Intelligent Factory Research Articles

Distributed deep learning approach for intrusion detection system in industrial control systems based on big data technique and transfer learning

ABSTRACT Industry 4.0 refers to a new generation of connected and intelligent factories that is driven by the emergence of new technologies such as artificial intelligence, Cloud computing, Big Data and industrial control systems (ICS) in order to automate all phases of industrial operations. The presence of connected systems in industrial environments poses a considerable security challenge, moreover with the huge amount of data generated daily, there are complex attacks that occur in seconds and target production lines and their integrity. But, until now, factories do not have all the necessary tools to protect themselves, they mainly use traditional protection. In order to improve industrial control systems in terms of efficiency and response time, the present paper propose a new distributed intrusion detection approach using artificial intelligence methods, Big Data techniques and deployed in a cloud environment. A variety of Machine Learning and Deep Learning algorithms, basically convolutional neural networks (CNN), have been tested to compare performance and choose the most suitable model for the classification. We test the performance of our model by using the industrial dataset SWat.

The digitization work of cement plant in China

This article is to introduce the digitization work of smart cement plant in China. To save power and promote the product quality, delicacy management in variety ways of digital transformation is the best tool for the new built, or existing cement plants. New technologies, such as special integrated robots, online equipment, information plant form, cement process digital applicant (SAAS), Artificial Intelligence (AI) and Advanced Process Control (APC) are implemented to approve the traditional ways of operation.This paper introduces the application of intelligent factory technology in China, and focuses on the intelligent optimization control system. The system has self-learning and global optimization functions, with strong adaptability and robustness. In combination with quality prediction, kiln condition identification and quantitative identification of fire watching TV video images, the kiln system is feed-forward controlled to solve the control problem of long time delay, large inertia and strong coupling of the burning system. The system gives the global optimal recommended value of the control target through calculation, with which to ensure the best control and operation of the system.With the application of comprehensive intelligent construction, the cement plant will usually save 2–5 % of energy and 20 % staff simplification, and make the management mode more concise and efficient. Meanwhile, the cement groups are also developing the production information plat form, by merging the Enterprise Resource Planning (ERP) data, sales business to lower down the total cost. This paper tries to introduce the digital efforts that are carried out in the smart cement plants and shows the methodology of implementation.

Real-time multi-factory scheduling in Industry 4.0 with virtual alliances

The ever-changing and dynamic market environment requires applying job shop systems based on real-time data. The establishment of physical-virtual systems in the production process has led to the emergence of intelligent factories. Compared with those employing traditional production methods, such factories manufacture products with higher quality, higher production speed, and other economic benefits. Regarding the virtual connections of factories, events such as the arrival of new jobs, and machine breakdowns, are identified by Radio Frequency Identification System between different production units, and related decisions are made quickly and carefully. In this intelligent job shop scheduling, it is assumed that independent factories create virtual production networks in which, each factory focuses on its own interests. Regarding the importance of this issue in today’s industry, this research explores the real-time scheduling problem in multi-agent production networks distributed in smart factories. Since this problem is a combination of static scheduling and real-time scheduling, a bi-objective model of mixed-integer linear programming is first developed. An approach is then proposed to solve the dynamic real-time scheduling problem. In addition, a learning-based memetic algorithm for solving large-size bi-objective instances is proposed due to the NP-hardness of the considered problem. Afterward, the results of the proposed algorithm are compared with the hybrid Pareto-based tabu search algorithm. The computational results show that in large-size instances, the proposed algorithm outperforms the competing algorithm.

Federated Deep Reinforcement Learning for Energy-Efficient Edge Computing Offloading and Resource Allocation in Industrial Internet

Industrial Internet mobile edge computing (MEC) deploys edge servers near base stations to bring computing resources to the edge of industrial networks to meet the energy-saving requirements of Industrial Internet terminal devices. This paper considers a wireless MEC system in an intelligent factory that has multiple edge servers and mobile smart industrial terminal devices. In this paper, the terminal device has the choice of either offloading the task in whole or in part to the edge server, or performing it locally. Through combined optimization of the task offload ratio, number of subcarriers, transmission power, and computing frequency, the system can achieve minimum total energy consumption. A computing offloading and resource allocation approach that combines federated learning (FL) and deep reinforcement learning (DRL) is suggested to address the optimization problem. According to the simulation results, the proposed algorithm displays fast convergence. Compared with baseline algorithms, this algorithm has significant advantages in optimizing the performance of energy consumption.

