Quality Of Industrial Products Research Articles

Metallic surface defect recognition network based on global feature aggregation and dual context decoupled head

Surface defect detection is a crucial inspection phase in ensuring industrial product quality and reliability, particularly in the context of metallic manufacturing components. While existing deep learning-based approaches have demonstrated some effectiveness, the design of certain network structures still lacks consideration for industrial scenarios. This paper proposed a complex metallic surface defect detection neural network which incorporates two innovatively designed modules along with other targeted enhancements. Firstly, the spatial pyramid pooling (SPP) structure is redesigned for richer global-local information fusion. A global feature fusion and redistribution module (GFAR) is designed to fully leverage feature information across different scales. With the combination of a novel global positional attention mechanism (GPAM), GFAR is capable of distinguishing defects from complex backgrounds. Moreover, the asymmetric convolution blocks (ACB) are employed to process the final features, enhancing the representational capacity for intricate defect shapes. Subsequently, the detection head is decoupled in both spatial and task domains to addresses the problems posed by inconsistent focus between classification and regression tasks. Finally, the reassignment of loss functions is undertaken for each respective task to enhance their alignment with the specific requirements of the defect detection task. Extensive ablation experiments are conducted on two steel surface defect datasets, NEU-DET and GC10-DET, to demonstrate the effectiveness of each module. The defects detection accuracy of our method respectively reaches 80.9 % and 84.8 % for the two datasets, which outperforms the majority of state-of-the-art detection neural networks.

Study of color degradation in package prints: Analyzing kinetics with principal component analysis

AbstractColor degradation is a pressing issue in various fields like food and medicine packaging, as well as printed materials, leading to a loss or distortion of vital information. Therefore, understanding how the color of package prints evolves over time is essential for product quality assessment and effective product life cycle management. This study focuses on investigating color changes in prints on blister foils, widely used in the medicine and packaging industries. To assess fading characteristics, samples printed via the gravure process have been exposed to a Xenon Test Chamber (B‐SUN). Spectrophotometric measurements of the prints have been conducted using a spectroradiometer, and the data have been analyzed using principal component analysis (PCA). The fading nature of cyan and black prints aligns with Type IV, while magenta and yellow prints correspond to Type II, as described by Giles. Notably, cyan and black inks demonstrated resistance toward fading at the end of the exposure, with a minor exception. The investigation involved curve fitting, revealing a consistent first‐order kinetic model governing fading rates throughout the exposure time. High R2 values (0.96–0.99) and low root means square error values (0.02–0.08) are observed across all cases. While color fading involves diverse kinetics, the fading rate consistently adheres to a singular first‐order kinetic path. The study highlights that yellow ink is most sensitive to light, followed by magenta ink. These findings offer valuable insights for optimizing packaging materials and ensuring product quality in industries reliant on printed materials. By mitigating color degradation, manufacturers can enhance the longevity and appeal of their products, reinforcing consumer trust and satisfaction. Furthermore, the effective utilization of PCA as a data analysis tool offers a valuable approach for future investigations in color science and related fields.

TD-Net:tiny defect detection network for industrial products

AbstractThe detection of tiny defects in industrial products is important for improving the quality of industrial products and maintaining production safety. Currently, image-based defect detection methods are ineffective in detecting tiny and variously shaped defects. Therefore, this paper proposes a tiny defect detection network (TD-Net) for industrial products to improve the effectiveness of tiny defect detection. TD-Net improves the overall defect detection effect, especially the detection effect of tiny defects, by solving the problems of downsampling of tiny defects, pre-filtering of conflicting deep and shallow semantic information, and cascading fusion of multi-scale information. Specifically, this paper proposes the Defect Downsampling (DD) module to realize the defect information supplementation during the backbone downsampling process and improve the problem that the stepwise convolution easily misses the detection of tiny defects. Meanwhile, the Semantic Information Interaction Module (SIIM) is proposed, which fuses deep and shallow semantic features, and is designed to interact the fused features with shallow features to optimize the detection of tiny defects. Finally, the Scale Information Fusion Module (SIFM) is proposed to improve the Path Aggregation Network (PANet) for cascading fusion and information focus on different scale information, which enables further improvement of defect detection performance of TD-Net. Extensive experimental results on the NEU–DET data set (76.8$$\%$$ % mAP), the Peking University PCB defect data set (96.2$$\%$$ % mAP) and the GC10-DET data set (71.5$$\%$$ % mAP) show that the proposed TD-Net achieves competitive results compared with SOTA methods with the equivalent parameter quantity.

