model is validated through a representative numerical case study involving L-shaped composite parts. The findings reveal that cure-induced residual stresses, in conjunction with discontinuities, play a critical role in matrix cracking and significantly affect the structural load-carrying capacity. The proposed coupled numerical approach provides an initial estimation of the influence of manufacturing defects and streamlines the optimization of cure profiles to enhance manufacturing quality. Among the investigated curing strategies, the three-dwell cure cycle emerged as the most effective solution. [ABSTRACT FROM AUTHOR]

The current study investigates the process of preparing and analysing porous-structured ceramics made from zirconium, aluminium, and magnesium ceramic oxides. The starch consolidation casting (SCC) technique, with different types of starches (potato and tapioca), was used for this purpose. Our objective was to methodically examine the impact of different processing factors, such as the temperature at which pre-sintering and sintering occur, and the proportions of ceramic powders, on the microstructure, mechanical characteristics, and porosity of the resultant composites. Pre-sintering effectively reduced the rate of shrinkage during the final sintering stage; this resulted in more controlled and predictable shrinkage, leading to better dimensional stability and reduced risk of defects in the final product. A higher alumina content was associated with an increase in apparent porosity and a reduction in volume shrinkage and apparent densities. The mercury intrusion porosimetry (MIP) findings concluded that the prepared porous ceramics have a multi-modal pore structure. The highest calculated compressive strength was 76.89 MPa for a sample with a porous structure, which was manufactured using 20 wt.% tapioca starch and 30 wt.% alumina content. The main advantage of alumina is its ability to improve compressive strength by refining the grain structure and serving as a barrier against fracture development. [ABSTRACT FROM AUTHOR]

Isolation bearings play an important role in the seismic resilience of highway bridges. Flexible and high‐strength reinforcement has been applied in elastomeric isolation bearings to substitute conventional rigid steel plate reinforcement to enhance their lateral performance, for example, lower lateral stiffness and larger deformability. However, the main literature shows that existing flexible reinforcement, such as carbon/glass fiber fabric, may not guarantee a sufficient vertical load‐carrying capacity of elastomeric bearings to meet the design requirement of 30 MPa considering the vertical seismic effect. To this end, the emerging high‐strength steel woven wire mesh was introduced as an alternative flexible reinforcement for the bearings in this study to increase their ultimate compression capacity while maintaining superior lateral performance. Vertical compression tests were conducted on 34 specimens of the proposed unbonded steel‐mesh‐reinforced bearings (USRBs) to investigate the ultimate compression capacity. In addition to the general ultimate behavior of USRBs under vertical loading, the influence of various design parameters (i.e., individual rubber layer thickness, number of reinforcement layers, bearing design load) was investigated through comparisons among the specimens. From the test results, the compressive failure mechanism of USRBs was unveiled, which originated from the tensile failure of the steel mesh reinforcement. The steel mesh reinforcement was proved to increase the bearing ultimate compression capacity to an average of 52.0 MPa compared to fiber‐reinforced bearings, with 85% of specimens exceeding 30 MPa. Moreover, the compression capacity of USRBs was identified to be significantly affected by the individual rubber layer thickness. Specific discussions were further provided concerning the influence of potential manufacturing defects. Finally, suggestions were provided to further enhance the ultimate compression capacity of USRBs based on the results and discussions. [ABSTRACT FROM AUTHOR]

A study on the optimal design of preforms has previously been actively conducted as a method to solve defects such as voids and flash in forged products. However, previous research has generally been performed through deterministic optimization for ideal cases that do not take manufacturing tolerances into account. As a result, the application of such optimal designs in actual processes may be limited due to various factors such as material manufacturing tolerances and the machining environment of preforms. Therefore, this study conducted reliability-based optimization considering tolerances in billets and preforms. The objective of the study was to optimize the design of a bearing hub and minimize defects in the final product. When comparing deterministic optimization and reliability-based optimization, the former showed relatively superior results in terms of defect indicators but had a higher occurrence of voids and lower forming loads, increasing the probability of void occurrence. On the other hand, the reliability-based optimization showed relatively lower performance in quality improvement indicators, but it successfully met the target reliability of 99% by reducing the probability of defect occurrence. These results were derived using an approximate model based on the Kriging method, providing an optimal design that is practical and effective in actual manufacturing processes. [ABSTRACT FROM AUTHOR]

This research introduces an advanced algorithm based on convolutional neural networks for the detection and categorization of surface defects in manufacturing processes. At its core, the algorithm employs a deep learning model that integrates residual networks and attention mechanisms to effectively extract features. Additionally, we have developed a novel feature selection method, named NR, which synergistically combines neighborhood component analysis and ReliefF techniques. This approach enables the selection of more representative deep features for subsequent analysis. For the classification task, we utilize the support vector machine technique, which demonstrates versatility in handling both binary and multi‐class classification scenarios. The reliability and superiority of our algorithm are further validated through a comparative analysis using a dataset specifically tailored for this context. The results indicate that our approach outperforms existing algorithms in accurately identifying manufacturing defects. [ABSTRACT FROM AUTHOR]

