Manufacturing Vision Research Articles

A Universal Framework for Skill-Based Cyber-Physical Production Systems

In the vision of smart manufacturing and Industry 4.0, it is vital to automate production processes. There is a significant gap in current practices, where the derivation of production processes from product data still heavily relies on human expertise, leading to inefficiencies and a shortage of skilled labor. This paper proposes a universal framework for skill-based cyber–physical production systems (CPPS) that formalizes production knowledge into machine-processable formats. Key contributions include a novel conceptual model for skill-based production processes and an automated method to derive production plans from high-level CPPS skills for production planning and execution. This framework aims to enhance smart manufacturing by enabling more efficient, transparent, and automated production planning, thereby addressing the critical gap in current manufacturing practices. The framework’s benefits include making production processes explainable, optimizing multi-criteria systems, and eliminating human biases in process selection. A case study illustrates the framework’s application, demonstrating its current capabilities and potential for modern manufacturing.

A revolutionary approach to automate labelling of CAD files

Abstract This innovation represents a significant advancement in enhancing processing efficiency when dealing with engineering drawings. Traditionally, the review of customer drawings needs to be executed by experienced experts before manufacture. The identification and labelling of key features were reliant on human intervention for subsequent Design for Manufacturability (DFM) or G-code programming. However, the previous method not only consumed valuable time but also introduced the potential for human errors. In this context, we introduce a high-efficiency technology for recognizing engineering drawings, automatically labelling key features, and converting unstructured data into structured data. It is noteworthy that this technology distinguishes itself from traditional methods, such as R-CNN or similar neural network approaches and differs from pre-labeled engineering drawings reliant on experts manually inputting keywords for dimensions or layers. This work is a vital step towards the realization of intelligent manufacturing. For instance, our customers can now review engineering drawings online and receive immediate feedback. This online interaction is facilitated by processing structured data, a direct output of this presented technology, with other specific modules. Furthermore, the structured data seamlessly integrates into subsequent manufacturing procedures, including notifying the supply chain, generating 3D drawings, and CNC programming. In conclusion, this technology showcases its significance in advancing our vision of intelligent manufacturing at ZY Corp.

In-situ optical monitoring and analysis of weld pool based on machine vision for wire and arc additive manufacturing

In-situ optical monitoring and analysis of weld pool based on machine vision for wire and arc additive manufacturing

Non-destructive on-machine inspection of machining-induced deformed layers

Complete inspection of workpiece surface integrity invariably involves a form of destructive testing to enable the assessment of microstructural defects such as machining-induced white layers and near-surface plastic deformation. The incumbent offline and destructive microscopy inspection process is incompatible with both a digital and sustainable manufacturing vision of zero waste, as such, a non-destructive technique which utilises a novel X-ray diffraction surface integrity inspection method (XRD-SIIM) has been developed. This approach has been designed to complement traditional machinability-type assessments of tool life and machined surface topography, establishing a new process flow for validation. In this paper, for the first time, non-destructive on-machine validation of workpiece microstructural surface integrity is demonstrated, via a comparative investigation into the effect of insert grade, cutting speed and coolant delivery method on the depth of the imparted plastic deformation depth. It is shown that XRD-SIIM allows repeatable, non-destructive determination of deformed layers within a typical machining centre enclosure, with comparable findings to the incumbent cross-sectional microscopy approach. The generation of surface integrity digital fingerprints of a machining operation facilitates rapid comparison between testing variables, with a transition to an objective quantifiable assessment rather than one which open to subjectivity. In turn, XRD-SIIM expedites the development and benchmarking of new operations, tooling, materials, or coolant.