Characterization of molar mass and conformation of relevant (non-)starch polysaccharides in cereal-based beverages

Arabinoxylans, β-glucans, and dextrins influence the brewing industry's filtration process and product quality. Despite their relevance, only a maximum concentration of β-glucans is recommended. Nevertheless, filtration problems are still present, indicating that although the chemical concentration is essential, other parameters should be investigated. Molar mass and conformation are important polymer physical characteristics often neglected in this industry. Therefore, this research proposes an approach to physically characterize enzymatically isolated beer polysaccharides by asymmetrical flow field-flow fractionation coupled to multi-angle light scattering and differential refractive index detector. Based on the obtained molar masses, root-mean-square radius (rrms from MALS), and hydrodynamic radius (rhyd), conformational properties such as apparent density (ρapp) and rrms/rhyd can be calculated based on their molar mass and size. Consequently, the ρapp and rrms/rhyd behavior hints at the different structures within each polysaccharide. The rrms/rhyd 1.2 and high ρapp values on low molar mass dextrins (1–2·105 g/mol) indicate branches, while aggregated structures at high molar masses on arabinoxylans and β-glucans (2·105 –6·106 g/mol) are due to an increase of ρapp and a rrms/rhyd (0.6–1). This methodology provides a new perspective to analyze starch and non-starch polysaccharides in cereal-based beverages since different physical characteristics could influence beer's filtration and sensory characteristics.

Robust Self-Constructing Fuzzy Neural Network-Based Online Estimation for Industrial Product Quality

Online estimation for product quality is crucial for improving industrial process efficiency. However, model degradation and outliers usually challenge industrial quality estimation models. To tackle the above problems, a robust self-constructing fuzzy neural network (RSC-FNN) is developed in the article. In the RSC-FNN, the rules can be automatically created or pruned, obtaining the online self-constructing mechanism (OSCM). First, an online error compensation algorithm is developed to generate new rules. Second, the model performance and contribution of existing rules are evaluated online to delete redundant rules. Thus, the OSCM effectively improves the structural adaptability and compactness of the RSC-FNN. Moreover, the correntropy-induced criterion, which can handle complex outliers, is modified for the parameter learning algorithm. Hence, the adverse effect of outliers can be suppressed during the parameter updating process. Besides, we analyze the convergence of the RSC-FNN to ensure its feasibility in industrial applications. Finally, two industrial applications are studied to test the effectiveness of the RSC-FNN. Compared with other FNN <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</sub> , the results indicate that the RSC-FNN performs better in learning efficiency, structure compactness, and robustness.

A Multiphase Dual Attention-Based LSTM Neural Network for Industrial Product Quality Prediction

A Multiphase Dual Attention-Based LSTM Neural Network for Industrial Product Quality Prediction

Determination of Thermotolerant Yeast Population in Dairy Products

Thermotolerant yeasts constitute a critical indicator for the effective management of product quality, food safety, and production processes in the dairy industry. This study aims to investigate the presence of thermotolerant yeasts in diverse dairy products within the context of Turkiye. In this study, pH and thermotolerant yeast properties of Turkish white cheese, butter, cream (Kaymak), and Tulum cheese were investigated. Thermotolerant yeast counts in butter samples ranged between 4.30 and 4.82 log CFU/g, with pH values ranging from 4.91 to 6.68. Thermotolerant yeast counts of Tulum cheese, white cheese and cream samples were &lt;2 log CFU/g. The pH values of cream and Tulum cheese were ranged from pH 5.07-6.79 and pH 4.91-6.68, respectively. The mean pH value for white cheeses was 4.98. As a results, thermotolerant yeast counts of Tulum cheese, white cheese and cream samples were below the detectable value; however, a high number of thermotolerant yeasts were determined in butter samples. In conclusions, these findings offer valuable insights for evaluating the microbiological quality of different dairy products, aiding producers in enhancing safety and quality for consumers.