An Optimization-Based High-Precision Flexible Online Trajectory Planner for Forklifts

There are numerous prospects for automated unmanned forklifts in the fields of intelligent logistics and intelligent factories. However, existing unmanned forklifts often operate according to offline path planning first followed by path tracking to move materials. This process does not meet the needs of flexible production in intelligent logistics. To solve this problem, we proposed an optimized online motion planner based on the output of the state grid as the original path. Constraints such as vehicle kinematics; dynamics; turning restriction at the end of the path; spatial safety envelope; and the position and orientation at the starting point and the ending point were considered during path optimization, generating a precise and smooth trajectory for industrial forklifts that satisfied non-holonomic vehicle constraints. In addition, a new rapid algorithm for calculating the spatial safety envelope was proposed in this article, which can be used for collision avoidance and as a turning-angle constraint term for path smoothing. Finally, a simulation experiment and real-world tray-insertion task experiment were carried out. The experiments showed that the proposal was effective and accurate via online motion planning and the tracking of automated unmanned forklifts in a complicated environment and that the proposal fully satisfied the needs of industrial navigation accuracy.

Hybrid Optimization of Green Supply Chain Network and Scheduling in Distributed 3D Printing Intelligent Factory

Considering the advantages of 3D printing, intelligent factories and distributed manufacturing, the 3D printing distributed intelligent factory has begun to rise in recent years. However, because the supply chain network of this kind of factory is very complex, coupled with the impact of customized scheduling and environmental constraints on the enterprise, the 3D printing distributed intelligent factory is facing the great challenge of realizing green supply chain networks and optimizing production scheduling at the same time, and thus a theoretical gap appears. This paper studies the hybrid optimization of green supply chain networks and scheduling of the distributed 3D printing intelligent factory. Firstly, according to the green supply chain network architecture of the distributed 3D printing intelligent factory, the cost minimization model is constructed. Secondly, mathematical software is used to solve the model, and the scheduling plan can be worked out. Finally, through the simulation analysis, it is concluded that the influencing factors such as demand, factory size and production capacity complicate the production distribution, and it can be observed that the carbon emission cost has gradually become the main factor affecting the total cost. The study has a reference value for the management decision making of the distributed 3D printing intelligent factory under the background of carbon emissions.

Strategies for the development of industry 4.0 and the formation of effective demand

The article is devoted to the development of industry 4.0 and the formation of effective demand. The authors are convinced that industry 4.0 is essentially a strategy of intelligent factories, i.e. the process of managing the production chain is carried out in real time using information and communication technologies, as well as the installation of robots. Effective demand for goods depends not only on the income of the buyer, but also on the prices of innovative products that are structurally formed by industry 4.0. It is proved that the main reason for the economic constraints of the development of industry 4.0 is the low effective demand for high-tech products. Therefore, the Russian industry must move to new types of economic activity and ensure the growth of productivity potential through the use of smart technologies, collaborative space, the formation of intelligent operational units and the digitalization process. Achieving sustainable long-term growth will depend on the ability to build a stronger domestic demand base and a balancing strategy: customer satisfaction and cost savings. The authors, based on international experience, prove the feasibility of forming a differentiated core, i.e. an effective combination of products and services formed by industry 4.0 while simultaneously integrating cyber-physical systems and the Internet of Things, as well as services. The analysis shows the huge growth potential of industry 4.0 as the revenue growth of leading international companies is estimated in the range of 8.1–13.6 % in the next five years. At the same time, an annual increase in customer spending on products is projected by 2.4 %. The conducted research shows that industry 4.0 allows for an annual increase in efficiency by an average of 3.3 % due to increased productivity, energy efficiency and resource efficiency. The authors are convinced that such a setup is impossible without the accumulation of capital and investments, which are the main factors contributing to economic growth and sustainable solvency of customers.

