Actual Factory Research Articles

Design and Implementation of Digital Twin Factory Synchronized in Real-Time Using MQTT

As information technology progresses, the need for digital transformation within the industrial sector has become increasingly apparent, and digital twin technology has emerged as a significant trend in manufacturing. Digital twins synchronize physical and digital environments, overcoming spatial and temporal limitations to create various added values that are unattainable in reality. This paper presents a model that integrates digital twin technology with production and operational technologies at manufacturing sites, enabling remote, centrally controlled manufacturing services that transcend physical constraints. Specifically, by utilizing Message Queuing Telemetry Transport (MQTT) for real-time synchronization, this approach ensures efficient and timely data transfer between physical and digital environments. While traditional approaches often encounter challenges due to high investment costs and design complexities, this paper proposes a cost-effective and practical solution that reflects actual factory conditions.

Research and clearance analysis on of steam twin-screw expander employed in indutrial waste heat recovery

In actual factories, screw expanders frequently operate under partial loads.To address the lack of theoretical research and the risks of errors in engineering design associated with steam pressure differential power generation under these conditions,a simulation of the twin-screw expander's working process was developed. The rotor profile was designed, geometric variables such as rotor mesh clearance were calculated, and models for leakage and heat transfer were established. A suction pressure correction model was proposed for the suction process under partial loads, along with correction parameters to simplify simulations.Experimental results demonstrate that the twin-screw expander operates stably under highly fluctuating conditions, demonstrating excellent variable load performance.As the load decreases, the loss of suction pressure reduces the expander's pressure differential, leading to lower leakage and improved efficiency. When the load is reduced from 100 % to 65 %, the volumetric efficiency increases from 78.87 % to 84.76 %, and the isentropic efficiency rises from 66.81 % to 74.25 %.Compared to experimental data, the flow calculation error of the suction pressure correction model presented in this study is controlled within 3 %.This model addresses performance calculations under partial loads, ensuring that theoretical models better align with real-world applications and supporting accurate selection of twin-screw expanders.This model enables appropriate selection of expanders to reduce pressure losses and enhance economic efficiency.

A Simulation Study on the Effect of Filament Spacing on the Temperature Field Uniformity of an HFCVD System

Hot-filament chemical vapor deposition (HFCVD) has become the most widely used ways of preparing diamond film-coated tools due to the simple equipment used, its convenient operation, and its low cost. In the production process of an actual factory, a large number of coated tools need to be prepared in batches. Factors such as the hot-filament arrangement often affect the uniformity of coating on tools, making the performance of the tools prepared in the same batch unstable. This article uses ANSYS R15.0 software software in the context of computational fluid dynamics (CFD) to calculate the temperature field in the HFCVD system, and study the effect of filament spacing on the uniformity of the temperature field of the surface of the substrate. It was found that when the distance between filaments was 14 mm, 10 mm, 10 mm, 8 mm, 8 mm, the temperature field on the surface of the substrate was the most uniform. The experiments are consistent with the results of the simulation, indicating that simulation research has practical significance.

Synthesis of soy-protein adhesives with excellent cohesion and cold-pressing bonding strength through multiple noncovalent crosslinking network of waterborne epoxy resin