An LLM-based vision and language cobot navigation approach for Human-centric Smart Manufacturing

Industry 5.0 prioritizes Human-centric Smart Manufacturing (HSM), aiming to enhance human operators’ well-being and needs. This necessitates collaborative robots with advanced natural interaction capabilities and improved perception, cognition, and action intelligence. The Large Language Model (LLM) exhibits strong reasoning abilities and generalization capabilities, which can significantly advance the development of HSM once integrated into human–robot interaction and collaboration. Accordingly, this paper explores part of LLM’s ability in the context of smart manufacturing, focusing on addressing interruptions in the manufacturing process caused by repetitive tool fetching. To alleviate this issue, a vision and language cobot navigation approach is innovatively adopted in the manufacturing environment that could be further used to assist operators in retrieving tools. Specifically, a real Human–Robot Collaboration (HRC) manufacturing scene is first reconstructed and annotated using Three-Dimensional (3D) point cloud techniques. Then the LLM is utilized for the Automated Guided Vehicle (AGV) to comprehend natural language commands and generate Python code to triger navigation actions. Finally, the Pathfinder algorithm is applied for corresponding path planning. The framework is implemented in the Artificial Intelligence (AI) Habitat simulator, and the case studies demonstrate that the AGV can accurately comprehend complex language instructions, empowering human operators to complete manufacturing tasks efficiently.

A vision for sustainable additive manufacturing

A vision for sustainable additive manufacturing

Continuous high viscosity biphasic liquid separation

Enhancing the efficiency of separation in high-viscosity biphasic liquid purification processes can significantly reduce energy consumption and costs. This study has developed a co-axial biphasic liquid separator that can continuously and instantaneously separate two-phase mixtures with viscosities up to 2000 mPa∙s at room temperature, to meet real industrial needs. This co-axial biphasic liquid separator uses a special helix wire structure that separates the two-phase fluids based on the difference in surface tension, unaffected by the viscosity of the liquids. In experiments with PMMA-toluene-water mixtures, the flow rate can reach up to 4000 μl/min, with PMMA recovery rates consistently above 83.6 %, and as high as 91.5 %. In experiments with undiluted heavy oil–water, a purity of up to 87.58 % was achieved for heavy oil, and for diluted heavy oil, the purity even reached as high as 92.15 %. The separator demonstrated remarkable universality and durability, maintaining stable and high-purity biphasic separation across varying flow rates, through 15 cycles of repeated use, and during prolonged operations of up to one hour, without any system failures. The co-axial biphasic liquid separator developed in this study is currently the only few microseparator in continuous processes capable of separating liquids with viscosities over 2100 mPa∙s, and it could decrease the cost of relevant process development in industrial setups, realizing the vision of green manufacturing.

An advanced product inspection and sorting system using artificial intelligence

The manufacturing sector is experiencing a notable transformation due to the incorporation of Industry 4.0 and the emerging concepts of Industry 5.0. Artificial intelligence (AI) plays a significant role in driving this transformation, particularly in the domain of product inspection and sorting systems. Incorporating digital computer vision, high accuracy and resolution sensors, and bigdata-driven simulations into manufacturing processes, the vision of smart manufacturing becomes tangible. These technologies offer practical solutions for automating product inspection and sorting processes, providing non-destructive and cost-effective alternatives. This ongoing research aims to develop a real-time product inspection and sorting system utilising artificial intelligence, specifically focusing on convolutional neural networks (CNNs) and machine learning algorithms. The proposed approach adopts a dynamic methodology, leveraging the synergistic capabilities of CNNs and machine learning algorithms. To extract features from images, CNN is trained on datasets containing both none-defective and defective product samples. These features are then further refined and classified by machine learning algorithms. Through rigorous training on diverse datasets, the system developed a robust ability to distinguish between none-detective and defective products and achieving an accuracy close to 98.10%

Electrochemical Wastewater Refining: A Vision for Circular Chemical Manufacturing.

Wastewater is an underleveraged resource; it contains pollutants that can be transformed into valuable high-purity products. Innovations in chemistry and chemical engineering will play critical roles in valorizing wastewater to remediate environmental pollution, provide equitable access to chemical resources and services, and secure critical materials from diminishing feedstock availability. This perspective envisions electrochemical wastewater refining─the use of electrochemical processes to tune and recover specific products from wastewaters─as the necessary framework to accelerate wastewater-based electrochemistry to widespread practice. We define and prescribe a use-informed approach that simultaneously serves specific wastewater-pollutant-product triads and uncovers a mechanistic understanding generalizable to broad use cases. We use this approach to evaluate research needs in specific case studies of electrocatalysis, stoichiometric electrochemical conversions, and electrochemical separations. Finally, we provide rationale and guidance for intentionally expanding the electrochemical wastewater refining product portfolio. Wastewater refining will require a coordinated effort from multiple expertise areas to meet the urgent need of extracting maximal value from complex, variable, diverse, and abundant wastewater resources.