A Diffusion Model Based on Network Intrusion Detection Method for Industrial Cyber-Physical Systems.

Industrial Cyber-Physical Systems (ICPS) connect intelligent manufacturing equipment equipped with sensors, wireless and RFID communication technologies through data interaction, which makes the interior of the factory, even between factories, become a whole. However, intelligent factories will suffer information leakage and equipment damage when being attacked by ICPS intrusion. Therefore, the network security of ICPS cannot be ignored, and researchers have conducted in-depth research on network intrusion detection for ICPS. Though machine learning and deep learning methods are often used for network intrusion detection, the problem of data imbalance can cause the model to pay attention to the misclassification cost of the prevalent class, but ignore that of the rare class, which seriously affects the classification performance of network intrusion detection models. Considering the powerful generative power of the diffusion model, we propose an ICPS Intrusion Detection system based on the Diffusion model (IDD). Firstly, data corresponding to the rare class is generated by the diffusion model, which makes the training dataset of different classes balanced. Then, the improved BiLSTM classification network is trained on the balanced training set. Extensive experiments are conducted to show that the IDD method outperforms the existing baseline method on several available datasets.

How Industry 4.0 and Sensors Can Leverage Product Design: Opportunities and Challenges.

The fourth industrial revolution, also known as Industry 4.0, has led to an increased transition towards automation and reliance on data-driven innovations and strategies. The interconnected systems and processes have significantly increased operational efficiency, enhanced organizational capacity to monitor and control functions, reduced costs, and improved product quality. One significant way that companies have achieved these benefits is by integrating diverse sensor technologies within these innovations. Given the rapidly changing market conditions, Industry 4.0 requires new products and business models to ensure companies adjust to the current and future changes. These requirements call for the evolutions in product design processes to accommodate design features and principles applicable in the current dynamic business environment. Thus, it becomes imperative to understand how these innovations can leverage product design to maximize benefits and opportunities. This research paper employs a Systematic Literature Review with Bibliometric Analysis (SLBA) methodology to explore and synthesize data on how Industry 4.0 and sensors can leverage product design. The results show that various product design features create opportunities to be leveraged to guarantee the success of Industry 4.0 and sensor technologies. However, the research also identifies numerous challenges that undermine the ongoing transition towards intelligent factories and products.

Intelligent Retrofitting Paradigm for Conventional Machines towards the Digital Triplet Hierarchy

Industry 4.0 is evolving through technological advancements, leveraging information technology to enhance industry with digitalisation and intelligent activities. Whereas Industry 5.0 is the Age of Augmentation, striving to concentrate on human-centricity, sustainability, and resilience of the intelligent factories and synergetic industry. The crucial enhancer for the improvements accomplished by digital transformation is the notion of ‘digital triplet D3’, which is an augmentation of the digital twin with artificial intelligence, human ingenuity, and experience. digital triplet D3 encompasses intelligent activities based on human awareness and the convergence among cyberspace, physical space, and humans, in which Implementing useful reference hierarchy is a crucial part of instigating Industry 5.0 into a reality. This paper depicts a digital triplet which discloses the potency of retrofitting a conventional drilling machine. This hierarchy included the perceptive level for complex decision-making by deploying machine learning based on human ingenuity and creativity, the concatenated level for controlling the physical system’s behaviour predictions and emulation, the observing level is the iterative observation of the actual behaviour of the physical system using real-time data, and the duplicating level visualises and emulates virtual features through physical tasks. The accomplishment demonstrated the viability of the hierarchy in imitating the real-time functionality of the physical system in cyberspace, an immaculate performance of this paradigm. The digital triplet’s complexity was diminished through the interaction among facile digital twins, intelligent activities, and human awareness. The performance parameters of the digital triplet D3 paradigm for retrofitting were eventually confirmed through appraising, anomaly analysis, and real-time monitoring.