In order to achieve the purpose of reducing cost and increasing efficiency of plywood fabrication, development of high-temperature soy meal (HSM)-based adhesives with high cold-pressing bonding strength without compromising its internal cohesion and interfacial adhesion is a meaningful goal in production. However, due to the agglomerated intrinsic structure, weak intermolecular interactions and strong hydrophilicity, the application of HSM adhesive was limited by its poor cohesion and unstable bonding performances. Taking consideration of the abundant hydrogen bonds sites and toughened microphase separation structure, in this work, polyurethane (PU) fragments served to provide bonding strength during prepressing stage. Sulfonate groups were grafted to prepare a stable waterborne epoxy emulsion (WEU) and then used as a crosslinker for HSM adhesives. Besides the introduced carbamate and polyether segments that provided regular hydrogen bonding platforms within HSM matrix, the soft polymer long-chain coils overlapped to increase the initial viscosity of the adhesives. This enhanced cohesive network strengthened the cold-pressing bonding strength of HSM-WPA2–15, reaching 0.76 MPa, which is enough for actual factory requirements. Pressure and thermal condition can stimulate resin penetrating into wood voids and cure it, opening active epoxy groups to covalent bonding within HSM chains. The secondary crosslinking network optimized the water resistance of cured adhesives, achieving 1.00 MPa of the adhesives under wet shear strength testing. This novel and feasible method demonstrates a foundational design, being able to be utilized for developing other plant-based wood adhesives with improved performances.

Biogas production prediction model of food waste anaerobic digestion for energy optimization using mixup data augmentation-based global attention mechanism.

Achieving rapid, efficient, and cost-effective anaerobic digestion (AD) of food waste is a key means to improve the efficiency of food waste treatment. However, in view of the shortage of historical anaerobic digestion data, the limitation of general neural networks in predicting biogas production, and its sensitivity to abnormal variation points, achieving accurate prediction of biogas production is not easy. This paper proposes a novel biogas production prediction model of food waste AD for energy optimization based on the mixup data augmentation integrating an improved global attention mechanism long short-term memory (LSTM). Taking the AD data of the actual factory as samples, the mixup data augmentation is introduced to generate virtual samples with the similar distribution as original samples. Then original samples and generated virtual samples are used as the input of the global attention mechanism LSTM to establish the food waste AD biogas production prediction model. Finally, the proposed method is applied in the biogas production prediction of actual food waste treatment plants. Compared with other industrial modeling models, the experimental results show that the proposed method has the highest prediction accuracy of 0.988, which performs well in predicting biogas production and can effectively guide and timely adjust feed configuration of AD plants.

Environmental impacts of graphene at industrial production scale and its application in electric heating technology

Although graphene has shown great promise in various fields with advantageous performance, the potential environmental impacts of the consequent large-scale production and application are of increasing concern. However, most studies have focused on laboratory-scale or hypothetical mass manufacture of graphene. This study evaluates the environmental impacts of graphene at industrial production scale based on actual factory and its application in electric heating technology using the life cycle assessment. Results show that electricity is the hotspot of environmental impacts for graphene and the novel graphene electric heating element, and their impact categories concentrate on global warming potential, fine particulate matter formation potential, human carcinogenic toxicity potential, and human non-carcinogenic toxicity potential. The thermal exfoliation route has significantly superior potential environmental performances than the oxidation–reduction route in industrial production of graphene. The environmental burdens of graphene and its electric heating applicants would further reduce in the future.

Actual NOX and SOX removal rates in the atmospheric environment of concrete permeable blocks containing TiO2 powders, coconut shell powders, and zeolite beads

This study examined actual nitrogen oxide (NOX) and sulfur oxide (SOX) removal rates of concrete permeable blocks in the atmospheric environment. To give the concrete permeable block a NOX and SOX removal function, TiO2 submicron particle or coconut shell powder was embedded in the surface layer, and zeolite bead was embedded in the base layer. Using an actual factory production line, concrete permeable block specimens were prepared by adding TiO2 submicron particles and coconut shell powder to the surface layer of the concrete permeable block. The prepared concrete permeable block specimens were placed in a parking lot, and then NOX and SOX removal rates were directly measured for 6 months on the concrete permeable block specimens placed in the parking lot. Test results showed that the clear day removal rates of NOX and SOX for the concrete permeable block specimens embedded with TiO2 submicron particles during the day was 84% and 70%, respectively. Compared to the concrete permeable blocks embedded with coconut shell powder, the NOx removal rate of the specimens embedded with TiO2 submicron particles was 1.45 times higher, while the SOx removal rate was similar. However, during cloudy days and nights, the NOX removal rate of the concrete permeable block specimens embedded with TiO2 submicron particles was zero. On the other hand, NOX and SOX removal rates of the concrete permeable block specimens embedded with coconut shell powder and zeolite bead were maintained at approximately 51% and 77% for 6 months, respectively, in the atmospheric environment regardless of weather and day light conditions. Consequently, the coconut shell powder added to the surface layer of the concrete permeable block is a practically promising material in reducing the NOX in the atmospheric environment regardless of weather and day light conditions.