An accelerated strategy to characterize mechanical properties of polymer composites using the ensemble learning approach

Screening new materials and decoding their structure-property relationships is a time-consuming and costly task in the laboratory. This research proposes the use of machine learning (ML) algorithms with a new train-test splitting strategy to predict the complete stress-strain curves of polymer composites based on their compositional, environmental, and processing conditions using data extracted from an open-source database. Both the artificial neural network (ANN) and ensemble learning (EL) models demonstrated acceptable fit to the training datasets, yielding root mean squared errors (RMSEs) below 3.2 MPa. However, EL models demonstrated superior performance by reducing the RMSE of the testing data by an average of 61% for polyethylene terephthalate (PET) composites and 30% for polycarbonate (PC) composites, compared to the wide ANN models. Additionally, the developed EL models reduced the training time from a few hours (∼4 h on average) for wide ANNs to just a few seconds. The results of this study imply that relying solely on ANN models may not always provide the best solution for a given problem in materials science, and that EL models should be considered as a viable alternative, particularly in cases where data are limited. This study will pave the way for the automated design and characterization of advanced composites, while reducing the costly and laborious experiments in keeping with the vision of smart manufacturing and Industry 4.0.

An attention-based deep learning approach for inertial motion recognition and estimation in human-robot collaboration

In line with a human-centric smart manufacturing vision, human-robot collaboration is striving to combine robots’ high efficiency and quality with humans’ rapid adaptability and high flexibility. In particular, perception, recognition and estimation of human motion determine when and what robot to collaborate with humans. This work presents an attention-based deep learning approach for inertial motion recognition and estimation in order to infer when robotic assistance will be requested by the human and to allow the robot to perform partial human tasks. First, in the stage of motion perception and recognition, quaternion-based calibration and forward kinematic analysis methods enable the reconstruction of human motion based on data streaming from an inertial motion capture device. Then, in the stage of motion estimation, residual module and Bidirectional Long Short-Term Memory module are integrated with proposed attention mechanism for estimating arm motion trajectories further. Experimental results show the effectiveness of the proposed approach in achieving better recognition and estimation in comparison with traditional approaches and existing deep learning approaches. It is experimentally verified in a laboratory environment involving a collaborative robot employed in a small part assembly task.

The Effect of Thickness on Strength of Single Lap Orbital Riveted Aluminum/Composite Joints Used in Marine Environments

In an innovative vision of manufacturing, orbital riveting is a joining technique characterized by high efficiency, energy saving, low costs and low noise. It is a cold forming process where a tool rotates at a fixed angle (i.e., typically 3° to 6°) to create a sweeping line of pressure around a rivet. This movement progressively promotes, with each rotation, the collapse of the rivet shank down onto the upper substrate of a joint, permanently forming a rivet head. The aim of this research is to make and test multi-material joints between an aluminum AA5083 H111 sheet and a glass fiber reinforced laminate. Specifically, nine configurations of single lap joints were studied by investigating the effect of the thickness (i.e., 2.5, 3.0 and 4.0 mm for the aluminum and 2.5, 3.0 and 4.0 mm for the composite laminate) both on the mechanical characteristics and on the failure modes.

Cellular Cancer Immunotherapy Development and Manufacturing in the Clinic.