AI-Smart Factory: Design and verification of Korean SME AI smart factory using level diagnosis system

This paper presents a new AI-based smart factory construction model for SMEs based on the Korean manufacturing industry and proposes an effective construction process that implements the informatization, automation, and intelligentization characteristics of smart factories. The Korean AI-based smart factory construction process starts with the accumulation of digital information by attaching sensors to the manufacturing processs, and then creates a customized AI model for the company through cloud accumulation of sensor data and big data analysis through CNN techniques. The Korean SME Smart Factory is an AI-based smart factory through AHP analysis and smart factory level diagnosis system, which is a construction process, which means a consumer-oriented intelligent factory that combines next-generation digital technology and manufacturing technology beyond the existing level of factory automation (FA). Various products can be produced on one production line, and modularization is expected to change from mass orders to individual flexible production systems. The transition to smart factories is expected to dramatically improve the productivity of manufacturing, as well as realize energy savings and a more human-centered work environment. By monitoring and controlling manufacturing sites in a virtual space, it is expected to facilitate factory management and lead to improved quality and cost competitiveness. Keywords: AI Smart Factory, Level Diagnosis System, Korean SMEs, AHP Analysis

Research on Steel Surface Defect Detection Algorithm Based on Improved Deep Learning

With the development of industrial economy, more and more enterprises use machine vision and artificial intelligence to replace manual detection. Therefore, the research of steel surface defect detection based on artificial intelligence is of great significance to promote the rapid development of intelligent factory and intelligent manufacturing system. In this paper, Yolov5 deep learning algorithm is used to build a classification model of steel surface defects to realize the classification and detection of steel surface defects. At the same time, on the basis of Yolov5, combined with the attention mechanism, the backbone network is improved to further improve the classification model of steel surface defects. The experiment shows that the Recall and mAP of improved Yolov5 perform better on the steel surface defect data set. Compared with Yolov5, the number of C3CA-Yolov5 parameters decreased by 13.02%, and the size of pt files decreased by 12.72%; the number of C3ECA-Yolov5 parameters decreased by 13.36%, and the size of pt files decreased by 13.22%.

Research on the agile management system of workshop production resources

Industry 4.0 has a key technical point that the intelligent factory uses the ‘ raw materials ’ containing information to complete the production tracking. The high-performance real-time positioning system is an important part of the real-time tracking and positioning of materials in the industrial Internet of Things. Passive Ultra High Frequency (UHF) Radio Frequency Identification (RFID) technology has become the preferred solution for the industrial 4.0 material real-time positioning system. This paper first analyzes the RFID positioning technology, gives the process architecture of workshop manufacturing resource RFID real-time information-driven production line and the classification and management scheme of multi-source real-time manufacturing information, and realizes the RFID tracking and positioning of workshop manufacturing resource.

Preface

The 3rd International Conference on Advanced Materials and Intelligent Manufacturing (ICAMIM 2022) was successfully held in Guangzhou, China from August 5-7, 2022. ICAMIM 2022 provided a high-level technical exchange platforms for experts, scholars, and entrepreneurs worldwide to communicate face-to-face with each other, strengthening the exchange of advanced materials and intelligent manufacturing technologies in various countries around the world.We were greatly honored to have both Prof. Kaiming Wu from Wuhan University of Science and Technology, China and Prof. Liejun Li from South China University of Technology, China to serve as our Conference Chairmen. 120 individuals attended the conference. The conference agenda was composed of keynote speeches and oral presentations. Firstly, keynote speakers were each allocated 30-45 minutes to hold their speeches. Then in the oral presentations, the excellent papers we had selected were presented by their authors by turns.At the beginning of the conference, nine sophisticated keynote speakers were invited to deliver their speeches. Among them, Prof. Mingxin Huang from The University of Hong Kong, China made a report on the Hydrogen Embrittlement of 1500 and 2000 MPa Press-Hardened Steel. This presentation compared the hydrogen embrittlement of two PHS grades with ultimate tensile strength of 1500 and 2000 MPa respectively. And it was found that hydrogen embrittlement depends strongly on the deformation strain rate. In addition, Prof. Tongguang Zhai from University of Kentucky, USA showed us his research: Physical Models for Quantification of Fatigue Crack Initiation Life in Engineering Alloys Based on Microstructure. He introduced the relationships between fatigue crack initiation and 3-D microstructure discovered recently in high performance Al alloys and established two physical models to quantify the stochastic behaviors of fatigue crack initiation. Last but not least, Researcher. Yu Chen from Institute of Intelligent Manufacturing, GDAS delivered a speech on the Industrial Internet Technologies and Manufacturing Digital Transformation. This speech introduced the concepts, intension and features of Industrial Internet, expounded the close relationship between Intelligent Manufacturing and Industrial Internet, and illustrated the typical Industrial Internet application scenarios in the Intelligent Factory and Manufacturing Digital Transformation. The brilliant speeches of each keynote speaker had triggered heated discussion in the conference. Moreover, their speeches won the praise of every participant for disseminating useful and insightful knowledge.The proceedings of ICAMIM 2022 cover various topics, which include Non-ferrous Metal Materials, Ceramic Metallurgy, Biomimicry Mechanisms, Engineering Optimization, Digital Manufacturing, etc. All the papers have been checked through rigorous review and processes to meet the requirements of publication.On behalf of the conference organizing committee, we’d like to show our congratulation and sincere gratitude to all the keynote speakers, peer reviewers, and all the participants. In addition, our special thanks go to the Journal of Physics: Conference Series, for all efforts of the colleagues in publishing this paper volume. The help and contribution of each individual and institution was instrumental in the success of the conference.Committee of ICAMIM 2022List of Committee member is available in this Pdf.