Product Quality and Product Compatibility in Network Industries

Abstract Using an appropriate game-theoretic approach, this article develops a non-cooperative two-stage game in a Cournot duopolistic network industry in which firms strategically choose whether to produce compatible goods in the first decision-making stage. Quality differentiation affects the sub-game perfect Nash equilibrium (SPNE): (i) the network effect acts differently between low- and high-quality firms, depending on their compatibility choice; (ii) if the network externality is positive (resp. negative), to produce compatible (resp. incompatible) goods is the unique SPNE; however, this equilibrium configuration leads the high-quality firm to be worse off; (iii) there is room for a side payment from the high- to the low-quality firm to deviate toward incompatibility (resp. compatibility) under positive (resp. negative) network externality. This payment represents a Pareto improvement on the firm side but not from a societal perspective, as consumers would be worse off. The article also pinpoints the social welfare outcomes corresponding to the SPNE.

A quality detection method of the unbalanced data based on the non‐parameter Log–Log prediction model with the feature extraction

In quality detection, it is important to classify and predict the unbalanced data sets with a high proportion of qualified and unqualified products. There exist already some machine learning methods available. However, these existing methods assume that the samples are evenly distributed among the different classes and ignore the unbalanced characteristics of data. In addition, existing methods cannot be directly applied to high‐dimensional data and cannot accurately express the relationship between data features and the quality of industrial engineering products. In this paper, we propose a new quality detection method of the unbalanced data by establishing a non‐parameter Log–Log classification model. The principal component analysis (PCA) is used to extract the features and reduce the dimension of the original data sets. We develop a sieve maximum likelihood algorithm to obtain the non‐parameter function classifier. The proposed method is applied to the product quality detection of industrial semiconductor manufacturing. The results show the proposed method has high detection performance and classification ability. Compared with traditional machine learning methods, the proposed method has a higher classification accuracy can better describe the relationship between product characteristics and product quality and has a strong generalization ability for different data sets.

DsP-YOLO: An anchor-free network with DsPAN for small object detection of multiscale defects

Industrial defect detection is of great significance to ensure the quality of industrial products. The surface defects of industrial products are characterized by multiple scales, multiple types, abundant small objects, and complex background interference. In particular, small object detection of multiscale defects under complex background interference poses significant challenges for defect detection tasks. How to improve the algorithm’s ability to detect industrial defects, especially in enhancing the detection capabilities of small-sized defects, while ensuring that the inference speed is not overly affected is a long-term prominent challenge. Aiming at achieving accurate and fast detection of industrial defects, this paper proposes an anchor-free network with DsPAN for small object detection. The core of this method is to propose a new idea i.e., feature enhancement in the feature fusion network for the feature information of small-size objects. Firstly, anchor-free YOLOv8 is adopted as the basic framework for detection to eliminate the affections of hyperparameters related to anchors, as well as to improve the detection capability of multi-scale and small-size defects. Secondly, considering the PAN path (top layer of neural networks for feature fusion) is more task-specific, we advocate that the underlying feature map of the PAN path is more vulnerable to small object detection. Hence, we in-depth study the PAN path and point out that the standard PAN will encounter several drawbacks caused by losing local details and position information in deep layer. As an alternative, we propose a lightweight and detail-sensitive PAN (DsPAN) for small object detection of multiscale defects by designing an attention mechanism embedded feature transformation module(LCBHAM) and optimizing the lightweight implementation. Our proposed DsPAN is very easy to be incorporated in YOLO series framework. The proposed method is evaluated on three public datasets, NEU-DET, PCB-DET, and GC10-DET. The mAP of the method is 80.4%, 95.8%, and 76.3%, which are 3.6%, 2.1%, and 3.9% higher than that of YOLOv8 and significantly higher than the state-of-the-art (SOTA) detection methods. Also, the method achieves the second-highest inference speed among the thirteen models tested. The results indicate that DsP-YOLO is expected to be used for real-time defect detection in industry.