A gray correlation based Bayesian network model for fault source diagnosis of multistage process – Small sample manufacturing system

Fault source diagnosis methodology is one of the key technologies of quality control and assurance for multi-source & multi-stage manufacturing processes, especially in small sample manufacturing systems. By analyzing the existing research on fault source diagnosis methods, a Bayesian network-based methodology is proposed. Gray correlation theory and mechanism analysis method are used in the process of Bayesian network model construction to reduce the dependence of sample data size for structure learning in the process of small sample manufacturing of complex products. In addition, two fault source diagnosis methods based on manufacturing principle analysis and reverse Bayesian network respectively are proposed. The strategy of the combined use of the two methods in the actual manufacturing scenes is given to cope with the fault source diagnosis scenario in the real manufacturing process. In the end, an example from an actual factory is provided to validate the effectiveness and efficiency of the proposed model and methodologies.

A Novel Curve Pattern Recognition Framework for Hot-Rolling Slab Camber

Camber is a typical asymmetrical defect of slabs in the hot-rolling process. The identification of camber plays a key role in improving the quality of the finished strip. To obtain the shape information of the camber, we select the derivative dynamic time warping (DDTW) distance as the measure of curve similarity. An iterative self-organizing data analysis clustering algorithm integrated with DDTW is proposed to divide the sample into different clusters. A curve template is generated via polynomial fitting in each cluster based on the dynamic mechanism of camber. Subsequently, the dynamic time warping (DTW) and DDTW distances are adopted as the input and a weighted random forest (WRF) algorithm is developed for the camber classification to address the sample imbalance problem. Comparative experiments are conducted on the camber sample collected in the actual factory to test the clustering and classification performances. The experimental results indicate that the error rates (ERRs) of four distance-based methods─Euclidean distance, Pearson distance, DTW, and DDTW are 32.7%, 28.5%, 19.1% and 12.3%, respectively, and ERRs of the four classification algorithms—DDTW-SVM, DDTW-KNN, DDTW-RF, and DDTW-WRF are 9.8%, 5.7%, 4.7%, and 2.1%, respectively, which verifies the superior performance of the proposed method in this article.

The Flow Stress-Strain and Dynamic Recrystallization Kinetics Behavior of High-Grade Pipeline Steels.

The hot deformation behavior of high-grade pipeline steels was studied in the strain rate range of 0.001~0.1 s−1 and the temperature range of 1050~1200 °C by using hot compression tests on a Gleeble 3500 thermomechanical simulator. The flow stress increases with the increase in strain rate and the decrease in deformation temperature, and the deformation activation energy is about 358 kJ/mol. The flows stress–strain behavior of the work-hardening and dynamic recovery (DRV) was calculated using the Estrin–Mecking equation, and the kinetics model of the dynamic recrystallization (DRX) was established based on the Avrami equation through characteristic strains. Furthermore, the flow stress–strain behavior of high-grade pipeline steels was predicted by the established model based on the coupling effects of DRV and DRX. The corresponding predicted results are in good agreement with the experimental results according to standard statistical parameters analysis. Finally, the economic strain (ε3) is proposed by the third derivative of the given kinetic model. Based on these calculation results, when the economic strain (ε3) is reached, uniform and refined DRX grains can be obtained, the energy consumption reduced, and the production costs controlled, which is of great significance to actual factory production.