Selectively Powering Multiple Small-Size Devices Spaced at Diffraction Limited Distance With Point-Focused Electromagnetic Waves

The issue of accurately distinguishing and selectively powering multiple closely spaced devices is challenging but vital for wireless power transfer (WPT). In this article, a continuous-wave time-reversal selective radiative wireless power transfer (CWTR-SRWPT) method based on a multipoint focusing idea is proposed. With the proposed method, we realize a selective radiative WPT prototype system, which can arbitrarily select multiple devices spaced at a diffraction limited distance and power them simultaneously. The accurate positioning and high-precision spatial selection of the closely spaced energy-dissipated devices are achieved by the multipoint focusing field (MPFF) without any auxiliary positioning equipment. The MPFF is produced by a set of weight-optimized single-point focusing fields that are generated by the fundamental time-reversal technique. Theoretical analysis and full-wave simulations to the performances of the CWTR-SRWPT system are presented. The experimental demonstrations of selectively powering multiple LEDs are also conducted. The results indicate that the proposed method exhibits exciting performances of high-precision selectivity, accurate positioning, and satisfactory over-the-air efficiency. It demonstrates potential applications in selectively powering the wireless sensors distributed densely in future intelligent factories.

Risk Related to AGV Systems—Open-Access Literature Review

AGV systems are considered a necessary element of the recently created cyber-physical systems whose task is to support material flows in intelligent factories. The implementation of AGV systems in cyber-physical systems generates new types of threats, and the level of risk of adverse events related to the operation of these systems is changing. This article aims to present the results of an open-access literature review, the purpose of which was to identify issues addressed in studies of the analysis and risk management of AGV systems, to group publications according to the proposed topical framework, and to identify gaps in the current research. We used the PRISMA concept to review the literature. After completing all stages of the procedure, 153 publications on risk in AGV systems were selected for analysis. To ensure the high quality of the analysis, access to the full texts of the selected documents was necessary. For this reason, we have limited the collection to open-access publications. The results of the analysis allowed us to distinguish 12 primary categories and 26 additional subcategories within the literature. These categories are characterized by different numbers of publications and cover different AGV systems operation areas. The simultaneous assignment of documents to several subcategories indicates the need for a holistic view of issues related to the risk of AGV operation in operating environments with varying degrees of automation. An essential scientific contribution of our analysis is the identification of a research gap. The analyzed publications lacked studies on methods of risk identification and assessment that take into account the specificity of AGV operation, especially in the cyber-physical systems of Logistics 4.0.