This work aims to establish a quality assurance methodology for additively manufactured (AM) samples, produced from laser powder bed fusion (LPBF) method. The method incorporates resonance frequency method (RFM), where reference samples from wrought 316 L will be used to establish a data‐base with a set of reference frequencies. The data‐base is enhanced further with simulated frequencies, via FEM method, which was carried out on samples with the same dimensions and material properties as those of the reference. The quality of LPBF samples were benchmarked against this database. Four sets of LPBF samples (termed as A, B, C, and D) were printed with different parameters, and their densities were measured to understand deviations from the reference database. It was observed that Set‐C had the least drop in density of approx. 0.65% when compared to the wrought samples. Microscopic analysis revealed that the melt pools were clearly visible in all the samples, with no significant effect from different print parameters. Subsequently RFM was performed on all the sets and clear shifts in frequencies observed. Set‐C had the least deviation when compared to the reference (averaged at 200 Hz), whereas it was 250, 300, and 400 Hz for Set‐D, Set‐A and Set‐B respectively. There are several reasons for the frequency shift, the presence of porosity being one of them. Set‐B had the highest concentration of porosity in the ‐YZ plane. An algorithm was developed to sort the samples based on the frequency shifts seen from those of the samples from wrought 316 L. The sorting methodology was based on the shift frequencies, and the farther the sift is from the wrought the worst it get in terms of quality. The algorithm, which is programmed based on this methodology, was tested on a new set of LPBF samples and its effectiveness validated. [ABSTRACT FROM AUTHOR]

Compression resin transfer moulding (CRTM) has been widely used to manufacture automotive parts with reduced production cycle times. With the development of fast curing thermosetting resins, the CRTM process is a viable option for the high production rates in the transportation industry. However, the dynamic resin curing behaviour poses a potential risk of manufacturing defects in the part. In order to reduce the risk during the development of the tool and the process parameters, this paper proposes a modelling framework for the CRTM process when using fast curing resin systems. The work specifically focused on the coupling between heat transfer, resin cure, resin flow and preform compaction using a commercial code, PAM-RTM. The tool captures accurately the preform filling, temperature and resin pressure evolution during the injection and compression phase. The application of the framework was demonstrated for a complex 3D demonstrator. The predicted preform filling had an accuracy of 73% for the flow front evolution compared to the experimental results. This work demonstrates the validity of the framework proposed when dealing with resin systems that are challenging to process. [ABSTRACT FROM AUTHOR]

Injection‐moulded products may have a variety of defects in production. Failing to detect and fix the defects may reduce product quality and lead to safety issues. An injection‐moulded product defect detection model, injection‐moulded product‐detection transformer (IMP‐DETR), is proposed to address the challenges of diversity, small size, and complex background in injection‐moulded products. The model constructs a feature extraction backbone network with the inverted residual mobile block module to extract key information and reduce interference from irrelevant backgrounds while maintaining lightweight. The small object fusion pyramid feature fusion network is used to capture rich texture information from small objects to improve the detection performance of fuzzy and small‐sized defects. Additionally, the Conv3XC‐Fusion module is designed to resolve the problem of integrating multi‐scale features, improving the stability of detection. Due to the lack of publicly available datasets for injection‐moulded product defects, custom dataset containing 2500 defect images was constructed. The experimental results indicate that the mean average precision of the IMP‐DETR model reaches 82.4%. Compared to other benchmark object detection models, IMP‐DETR demonstrates superior detection performance and a smaller model size, which is suitable for application in real scenarios. [ABSTRACT FROM AUTHOR]

A company often faces incidents in which its offerings cause bodily (e.g., product safety defects) or psychological (e.g., data breach) harm to its consumers. Such incidents may invoke product liability lawsuits against the company. The company may try to recover from the liability-invoking failure by notifying the affected consumers, offering a remedy, and persuading them to comply with the company message. The authors theorize and experimentally demonstrate that, on average, a prevention-focused message receives greater compliance than a promotion-focused message. Further, a prevention-focused message is more effective with consumers from high-uncertainty-avoidance cultures, whereas a promotion-focused message is more effective in low-uncertainty-avoidance cultures. Perceived compatibility of prevention or promotion goals with low or high values of uncertainty avoidance mediates the interaction effect on compliance. The findings can help companies overcome consumer apathy to product recall or data breach notices and offer managers ways to promote consumer safety and protection. [ABSTRACT FROM AUTHOR]

In this paper, we analyze the literature concerning the implementation of digital twins (DTs) for zero-defect manufacturing (ZDM) following a systematic method and, guided by a preliminary finding that a structured and standardised approach to the development of the DT applications is lacking, we provide a standardised design methodology to guide researchers and practitioners in their efforts to develop DTs regardless of the domain. After examination and interpretation of the literature, we also present the results of our state-of-the-art analysis, discuss the current state and limitations of research and practice, and provide useful insights on this important and complex topic. The design methodology proposed in our study will benefit both practitioners and academicians by covering the essential elements to be considered when developing DTs for ZDM for any applications in this domain. The study also contributes to knowledge by presenting a structured overview of the specific research area with a comprehensive, systematic, and critical analysis of the literature and by providing answers to some fundamental questions in the context of DTs for ZDM. Finally, we provide suggestions for further developments in research and practice. [ABSTRACT FROM AUTHOR]

