Quality Of Industrial Products Research Articles

Research on steel surface defect detection system based on YOLOv5s-SE-CA model and BEMD image enhancement

ABSTRACT The nondestructive testing capability of steel surface defects has a significant impact on quality of industrial production. However, the single-stage detection network YOLOv5s model for steel surface defect detection is of insufficient feature extraction ability and susceptibility to noise interference in extracting surface defect images. Accordingly, a steel surface defect detection system based on YOLOv5s-SE-CA model and bi-dimensional empirical mode decomposition (BEMD) image enhancement is proposed. Firstly, the BEMD is used for image threshold denoising to improve the quality of surface defect images and achieve preprocessing of the dataset. Secondly, the squeeze and excitation network (SENet) attention mechanism is introduced into the backbone network of YOLOv5s model to adaptively adjust weights of each feature channel. Meanwhile, the coordinate attention (CA) attention mechanism is integrated to improve network model’s ability to focus on surface defects. Therefore, the YOLOv5s-SE-CA model is designed to enhance detection capability of steel surface defects. Experimental verification was carried out using the NEU-DET noisy dataset with specific evaluation indicators of image enhancement detection system. Results showed the proposed system has superior performance when compared to traditional models trained on noisy and denoising datasets and achieved more accurate detection of surface defects in steel.

Power of photo-stress analysis in unravelling the mechanics of granular materials and its applications in interdisciplinary research

Granular materials, such as various grains and granulated powders, are essential in numerous industrial applications and natural processes, including the chemical, pharmaceutical, food, and materials processing industries. They also play a significant role in geotechnical phenomena like the shear-driven collapses of embankments and landslides, which can lead to extensive damage and disruptions. Despite a significant surge in research activities and advancements in discrete element modelling (DEM) tools since the late 1970s, understanding the mechanical characteristics of granular materials remains complex. Experimental techniques such as photo stress analysis (PSA), Nuclear Magnetic Resonance (NMR), and X-ray tomography (XRT) offer valuable insights, yet there is still a lack of detailed information on how granular materials respond mechanically at the microscale under various loading conditions. This knowledge gap affects industrial product quality, equipment safety, production efficiency, operational costs, environmental compliance, and overall safety. In this paper, we initially focus on our scientific contributions to the development of photo stress analysis (PSA) in unravelling the load-transmission characteristics of granular materials of varying sizes and structures (two-dimensional and quasi-three-dimensional) under different mechanical loading environments and applications. Where relevant, discrete element modelling (DEM) complements PSA to enhance our understanding of granular mechanics. We then present the application of PSA to selected key interdisciplinary areas from our recent work, demonstrating the power of PSA. Despite current challenges in applying PSA to three-dimensional granular systems, the results reported here significantly advance our fundamental understanding of the particle-scale and micromechanical characteristics in granular mechanics and particulate engineering. This work highlights the potential of PSA in addressing various inter- and multidisciplinary research problems in the future.

Short-Term Multistep Prediction of Process Industry Product Quality With CNN-BiLSTM Network Based on Parallel Attention Mechanisms

Short-Term Multistep Prediction of Process Industry Product Quality With CNN-BiLSTM Network Based on Parallel Attention Mechanisms

The Impact of Quality Control Implementation on Productivity and Product Quality in Industry

Furniture production is one of the activities that has the potential to generate income for educational institutions. The furniture production process is generally carried out by experienced students utilizing industry-standard facilities, namely the Wood Construction workshop along with its machinery equipment. Therefore, the implementation of Quality Control in the production process is necessary. This study aims to determine the impact of Quality Control implementation on improving the productivity and quality of furniture products. The method used is quantitative with 4 stages of Quality Control, including product design, process planning, Incoming Quality Control (IQC), and In-Process Quality Control (IPQC). Productivity measurement is divided into 3 parts: partial productivity, total factor productivity, and total productivity, which includes labor productivity, time, and production output. Meanwhile, product quality measurement is obtained from assessments by practitioners and product quality testing according to ISO 19682. Data is processed using the path analysis method with the help of a smart PLS application. The research results show that the productivity of production output is 87.5%, categorized as good, with labor productivity of 0.125 pieces/person/day and work hour productivity of 0.25 pieces/hour/day. Product quality based on production processes is categorized as good at 83%. Meanwhile, product quality testing, including visual aspects is good (82.3%), product stability (normal), strength (normal), durability (normal), stiffness (normal 34mm/m), and drop test (normal). Based on path analysis, it is known that quality control including product design, process planning, Incoming Quality Control (IQC), and In-Process Quality Control (IPQC) has an impact on improving the productivity and quality of furniture products. the quality control carried out to increase productivity and product quality can be said to be good with the validation of the final result of 0.188 including in the good category.