Temperature-adaptable pressure sensors based on MXene-coated GO hierarchical aerogels with superb detection capability

With the broader application of intelligent sensors in various complex conditions, sensing materials with better mechanical and conductive properties as well as higher temperature adaptability are required. For this purpose, a GO/PVA/MXene (GPM) composite aerogel was designed and prepared via a two-step strategy, including self-assembly and vacuum-assisted dip-coating processes. The as-prepared GPM has a unique structure composed of hierarchical GO network skeleton and MXene coating, endowing it with high Young's modulus (11.2 kPa) and compressibility (90%). Demonstrated with the finite element (FE) simulation, the elasticity and toughness of the GPM are originated from the hierarchical microstructure and the improved thickness of shells. Besides, the GPM exhibits a long-term durability of 5,000 compression cycles at 50% strain, a high sensitivity of 1.744 kPa−1, and a rapid response time of 4.5 ms. Further application examples reveal that the pressure sensor based on GPM has superior physiological signal capture capability, realizing fingertip pulse and temporal artery detections. Moreover, the GPM aerogel possesses high thermal insulation efficiency, resulting in good temperature adaptability in harsh environments from −196 °C to 300 °C. The sensitive pressure sensors should have the potential application in many fields of intelligent factories, complex militaries, and space discoveries.

A Secure Group Data Encryption Scheme in Intelligent Manufacturing Systems for IIoT

In recent years, there are many industries that imported intelligent systems to help them make the intelligent factory and get product record analysis to evaluate product rate. These intelligent systems could generate product records, store them to the on-line database, and provide product rate analysis from these records. Due to the rapid development of internet of things (IoT), the stockholder can construct its own smart factory with the smart intelligent system to develop its own industrial internet of things (IIoT) architecture. With the help of IIoT, the smart intelligence system can collect data information with IoT sensors embedded into each machine in the production line. However, there are some security issues arising between smart intelligent systems and IoT devices. In addition, the authors also discovered that there are fewer methodologies to talk about the data security during the machine transmitting its censored data to the other machines under the same network environment.

Evolution of Knowledge Structure in an Emerging Field Based on a Triple Helix Model: the Case of Smart Factory

As an important emerging field of science and technology, smart factories have received attention from academia, industry, and the government. Currently, although some scholars have examined the research trends in the field of smart factories, we have not found any research on the analysis of the knowledge output of relevant organizations among smart factories. An urgent problem is whether cooperation between organizations with different characteristics will affect the overall development of intelligent factories. This study aimed to perform a comprehensive analysis of the knowledge content and structure of a smart factory and the characteristics of its knowledge production structure. We also evaluated whether the triple helix structure was stable, and whether the research topics of different issues were similar. The triple helix model was used to study three aspects of the knowledge structure of a smart factory: university, government, and industry. Furthermore, the research contents of different organizations were analyzed in detail using network analysis. It was found that research funding at the national level leads to a knowledge spillover effect. After 2015, a triple helix knowledge structure was formed in the field of smart factories, which maintained a certain stability until 2020. The output of triple helix cooperation research has a significant impact. University research focuses more extensively and intensively on technology, government research on macro aspects, and Industry 5.0 has become a hotspot in industry research. The government needs to provide new platforms to integrate and promote the development of smart factories.

A Storage Optimization Model for Cloud Servers in Integrated Communication, Sensing, and Computation System

The massive amount of sensing and communication data that needs to be processed during the production process of complex heavy equipment generates heavy storage pressure on the cloud server-side, thus limiting the convergence of sensing, communication, and computing in intelligent factories. To solve the problem, based on machine learning techniques, a storage optimization model is proposed in this paper for reducing the storage pressure on the cloud server and enhancing the coupling between communication and sensing data. At first, based on the operation rules of the distributed file system on the cloud server, the proposed model screens and organizes the system logs. With the filtered logs, the model sets feature labels, constructs feature vectors, and builds sample sets. Then, based on the ID3 decision tree, a file elimination model is trained to analyze the files stored in the cloud server and predict their reusability. In practice, the proposed model is applied in the Hadoop Distributed File System and helps the system delete underutilized and low-value files and save storage space. Experiments show that the proposed model can effectively reduce the storage load on the cloud server and improve the integration efficiency of multisource heterogeneous data during complex heavy equipment production.