The transfusion of naturally derived or modified cellular therapies, referred to as adoptive cell therapy (ACT), has demonstrated clinical efficacy in the treatment of hematologic malignancies and metastatic melanoma. In addition, cellular vaccination, such as dendritic cell-based cancer vaccines, continues to be actively explored. The manufacturing of these therapies presents a considerable challenge to expanding the use of ACT as a viable treatment modality, particularly at academic production facilities. Furthermore, the expanding commercial interest in ACT presents new opportunities as well as strategic challenges for the future vision of cellular manufacturing in academic centers. Current trends in the production of ACT at tertiary care centers and prospects for improved manufacturing practices that will foster further clinical benefit are reviewed herein.

Digitization in bioprocessing: The role of soft sensors in monitoring and control of downstream processing for production of biotherapeutic products

Owing to the advancement in the technologies, the vision of smart manufacturing is not implausible. Development of sophisticated measuring tools, modelling approaches, improved computational efficiency, wireless technology and emphasis on automation and digitalization has promoted soft sensor implementation in bioprocessing. With multiple driving forces including decrease in operational cost, reliable operation, ease in calibration, increased long term stability etc., soft sensor has become a central process analytical tool for accessing and maintaining real time critical quality attributes. Fundamentally, soft sensor is an indirect measurement method combining process data and a mathematical model to estimate a target quantity. The model input is primarily composed of hardware signals from inline process sensors, analytical devices, actuators, and the instruments. This article offers an overview of soft sensing techniques in downstream bioprocessing for production of biotherapeutic products. Novel possibilities that its implementation presents towards real time measurement, monitoring, modelling and control (M3C) during commercial manufacturing have been highlighted along with the technology challenges and potential solutions from a data scientist perspective.

Secure Blockchain Middleware for Decentralized IIoT towards Industry 5.0: A Review of Architecture, Enablers, Challenges, and Directions

Resilient manufacturing is a vision in the Industry 5.0 blueprint for satisfying sustainable development goals under pandemics or the rising individualized product needs. A resilient manufacturing strategy based on the Industrial Internet of Things (IIoT) networks plays an essential role in facilitating production and supply chain recovery. IIoT contains confidential data and private information, and many security issues arise through vulnerabilities in the infrastructure. The traditional centralized IIoT framework is not only of high cost for system configuration but also vulnerable to cyber-attacks and single-point failure, which is not suitable for achieving the resilient manufacturing vision in Industry 5.0. Recently, researchers are seeking a secure solution of middleware based on blockchain technology integration for decentralized IIoT, which can effectively protect the consistency, integrity, and availability of IIoT data by utilizing the auditing and tamper-proof features of the blockchain. This paper presented a review of secure blockchain middleware for decentralized IIoT towards Industry 5.0. Firstly, the security issues of conventional IIoT solutions and the advantages of blockchain middleware are analyzed. Secondly, an architecture of secure blockchain middleware for decentralized IIoT is proposed. Finally, enabling technologies, challenges, and future directions are reviewed. The innovation of this paper is to study and discuss the distributed blockchain middleware, investigating its ability to eliminate the risk of a single point of failure via a distributed feature in the context of resilient manufacturing in Industry 5.0 and to solve the security issues from traditional centralized IIoT. Also, the four-layer architecture of blockchain middleware presented based on the IIoT application framework is a novel aspect of this review. It is expected that the paper lays a solid foundation for making IIoT blockchain middleware a new venue for Industry 5.0 research.

Research on Industrial Machinery Design and Automation Technology under the Vision of Intelligent Manufacturing

Research on Industrial Machinery Design and Automation Technology under the Vision of Intelligent Manufacturing

First principles of smart manufacturing

AbstractThe democratizing of innovation and technology happens when tools and expertise are made affordable and accessible at scale to most manufacturers. To realize the democratization of smart manufacturing innovation, it is not only necessary to democratize the technology, but also necessary to provide wide access to the knowledge required to implement the strategies and leverage the solutions and insights in a more digitally enabled manufacturing ecosystem. We are at a point in the journey where it is time to converge on a concrete set of guiding principles and a framework that organizes expectations and requirements for the implementation of smart manufacturing. This paper summarizes the practices that define smart manufacturing as developed by the industry leaders, early adopters, and expert practitioners working in the ecosystem at CESMII, the smart manufacturing institute. In smart manufacturing, organizations, people, and technology work in synergy via processes and technology‐based solutions that follow these seven First Principles: Flat & real‐time, open & interoperable, proactive & semi‐autonomous, sustainable & energy efficient, secure, scalable, orchestrated & resilient. All the design principles must be considered to fully realize the vision for smart manufacturing. This paper will further explain how the seven principles work collectively to achieve new levels of connectivity, intelligence, and automation in the manufacturing ecosystem.