OPC-UA 연동 디지털 트윈 기반의 스마트팩토리 사이버 물리 생산시스템 통합 플랫폼 구축

This paper proposes constructing a digital twin-based CPPS (Cyber Physical Production System) linking OPC-UA that integrates smart factory and production processes. With the development of the CPPS integrated platform, the BOP (Bill of Process) management function was included from the pre-production process planning stage. An IoT-bas768ed MES (Manufacturing Execution System) and SCADA (Supervisory Control And Data Acquisition) were established to monitor resource allocation and status information and process information records of facilities by linking P4R. A 3D CPPS integrated platform was modeled and OPC-UA was integrated to enable the integrated management of sensors and software to predict product production for the entire manufacturing process. Through OPC-UA linkage, the integrated linkage technology of equipment composed of digital twins and detection data of virtual factories, actual factories, and actuator operation information were applied.

Acceleration-based Activity Recognition of Repetitive Works with Lightweight Ordered-work Segmentation Network

This study presents a new neural network model for recognizing manual works using body-worn accelerometers in industrial settings, named Lightweight Ordered-work Segmentation Network (LOS-Net). In industrial domains, a human worker typically repetitively performs a set of predefined processes, with each process consisting of a sequence of activities in a predefined order. State-of-the-art activity recognition models, such as encoder-decoder models, have numerous trainable parameters, making their training difficult in industrial domains because of the consequent substantial cost for preparing a large amount of labeled data. In contrast, the LOS-Net is designed to be trained on a limited amount of training data. Specifically, the decoder in the LOS-Net has few trainable parameters and is designed to capture only the necessary information for precise recognition of ordered works. These are (i) the boundary information between consecutive activities, because a transition in the performed activities is generally associated with the trend change of the sensor data collected during the manual works and (ii) long-term context regarding the ordered works, e.g., information about the previous and next activity, which is useful for recognizing the current activity. This information is obtained by introducing a module that can collect it at distant time steps using few trainable parameters. Moreover, the LOS-Net can refine the activity estimation by the decoder by incorporating prior knowledge regarding the order of activities. We demonstrate the effectiveness of the LOS-Net using sensor data collected from workers in actual factories and a logistics center, and show that it can achieve state-of-the-art performance.

Application of Monitoring Module Three-in-One Microsensor to Real-Time Microscopic Monitoring of Polarizer Sheet in Roll-to-Roll Process

The Roll-to-Roll (R2R) process refers to a high-efficiency, low-cost, continuous production method. The roll material used for processing is a flexible plastic or metal film. In many R2R processes, polarizing films are high-precision products with a high output value. In the production of conventional polarizers, product inspection will only be carried out after the production of the polarizing film is completed. The principal raw material of a polarizer sheet is a hydrophilic polymer, the properties of which may be influenced by water vapor, which degrades its quality. Whether or not the product is impacted can be ascertained by means of a quality inspection, but it must be performed after the process is finished. However, it is already too late when a defective product is detected: the production cost is increased, the schedule is influenced and the delivery date is delayed. The focus of this research was on environmental monitoring of the production drying process oven, but the commercial all-in-one sensors currently on the market cannot tolerate the temperature of the factory’s high-temperature oven. In particular, a commercial flow sensor is rarely suited to high-temperature applications. Some are expensive and cannot be widely distributed. Therefore, this study aimed to develop an integrated sensor to measure the internal environment of the drying process oven for the real-time monitoring of a polarizer sheet in the Roll-to-Roll (R2R) process. This study used micro-electro-mechanical systems (MEMS) technology to make a flexible three-in-one (temperature, humidity and flow) micro-sensor. We monitored the temperature, humidity and flow uniformity in a laboratory oven to simulate the environment of the actual factory oven, with the aim to provide data to confirm whether or not a polarizer sheet has dried. Our system can be monitored instantly by Arduino, or even Raspberry Pi 3, to realize the flexible micro-sensor layout and field verification, in order to optimize the R2R process and to enhance the yield and performance of the polarizer sheet process.