Noting the proliferation of product recalls and extensive use of lobbying in some critical product markets (e.g., automobiles, medical equipment), the authors examine the relationship between lobbying and product recalls. Lobbying does not alter product quality, so an efficiency perspective would suggest no relationship. However, a legitimacy-based institutional theory perspective and associated regulation models suggest that lobbying reduces voluntary firm-initiated and mandatory regulator-initiated recalls. To provide insights into these questions, the current study explores nine years of multisource data from the automotive industry, related to recalls and lobbying. The results, obtained with an instrumental variable approach, support dual impacts of lobbying for reducing both voluntary and mandatory recalls. Defect severity and media coverage moderate the effects, and the data support full indirect moderation, such that the interaction between media coverage and lobbying mediates the interaction between defect severity and lobbying. In terms of effect sizes, approximately $404,367 ($1.66 million) more in lobbying expenditures is associated with one fewer voluntary (mandatory) recall, assuming a typical average recall of 235,638 vehicles. This study highlights lobbying as an important (marketing) tool that automotive companies use to manage their regulatory environment, with deep implications for policy making, research, and practice. [ABSTRACT FROM AUTHOR]

Featured Application: Digital Twin used in real-time monitoring of operations performed on CNC technological equipment that will retrieve the relevant parameters of machine working. This article presents the development and implementation of a real-time monitoring solution designed for CNC machines, specifically applied to 150 industrial printing machines, leveraging Digital Twin (DT) technology. The system integrates an SQL database with Android and.NET interfaces, ensuring seamless data synchronization across all machines and optimizing production processes. The real-time monitoring enables immediate reflection of operational changes, enhancing predictive maintenance and reducing machine downtime. A notable feature of the system is its 1 s average data synchronization rate per machine, managing 150 resources distributed over a 10,000 mp area. This fast synchronization improves workflow coordination, reducing production time by approximately 10%, and minimizing operator delays caused by material issues, machine malfunctions, or product defects. The integration of advanced analytics further supports real-time decision-making, predictive maintenance, and performance optimization, aligning the solution with the objectives of Industry 4.0 and Industry 5.0 initiatives. This version reflects the specific results of the research, including the 1 s synchronization rate, the 10% reduction in production time, and the scalability of the system for 150 resources. [ABSTRACT FROM AUTHOR]

Honeycomb-structured materials are extensively utilised in both commercial and military aircraft. The occurrence of manufacturing defects and operational damage has emerged as a significant safety concern, consequently elevating the necessity for non-destructive testing (NDT) to identify flaws and damage during aircraft operation and maintenance. In addition to merely detecting defects, it is crucial to accurately characterise or classify them. In this paper, the honeycomb sandwich samples were designed and manufactured from two CFRP face-sheets and Nomex core with thick resin containing nano clay 30B mixed with ultrasonic coverage. Defects were placed into carbon sheets, resin, and core materials. An NDT technique based on infrared thermography was applied, along with PCA image processing, which automatically categorises prevalent defects found in honeycomb materials. These defects encompass debonding, adhesive issues, and fractures within the honeycomb core. Thermography results showed that defects such as delamination, core fracture, lack of adhesion, and resin inhomogeneity can be identified, especially with high accuracy using PCA image processing. [ABSTRACT FROM AUTHOR]

In traditional column grid array (CGA) packaging, column interconnections are achieved through precise mold positioning and reflow soldering of pre-printed solder paste. However, the presence of molds during the welding process increases the occurrence of welding defects and manufacturing costs. A friction inlay welding (FIW) method was proposed by the authors to achieve copper column connections without the assistance of molds. The study utilized the Coupled Eulerian–Lagrangian (CEL) method to model and simulate the FIW process. By integrating simulation results with experimental data, the temperature history, strain history, and solder flow behavior of the FIW micro-welding were determined. The results indicated that during the welding process, the peak values of temperature, strain, and solder flow velocity were all located at the base corners of the copper column. The peak temperature at the friction interface reached 158 °C (431 K), which was 88% of the solder melting point (490 K), with a peak flow velocity of 197 mm/s. The solder flow trajectory at the bottom of the copper column exhibits a downward spiral staircase pattern. After flowing to the edge of the bottom of the column, the solder rotates upward along one side of the column. On the side of the copper column, the thermoplastic solder near the copper column (within about 150–200 µm from the copper column) simultaneously flows in axial, radial, and circumferential directions. The flowing solder ultimately evolves into a microstructure composed of recrystallized fine grains in the stir zone (SZ) and deformed and elongated grains in the thermo-mechanically affected zone (TMAZ). The atomic diffusion between the copper column and the solder is excited and accelerated by the thermal–mechanical action caused by friction, and the connection layer composed of Cu6Sn5 intermetallic compound is formed at the interface. [ABSTRACT FROM AUTHOR]