Management of data flows in modern industry using blockchain

Recent research and publications. "Industry 4.0" is a concept of the industrial revolu-tion, which is based on the use of modern technologies and digital innovations in production and distribution processes. The introduction of the concept of "Industry 4.0" was designed to improve the competitiveness of European industry and increase productivity and product quality. A blockchain is a distributed data structure that is replicated and distributed among network members. The purpose of the study is to improve automation processes, increase efficiency, re-duce delays and errors in information systems of industry and supply chains by using block-chain technologies in the construction of information systems. Main material of the study. The paper makes an analysis of approaches and algorithms to data management in "Industry 4.0" information systems. Blockchain algorithms are com-pared to classical approach with other databases in the client-server architecture. Conclusions. By implementing algorithms based on blockchain technology, namely by using the Merkle Tree, digital signature technology, and by using consensus algorithms in the framework of decentralized data storage in Distributed Ledger Technology, the processes of automation and efficiency in data flow management are improved, providing a secure and transparent way to store and share data that reduces delays and errors in industry informa-tion systems and supply chains.

High-throughput thermodynamic study of SiC high-temperature chemical vapor deposition from TMS-H2

A systematic study on the tetramethylsilane-hydorgen (TMS-H2) system by high-throughput thermodynamic method is conducted with a large experimental condition range (500 ∼ 2700 K, 0 ∼ 75 kPa, and H2:TMS ratio of 0.1 to 10000) towards the deposition of pure single-crystal SiC by high-temperature chemical vapor deposition (HTCVD) method. The temperature, pressure and H2:TMS ratio dependence phase diagrams are calculated to describe the deposition condition of single-phase “gas + SiC” region. Due to the etching effect of hydrogen on graphite, the increasing temperature and pressure could suppress the formation of graphite and favor the growth of SiC in the H2:TMS ratio above 1000, however, excessive H2:TMS ratio could reduce the TMS consumption efficiency. TMS consumption efficiency maps focusing on the gas + SiC region based on the phase diagrams have been carried out to balance the deposition parameters, SiC mole fraction and TMS consumption efficiency. When H2:TMS ratio > 1000, an accepted window for preparing single-phase SiC with a high TMS consumption efficiency (>90%) is provided, and this window could expand furtherly with increasing the ratio. TMS consumption efficiency maps offer the theoretic basis and guidance for improving the quality and cost of industrial production of single-crystal SiC by HTCVD method.

Enhancement of virtual data quality using pre-trained Bayesian transfer learning under inaccurate and insufficient measurement data

The accurate application of designed dimensions in manufacturing is crucial but complicated by certain limitations in the manufacturing process. Therefore, manufacturing and assembly conditions should be considered in product designs. However, predicting and managing the deformation quality of industrial products, composed of multiple components, is often challenging owing to part deformation during the manufacturing process and structural deformation at the assembly stage. Owing to the advancements in artificial intelligence, highly precise predictive models can be created if sufficient data are available. However, the acquisition of extensive labeled data required for model training can be expensive and time-consuming in the industrial sector. Traditional approaches such as computer-aided engineering, which are commonly used to produce virtual data, have limitations owing to uncertainties in the experimental data, making consistency correction both time-consuming and expensive. This study proposes a method for enhancing the quality of virtual data created in contexts with insufficient and inaccurate experimental data. Engineering properties were extracted from experimental data and integrated with transfer learning using a pre-trained Bayesian neural network. We applied the proposed method to a micro-LED case to predict assembly and manufacturing variations during the design phase under inaccurate and insufficient data. Consequently, the accuracy was significantly improved over the predictive model that predicted the amount of assembly deformation using only experimental data, resulting in a predictive model with an average R-squared prediction accuracy of 98.9% for the deformation resulting from the assembly.