A One-Stage Ensemble Framework Based on Convolutional Autoencoder for Remaining Useful Life Estimation.

As the legislative pressure to reduce energy consumption is increasing, data analysis of power consumption is critical in the production planning of manufacturing facilities. In legacy studies, a machine conducting a single continuous operation has been mainly observed for power estimation. However, the production machine of a modularized line, which conducts complex discrete operations, is more like the actual factory system than an identical simple machine. During the information collection of this kind of production line, it is important to interpret mixed signals from multiple machines to ensure that there is no reduction in the information quality due to noise and signal fusion and discrete events. A data pipeline—from data collection (from different sources) to preprocessing, data conversion, synchronization, and deep learning classification—to estimate the total power use of the future process plan, is proposed herein. The pipeline also establishes an auto-labeled data set of individual operations that contributes to building an power estimation model without manual data preprocessing. The proposed system is applied to a modular factory, connected with machine controllers, using standardized protocols individually and linked to a centralized power monitoring system. Specifically, a robot arm cell was investigated to evaluate the pipeline, with the result of the power profile being synchronized with the robot program.

Operation-Driven Power Analysis of Discrete Process in a Cyber-Physical System Based on a Modularized Factory

As the legislative pressure to reduce energy consumption is increasing, data analysis of power consumption is critical in the production planning of manufacturing facilities. In previous studies, a machine, conducting a single continuous operation, has been mainly observed for power estimation. However, a modularized production line, which conducts complex discrete operations, is more like the actual factory system than an identical simple machine. During the information collection of such production lines, it is important to interpret and distinguish mixed signals from multiple machines to ensure that there is no reduction in the information quality due to noise and signal fusion and discrete events. A data pipeline from data collection from different sources to pre-processing, data conversion, synchronization, and deep learning classification to estimate the total power use of the future process plan is proposed herein. The pipeline also establishes an auto-labeled data set of individual operations that contributes to building power estimation models without manual data pre-processing. The proposed system is applied to a modular factory connected with machine controllers using standardized protocols individually and linked to a centralized power monitoring system. Specifically, a robot arm cell was investigated to evaluate the pipeline with the result of the power profile synchronized with the robot program.

Artificial Intelligence of Manufacturing Robotics Health Monitoring System by Semantic Modeling.

Robotics is widely used in nearly all sorts of manufacturing. Steady performance and accurate movement of robotics are vital in quality control. Along with the coming of the Industry 4.0 era, oceans of sensor data from robotics are available, within which the health condition and faults are enclosed. Considering the growing complexity of the manufacturing system, an automatic and intelligent health-monitoring system is required to detect abnormalities of robotics in real-time to promote quality and reduce safety risks. Therefore, in this study, we designed a novel semantic-based modeling method for multistage robotic systems. Experiments show that sole modeling is not sufficient for multiple stages. We propose a descriptor to conclude the stages of robotic systems by learning from operational data. The descriptors are akin to a vocabulary of the systems; hence, semantic checking can be carried out to monitor the correctness of operations. Furthermore, the stage classification and its semantics were used to apply various regression models to each stage to monitor the quality of each operation. The proposed method was applied to a photovoltaic manufacturing system. Benchmarks on production datasets from actual factories show the effectiveness of the proposed method to realize an AI-enabled real-time health-monitoring system of robotics.