Chemical composition prediction in goji (Lycium barbarum) using hyperspectral imaging and multi-task 1DCNN with attention mechanism

The bioactive components of goji berries (Lycium barbarum) are crucial determinants of their nutritional and commercial value. In this study, we combined hyperspectral imaging technology (HSI) with a one-dimensional convolutional neural network (1DCNN) to predict the content of chemical compositions in goji. To enhance the model's ability to focus on relevant information, we introduced the channel attention module (CAM), spectral attention module (SAM) and fused them together, which can focus on output features of the convolution kernels differently and adaptively emphasize more effective spectral bands, respectively. Moreover, considering the limitations of traditional single-task prediction methods and the inherent correlations among different constituents, we employed a multi-task CNN for the simultaneous prediction of various goji constituents. The results indicate that the attention-enhanced 1DCNN model outperforms both the partial least squares regression (PLSR) model and the vanilla 1DCNN. With multi-task learning, the model achieves optimal performance, achieving an average R2 of 0.9435 for the prediction of the three components. Our research develops an efficient and accurate method for predicting the constituents of goji berries, providing a valuable, convenient, and effective tool for assessing and detecting the quality of fruit products in the food industry.

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

CRGF-YOLO: An Optimized Multi-Scale Feature Fusion Model Based on YOLOv5 for Detection of Steel Surface Defects

The identification of imperfections on steel surfaces is vital for ensuring the quality of industrial products. It requires the capability of real-time detection with high accuracy. This paper proposes the CRGF-YOLO (Contextual Reparameterized Generalized Feature) model based on YOLOv5. In the network, BottleneckCSP structures and depthwise separable convolutions utilizing the structural reparameterization are introduced to reduce the model size and improve performance. In addition, contextual transformer modules are employed as self-attention mechanisms to improve feature representations by capturing long-range dependencies, outperforming conventional convolutional networks. Furthermore, the simplified generalized feature pyramid network is embedded to aggregate multi-scale feature maps and enhance the network’s robustness. Finally, four prediction heads with different sizes are employed to predict defects, which are supported by prior bounding boxes generated using k-means clustering algorithm. The Focal-EIOU (Exponential Intersection over Union) loss function is introduced to improve detection accuracy and expedite model convergence. The improved model achieves a mean average precision (mAP) of 82.2% on the NEU-DET dataset, outperforming the baseline YOLOv5s by 7.7% mAP while maintaining real-time speeds. Comparative evaluations demonstrate CRGF-YOLO’s superior performance over previous state-of-the-art methods like Faster R-CNN (77.4% mAP), YOLOv3 (77.4% mAP), YOLOv7s (72.1% mAP), and YOLOv8s (78.7% mAP) for steel surface defect detection. Overall, this study provides valuable insights and practical guidance for the advancement of defect detection technology.

Improving efficiency and quality of operational industrial production assets information management in customer–vendor interaction