Category-level selective dual-adversarial network using significance-augmented unsupervised domain adaptation for surface defect detection

Surface defect detection is very important to ensure the quality of industrial products. Traditional machine learning cannot be well extended to a non-identically distributed dataset, making surface defect detection a data-limited task. Unsupervised domain adaptation (UDA) can solve this problem by transferring knowledge from a labeled source domain to an unlabeled target domain. Adversarial learning is one of the latest heuristic methods to deal with domain shift in UDA tasks. Although impressive results have been achieved, the adversarial model still suffers from the equilibrium challenge, which may lead to under-transfer or negative transfer. To this end, we utilize joint distribution adaptation to propose a novel UDA model, named significance-augmented-based category-level selective dual-adversarial (CDASA) network, to learn a generalized model. Specifically, to promote positive transfer, we use the selective dual-adversarial (DA) learning strategy to further minimize the feature distribution difference at equilibrium to achieve better domain confusion. Meanwhile, guided by defect recognition performance, the transferability of confusion domain is measured to enhance the distributions of potential domains. In addition, to avoid under-transfer, we consider the relationship between the target data and the decision boundary, and then the significance-augmented (SA) mechanism is proposed to encourage class-level alignment. Thus, the alignment features with domain-invariant and category discrimination can be captured simultaneously. Extensive experiments on collected real industrial datasets and publicly available steel surface defect datasets confirm the effectiveness of our approach.

Industrial Product Quality Analysis Based on Online Machine Learning.

During industrial production activities, industrial products serve as critical resources whose performance is subject to various external factors and usage conditions. To ensure uninterrupted production processes and to guarantee the safety of the production personnel, a real-time analysis of the industrial product quality and subsequent decision making are essential. Conventional detection methods have inherent limitations in meeting the real-time demands of processing large volumes of data and achieving high response speeds. For instance, the regular inspection and maintenance of cars can be time-consuming and labor-intensive if performed manually. Furthermore, monitoring the damage situation of bearings in real time through a manual inspection may lead to delays and may hinder production efficiency. Therefore, this paper presents online machine-learning-based methods to address these two practical problems and simulates them on various datasets to meet the requirements of efficiency and speed. Prior to being fed into the network for training, the data undergo identity parsing to transform them into easily identifiable streaming data. The training process demonstrates that online machine learning ensures timely model updates as small batches of data are sent to the network. The test results indicate that the online learning method exhibits highly stable and effective performance, optimizing the training process.

Automatic recognition of surface defects of hot rolled strip steel based on deep parallel attention convolution neural network

In general, the existence of surface defects in hot rolled strip steel will seriously affect the quality and appearance of industrial products. In order to improve the classification efficiency and accuracy of surface defect recognition. Firstly, a lightweight and high-precision deep learning neural network, deep parallel attention convolutional neural network, is proposed in this paper. In this network, fusion attention module is combined with parallel multi-scale module to focus important feature information and capture context feature information without losing important feature information. Secondly, in order to verify the lightweight of the proposed model, the Parameters, FLOPs, Inference time and FPS of the deep parallel attention convolutional neural network are compared with the current popular classification networks. The experimental results show that the deep parallel attention convolutional neural network can be applied to the actual steel defect detection. The experimental results show that the recognition accuracy, precision and area under the curve of the deep parallel attention convolutional neural network model reach 99.57%, 99.6% and 99.8%, respectively. This method significantly improves the average accuracy of classification and is an effective tool for detecting steel defects.

ICA-Net: Industrial defect detection network based on convolutional attention guidance and aggregation of multiscale features

Detecting surface defects in the industry is essential for improving the quality of industrial products and maintaining product safety. However, problems such as the similarity of defects, significant variation in the scale of the target object, and the balance between detection speed and accuracy in industrial inspection scenarios have been considerable research topics in this field. This paper proposes an industrial defect detection network based on convolutional attention-guided and aggregated multiscale features to address these issues (ICA-Net). Firstly, for similarity defects in complex backgrounds, this paper proposes a backbone network with a combination of lightweight convolutional blocks and self-attentive modules to fully extract images’ local and global information and enhance the network’s expressiveness. Secondly, to make full use of the shallow fine-grained features and deep semantic features of the backbone network to improve the detection capability of defects with significant scale changes, this paper designs a cross-layer multiscale feature fusion network (CEF-Net), which fully fuses the features of adjacent layers and cross-layers through a reweighting feature strategy to enrich the network feature transfer path and ensure the efficient fusion of different scale features in the network. At the same time, the fine-grained feature fusion module (FFM) is used to fuse elements from multiple layers to extract more contextual information, enhance the extraction of fine-grained features and improve the detection capability of complex small targets. Finally, to address the problems of inaccurate regression localization and low detection accuracy of defects in existing industrial algorithms, a new IoU loss function (G-IOU) is proposed for regressing the intersection part of the predicted frame and the actual structure according to the aspect ratio of the real frame during the model regression to improve the accuracy and stability of detection. The experimental results show that 94.1%, 98.6%, 99.4%, 98.8% and 96.5% of mAP@.5 are obtained on steel, PCB, aluminium, automobile and Xsteel steel metal surface defect datasets, respectively, and 48 FPS is achieved, which is superior to the current mainstream detectors and meets the needs of practical industrial production.