Experimental Validation for Moving Particle Detection Using Acoustic Emission Method

Gas-insulated switchgears (GISs) are important pieces of power equipment used to improve the reliability of power facilities. As the number of GISs increases, more insulation failures occur every year. The most common cause of insulation failure is particles and foreign bodies producing a partial discharge (PD), which causes deterioration of the insulation materials and results in insulation breakdown. However, it is not easy to detect them by conventional PD and ultra-high frequency (UHF) PD measurements because it is difficult to apply the conventional method to the GISs in service, and the UHF method is not always applicable to GISs. Therefore, an appropriate method to detect particles and foreign bodies in GISs is needed. In this study, experimental validation was performed to detect particles moving in GISs using the acoustic emission (AE) method. Acoustic wave signals were produced by the particles moving on the surface of a flat plate when applying voltage. An AE sensor with a frequency range of 50 to 400 kHz was used, and a decoupler and low-noise amplifier were designed to detect the acoustic wave signals with high sensitivity. Twelve types of particles were used, and one was selected to confirm the detectable minimum output voltage. In an actual factory test, the output voltage of the acoustic wave signals was analyzed while considering the applied voltage and signal attenuation. Consequently, it was confirmed that the AE measuring system proposed in this paper could detect particles moving inside GISs.

A Suitable Shape of the Suction Head for a Cleaning Process in a Factory Developed by Computational Fluid Dynamics

The previous shape of the suction head (SH) employed in a cleaning process in a factory had a low performance, removed fewer particles, and generated an annoying noise. Therefore, new shapes of SH have been proposed to solve the issues and the cleaning performance was investigated by the Shear Stress Transport (SST) k-ω turbulence, Discrete Phase (DP), Large Eddy Simulation (LES), and Ffowcs Williams and Hawkings (FW–H) models in a transient state of computational fluid dynamics (CFD). The SST k-ω and DP models were applied to determine the airflow, suspension velocity, cleaning region, and particle trace. In addition, the LES and FW–H models were used to evaluate the noise, sound pressure level, and frequency generated from the proposed shapes. All simulation results were validated with the air velocity and noise measurements and were analyzed to find a suitable shape. The simulation and experimental results revealed that the shapes of the SH affected the cleaning performance and noise generation. The higher the air velocity, the higher the noise generation. The suitable shape delivered a 4.37% better particle removing performance and 11.1 dB less noise generation than the previous shape. The outcomes of this research are the suitable shape of the SH and the research methodology which enabled the application of both CFD and experiments to solve the issue to help enhance the efficiency of the cleaning process in an actual factory.

Developing the Smart Sorting Screw System Based on Deep Learning Approaches

The deep learning technique has turned into a mature technique. In addition, many researchers have applied deep learning methods to classify products into defective categories. However, due to the limitations of the devices, the images from factories cannot be trained and inferenced in real-time. As a result, the AI technology could not be widely implemented in actual factory inspections. In this study, the proposed smart sorting screw system combines the internet of things technique and an anomaly network for detecting the defective region of the screw product. The proposed system has three prominent characteristics. First, the spiral screw images are stitched into a panoramic image to comprehensively detect the defective region that appears on the screw surface. Second, the anomaly network comprising of convolutional autoencoder (CAE) and adversarial autoencoder (AAE) networks is utilized to automatically recognize the defective areas in the absence of a defective-free image for model training. Third, the IoT technique is employed to upload the screw image to the cloud platform for model training and inference, in order to determine if the defective screw product is a pass or fail on the production line. The experimental results show that the image stitching method can precisely merge the spiral screw image to the panoramic image. Among these two anomaly models, the AAE network obtained the best maximum IOU of 0.41 and a maximum dice coefficient score of 0.59. The proposed system has the ability to automatically detect a defective screw image, which is helpful in reducing the flow of the defective products in order to enhance product quality.

A Study on the PLC-Based Pre-Validation Simulation Design Method for Smart Factory

Cyber-physical systems (CPSs) are considered as the main technology in smart factories. In a CPS environment, it is important to monitor the operational processes to detect abnormal conditions and respond to them via real-time comparisons between real-world and virtual models implementing identical real-world factories. The main purpose here is detecting abnormalities in an actual factory and finding quick responses through a simulation model reflecting these abnormalities. This study demonstrates modeling of the key components, such as process equipment and production results, placement of virtualized equipment in virtual factories, synchronization of data between actual and virtual facilities to design digital twins corresponding to the physical assets, and confirming their utilities.