Industrial assets management has a key role in ensuring continuity of industrial production via purchasing, warehousing and maintenance of devices and spare parts used in manufacturing, but it suffers from data and information accessibility, completeness, consistency, interoperability, and timeliness challenges in the customer - equipment vendors interaction and information sharing. Current practices in devices and spare parts purchasing, warehousing and maintenance include a lot of manual information search, validation and update work done at the manufacturing companies to maintain high quality master data records for the production assets management. This article addresses the challenge of improving efficiency and quality of operational production assets information management in industry customer company interaction with their production equipment vendors, with a design science research approach. The research was executed in collaboration between research organizations, forest industry companies, industrial production equipment vendors and information technology (IT) companies on two parallel and complementary research tracks, which are described with their key research results. The contributions are an ex-ante evaluated digital marketplace concept with preliminary requirements, and an ex-post evaluated semi-automated master data harmonization tool, as artefacts for improving efficiency and quality of production assets information management in industrial customer–vendors interaction. As theoretical contribution, the article provides a synthesis of the research results in the framework of information systems (IS) design theory as a nascent and novel design theory on designing IS artifacts for improving efficiency and quality of production assets information management, both in single industrial customer company and ecosystem settings. The contributions provide a starting point for further research and development of a digital marketplace for improving production assets information management in industrial customers – vendors interaction, potentially also applicable for similar ecosystems in other continuous production process -based manufacturing industries.

Industrial automation and product quality: the role of robotic production transformation

ABSTRACT We investigate the impact of industrial automation, specifically robots, on product quality in international trade. By analysing detailed data from Chinese listed firms, we examine how the adoption of robots is associated with higher product quality compared to that of nonrobot users. This effect is more pronounced among heterogeneous products and in ordinary trade patterns. We identify several mechanisms through which robots contribute to product quality improvements, including resource reallocation, adaptability to complex production processes, and enhancement of input quality. Moreover, we demonstrate that replacing low-skilled labour with complex robots and employing high-skilled labour to collaborate with robots improved product quality, confirming that robotic displacement and reinstatement effects generate better product quality upgrading. Overall, our empirical evidence supports the positive influence of robot adoption on product quality and provides insights into the mechanisms underlying robotic production transformation.

Efficient packaging defect detection: leveraging pre-trained vision models through transfer learning

The inspection of packaging defects is a crucial aspect of maintaining the quality of industrial production, especially in the case of boxed products. This study introduces a novel approach for detecting physical defects in product packaging boxes by integrating image processing with deep learning, specifically transfer learning with two images as an input. The proposed method utilizes both top view and side view images of the packaging to determine its condition, a significant departure from the conventional single image input. Our approach incorporates 16 pre-trained model variants from EfficientNetV2, MobileNetV3, and ResNetV2 for transfer learning as feature extractors. The experimental findings demonstrate that the best model that leverages EfficientNetV2 variant achieves 100% accuracy and F1 score in terms of classification performance. However, the most optimal model in terms of classification performance and inference speed was the one that leveraged ResNetV2 variant. This model scored 95% accuracy and 95.24% F1 score, with an inference speed of 91 ms per prediction.

Meta learning based residual network for industrial production quality prediction with limited data

Due to the challenge of collecting a substantial amount of production-quality data in real-world industrial settings, the implementation of production quality prediction models based on deep learning is not effective. To achieve the goal of predicting production quality with limited data and address the issue of model degradation in the training process of deep learning networks, we propose Meta-Learning based on Residual Network (MLRN) models for production quality prediction with limited data. Firstly, the MLRN model is trained on a variety of learning tasks to acquire knowledge for predicting production quality. Furthermore, to obtain more features with limited data and avoid the issues of gradient disappearing or exploding in deep network training, the enhanced residual network with the effective channel attention (ECA) mechanism is chosen as the basic network structure of MLRN. Additionally, a multi-batch and multi-task data input approach is implemented to prevent overfitting. Finally, the availability of the MLRN model is demonstrated by comparing it with other models using both numerical and graphical datasets.

Dual-branch information extraction and local attention anchor-free network for defect detection

In the production process, the presence of surface defects seriously affects the quality of industrial products. Existing defect detectors are not suitable for surface with scattered distribution and complex texture of defects. In this study, a dual-branch information extraction and local attention anchor-free network for defect detection (DLA-FCOS), which is based on the fully convolutional one-stage network, is proposed to accurately locate and detect surface defects of industrial products. Firstly, a dual-branch feature extraction network (DFENeT) is proposed and used to improve the extraction ability of complex defects. Then, a local feature enhancement module is proposed, and a residual connection is established to enrich local semantic information. Meanwhile, the self-attention mechanism is introduced to form local attentional residual feature pyramid networks (LA-RFPN) to eliminate the influences of feature misalignments. The mean average accuracy (mAP) and frames per second (FPS) of the proposed DLA-FCOS on the cut layer of the tobacco packet defect dataset (CLTP-DD) are 96.8% and 20.7, respectively, which meets the requirements for accurate and real-time defect detection. Meanwhile, the average accuracy of the proposed DLA-FCOS on the NEU-DET and GC10-DET datasets is 78.4% and 67.7%, respectively. The results demonstrate that the DLA-FCOS has good feasibility and high generalization capability to perform defect detection tasks of industrial products.