Multi-layer film surface functionalized PMF sensor for real-time humidity measurement

Real-time monitoring of environmental relative humidity (RH) plays a crucial role in ensuring both the quality of industrial production and the quality of life. In this work, polarization maintaining fiber (PMF) was used as carriers to creatively constructed multi-layer film surface functionalization on its cladding surface. The indium tin oxide (ITO) film and polyvinyl alcohol (PVA) doped graphite oxide (GO) composite polymer constitute multi-layer humidity sensitive film, which has the effect of dual absorption of water and improves the humidity sensitivity. Microporous composite polymer films created by laser processing increase the area for water evaporation and absorption, which improve response speed, and reduce recovery time. The surface functionalized PMF sensor exhibits high sensitivity and rapid response to humidity, achieving bidirectional repeated measurement of 0.28 nm/% within 10 s in the measurement range of 30%-90%, rendering it a reliable means for real-time monitoring of humidity.

Enhancing Efficiency and Cost-Effectiveness: A Groundbreaking Bi-Algorithm MCDM Approach

Numerous scholars have thoroughly studied the topic of choosing machines considering the progress and technological growth seen in machinery options. This scholarly investigation explores decision-making methods specifically designed to aid the selection of machines in manufacturing businesses. Additionally, this research emphasizes the need for decision-making frameworks in manufacturing facilities, highlighting the importance of smart machine selection strategies in those contexts. In this research, we show a dual-MCDM approach that includes DEX—decision experts—and the EDAS method that are popularly employed to solve decision-making problems in both academic and practical industries. Throughout the previous decade, business leaders and managers increasingly use MCDM solutions to overcome machine selection challenges. At this time, while various decision-support technologies and procedures have been developed and used, it is essential that we discuss the sequence of our study objectives and drive the proposed method for widening use in practical firms. In short, this research may be helpful as a literature review for MDCM studies and related topics. It will also help executives, engineers, and specialists determine which equipment or machines to create and increase product quality in manufacturing and industry.

The Effect of the Aeration Condition on the Liquid–Solid Material Mixing in a Stirred Tank with a Single-Layer Impeller

In order to increase industrial production quality and efficiency, it is essential to understand how the aeration and no-aeration condition affects liquid and solid material mixing in the stirred tank. Due to complicated shear flows, the related mass-transfer mechanism confronts numerous difficulties. This paper put forward an improved computational fluid dynamics and discrete element method (CFD–DEM) modeling approach to explore the effect mechanism of aeration conditions on liquid–solid material mixing. Firstly, a mass-transfer dynamic model is set up with a volume of fluid and piecewise linear interface construction (VOF–PLIC) coupling strategy to explore flow modes and vorticity evolution trends under aeration control. Then, a self-developed interphase coupling interface is utilized to modify the coupling force and porosity of the porous media model in the DEM module, and random dispersion properties of the particle phase under non-aeration and aeration are obtained. Results show that the aeration and flow-blocking components transform fluid tangential speeds into axial and radial speeds, which can improve the material mixing quality and efficiency. The mixed flow field can reach a greater turbulent process under the impeller rotation, making the particles have an intensive disorder and complex flow patterns. The enhanced motion efficiency of the vortex clusters encourages their nesting courses and improves cross-scale mixed transport. It can serve as some reference for the three-phase flow mixing mechanism, vorticity distribution law, and particle motion solution and has a general significance for battery homogeneous mixing, biopharmaceutical processes, and chemical process extraction.