Prediction of Failure Categories in Plastic Extrusion Process with Deep Learning

Today’s manufacturing vision necessitates extracting insights from the data collected in real-time from manufacturing processes. Predicting failures with the predictive analysis of the collected process data and preventing these failures by taking necessary actions before they occur is a key factor in ensuring quality at the desired level, increasing productivity, and reducing costs in production systems. In the literature on predictive analysis of process data, machine learning and deep learning methods have attracted considerable attention, especially in recent years. This study has addressed a multi-class failure classification problem in the plastic extrusion process with a real case study. Classification models have been developed based on Long Short-term Memory (LSTM) as a deep learning method and Multilayer Perceptron (MLP) and Logistic Regression (LR) as machine learning methods to predict the failure categories. In the case study, real data taken from the extrusion process of one of the leading insulation companies operated in Izmir has been used. The final dataset includes actual measurements of seven parameters related to temperature and pressure and failure categories as the target variable. Three failure categories have been identified to define Category 0 (No failure), Category 1 (Filter change), and Category 2 (Feeding failures) states, and coded as 0,1 and 2 in the models, respectively. LSTM, MLP, and LR’s performance to predict the failure categories have been evaluated and compared based on accuracy, precision, recall, and F1 Score measures. LSTM is the highest performing among the three methods, with 100% prediction accuracy for each failure category. On the other hand, LR and MLP have achieved considerable and close results except for Category 1.

Applied computer vision for composite material manufacturing by optimizing the impregnation velocity: An experimental approach

The application of the cutting edge industrial solutions in composite manufacturing are optimizing the processes, quality control and resources usage. Computer vision is currently used in many quality control stages, although its potential advantages have not been applied in the process control. In this paper, computer vision is used to control the impregnation velocity in VARI (Vacuum Assisted Resin Infusion) process. As it is known, there is a relationship between the impregnation velocity and the final mechanical properties in LCM (Liquid Composite Manufacturing) processes. Constant and optimum flow front velocity mean optimum mechanical properties, although the nature of the process makes it difficult to keep these conditions. Then, a methodology has been proposed to identify and use the optimum velocity during the manufacturing process. Firstly, the flow front recognition algorithm was calibrated to be used in different reinforcement and fluid systems. Then, tensile and impact specimens have been manufactured and tested at different controlled and uncontrolled velocities. As a result, the tensile modulus has been increased up to 12.6%, as the tensile strength has increased up to 8.7%. Similarly, the maximum reaction force during the impact test has been increased up to 6.5%, as the damaged area has been reduced by 8.8%. For stitched laminates, force results increase up to 3.2%, as the damaged area has been reduced up to 31% when the optimum velocity is used. The experimental results have demonstrated the advantage of using mechanisms to control the impregnation process to achieve improved mechanical properties of composite materials.

«Faux Leather»: What Are We Talking About? A Sustainability Project

The tanning sector is a strong point of the so-called ‘Made in Italy’, but for a long time the leather supply chain has been considered responsible for pollution. This project, developed as practice during the training course on Agenda 2030 organized by AIB Emilia-Romagna, has the long-term objective of highlighting the stereotypes related to the vision of leather manufacturing, and of illustrating the possibility of a sustainable tanning process. The project includes some exhibitions (both on-site and virtual) focused on three aspects of the tanning activity: the scientific books on vegetal tanning process, the leather bindings and parchment in antiquarian books as materials of animal origin, and the plants traditionally used in leather tanning. A thematic web portal brings together the various exhibitions. The common goal is to widespread some adequate knowledge on the 2030 Agenda, in order to make people aware of the importance of the Sustainable Development Goals.