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.

Experiential Marketing and Quality Drive Coffee Shop Loyalty in Indonesia

This study analyzes the effects of Experiential Marketing, Product Quality, and Service Quality on consumer loyalty at a coffee shop in Pandaan. Using an associative quantitative approach, data from 100 consumers were collected via questionnaires and processed with SPSS. Results indicate that all three factors positively and significantly impact consumer loyalty. These findings suggest that improving these aspects can enhance customer loyalty. The study fills a knowledge gap by providing empirical evidence on the importance of experiential marketing, product quality, and service quality in the coffee shop industry. Highlight: Experiential Marketing positively impacts consumer loyalty. High product quality boosts customer retention. Excellent service quality enhances repeat purchases. Keywoard: Experiential Marketing, Product Quality, Service Quality, Consumer Loyalty, Coffee Shop

Development of an algorithm for evaluating the comprehensive ergonomic indicator of hand tools

Objective: development of an algorithm for evaluating the comprehensive ergonomic indicator of hand tools to reduce physical strain during their use.Materials and Methods. The algorithm development utilized the nomenclature of ergonomic indicators specified in DSTU 7895:2015 "Design and Ergonomics. Rules for evaluating the ergonomic level of quality of industrial products," which is categorized into two levels. The first level includes usability, control and monitoring convenience, product planning, serviceability, product hygiene, and safety. The second level involves physical and mental workload, fatigue, and compliance with anthropometric parameters.Results. An algorithm for evaluating the comprehensive ergonomic indicator of hand tools has been developed, consisting of five steps: compiling a list of ergonomic characteristics of hand tools according to production tasks, selecting a group of five or more experts, conducting preliminary assessment of the product's comprehensive ergonomic indicator, analyzing expert assessments, discussing strengths and weaknesses, construction, and making decisions regarding the choice of weight coefficients, verifying the obtained results, and developing recommendations for application conditions for a specific product. The distinctive feature of the proposed algorithm lies in establishing weight coefficients of the tool's characteristics considering safety, convenience, and working conditions. The results of evaluating the comprehensive ergonomic indicator of angle grinders are presented, allowing for the effectiveness of the developed algorithm based on feedback from study participants and providing recommendations for its further improvement.The novelty lies in establishing correlations between two levels of ergonomic indicators through weight coefficients, allowing for the consideration of the complexity of working conditions, safety, and convenience of hand tools.Practical value: сhecklists for processing score assessments based on established ergonomic criteria have been developed.

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.

Единая промышленная политика Союзного государства Белоруссии и России как инструмент обеспечения экономической безопасности

The issue of ensuring the economic security of the Union State and its industrial independence is a priority in the context of sanctions against Russia and Belarus, as well as outright economic pressure from Western countries. The assessment of the current level of economic security of the integration association based on three groups of indicators – the quality of life of the population, the state of foreign trade cooperation, the quality of industrial production and cooperation – revealed an insufficient level of development of industrial cooperation, investment, support for the introduction of R&amp;D, high dependence on imports of goods of the machine-building complex, etc. In addition, the dynamics of the values of most indicators is negative, which indicates the need to develop new economic policy measures, including at the interstate level. The key condition for the self-sufficiency of the Union State in necessary goods, including high-tech ones, is the qualitative development of the industrial sector through active assistance from both state and supranational bodies. Measures taken at the interstate level should synergistically strengthen support at the State level. It is essential to understand the level of demand for critical goods (including raw materials and components for the production of new technological goods) and the necessary resources for the creation and development of their production facilities on the territory of the Union State (including scientific developments, technologies and human resources).