This study proposes an automated framework for online cutting tool wear classification in CNC turning using low-cost optical equipment and Convolutional Neural Networks (CNNs). Longitudinal turning experiments were performed on CK45 medium carbon steel using a HAAS TL1 lathe under dry machining conditions. Tool wear evolution was monitored via a lathe-mounted digital microscope, with images classified into three distinct stages: Low (Vb<160 μm), Medium (160≤Vb≤200 μm), and Critical (Vb>200 μm). A shallow CNN architecture, consisting of three convolutional blocks and a Softmax output layer, was developed to balance model complexity with computational efficiency for potential edge deployment. To enhance robustness against positional changes, data augmentation techniques including random translations and rotations were applied. The results demonstrate good performance, with the model achieving 94.7% accuracy and a weighted F1-score of 95.4% on the testing subset. While the model showed exceptional performance in identifying Low and Medium wear, data scarcity in the Critical wear class remained a limiting factor for recall. Overall, the study confirms that shallow CNNs can accurately capture spatial hierarchies for image-based wear assessment.

High-level Computer-Aided Process Planning (CAPP) generates manufacturing process plans from part specifications. It suffers from limited dataset availability in industry, reducing model generalization. We propose a semi-supervised learning approach to improve transformer-based CAPP transformer models without manual labeling. An oracle, trained on available transformer behaviour data, filters correct predictions from unseen parts, which are then used for one-shot retraining. Experiments on small-scale datasets with simulated ground truth across the full data distribution show consistent accuracy gains over baselines, demonstrating the method's effectiveness in data-scarce manufacturing environments.

This study applies Large Language Models (LLMs) to high-level Computer-Aided Process Planning (CAPP) in a distributed manufacturing context. It aims to generate alternative, feasible process chains for production of a wide range of parts. Parts are encoded in a custom encoding scheme supporting diverse part overall shapes, geometrical features within them, and corresponding manufacturing processes. The CAPP problem is formulated as a sequence prediction task, where a GPT-2-based LLM generates process chains autoregressively. To train and test the LLM a synthetic dataset of 7,840 unique parts and their alternative process chains was generated using expert-driven rule-based logic. The LLM is trained from scratch using a tokenization scheme treating part features and processes uniformly as discrete tokens, special tokens being employed to control sequence flow. Performance evaluation was performed for systematically reducing the size of the dataset. Finally, even with as little as 5% of the training data, the LLM achieves over 99% accuracy at the process chain-level. The extremely few spotted errors mainly involve minor secondary process mispredictions without critical failures. For comparison, a Recurrent Neural Network (RNN) was also trained with the same dataset. Since manufacturing data stemming from experts and not from sensors is notoriously difficult to collect, training a machine learning model with a dataset that is as small as possible is of utmost importance. In this light, the LLM proved superior to RNN, in fact emphatically so, the more the training dataset was limited.

This article addresses the challenge of human activity recognition (HAR) in industrial environments, focusing on the effectiveness of various neural network architectures. In particular, simpler Feedforward Neural Networks (FNN) are focused on with an aim to optimize computational performance without compromising accuracy. Three FNN configurations-FNN1, FNN2, and FNN3-were evaluated alongside the Convolutional Neural Network (CNN 1D) model for comparative analysis. The results indicate that the FNN achieved accuracy rates ranging from 94.28 to 99.19%, while the CNN 1D exhibited an accuracy of 98.12%. Despite the CNN 1D's efficiency for real-time applications, the FNN's fast training times and high accuracy make them particularly valuable in resource-constrained environments such as mobile devices. The findings suggest that while more complex models such Long Short-Term Memory (LSTM)-Auto-Encoder configurations, that have been tried by the same research group before, may offer better adaptability, simpler architectures can provide effective results in HAR tasks. Notably, these simpler models can be adopted in cascading systems operating online, serving as detectors of known activities for real-time monitoring and classification.

As the name of this Special Issue indicates, its aim was to highlight advanced applications of Artificial Intelligence (AI) in metallic materials processing, containing a final total of 12 published papers [...]

Laser Powder Bed Fusion (LPBF) has emerged as a leading additive manufacturing (AM) process for producing complex metal components. Despite its advantages, the inherent LPBF process complexity leads to challenges in achieving consistent quality and repeatability. To address these concerns, recent research efforts have focused on sensor fusion techniques for process monitoring, and on developing more elaborate control strategies. Sensor fusion combines information from multiple in situ sensors to provide more comprehensive insights into process characteristics such as melt pool behavior, spatter formation, and layer integrity. By leveraging multimodal data sources, sensor fusion enhances the detection and diagnosis of process anomalies in real-time. Closed-loop control systems may utilize this fused information to adjust key process parameters–such as laser power, focal depth, and scanning speed–to mitigate defect formation during the build process. This review focuses on the current state-of-the-art in sensor fusion monitoring and control strategies for LPBF. In terms of sensor fusion, recent advances extend beyond CNN-based approaches to include graph-based, attention, and transformer architectures. Among these, feature-level integration has shown the best balance between accuracy and computational cost. However, the limited volume of available experimental data, class-imbalance issues and lack of standardization still hinder further progress. In terms of control, a trend away from purely physics-based towards Machine Learning (ML)-assisted and hybrid strategies can be observed. These strategies show promise for more adaptive and effective quality enhancement. The biggest challenge is the broader validation on more complex part geometries and under realistic conditions using commercial LPBF systems.

Flexible manufacturing cells provide significant advantages in low-volume mass-customization production but also induce added complexity and technical challenges in terms of integration, control, and extensibility. The variety of closed-source industrial protocols, the heterogeneous equipment, and the product’s manufacturing specifications are main points of consideration in the development of such a system. This study aims to describe the approach, from concept to implementation, for the development of the controller for a flexible manufacturing cell consisting of heterogeneous equipment in terms of functions and communication interfaces. Emphasis is put on the considerations and challenges for effective integration, extensibility, and interoperability. Scheduling and monitoring performed by the developed controller are demonstrated for a manufacturing cell producing microfluidic devices (bioMEMS) that consists of six workstations and a robot-based handling system. Communication between the system controller and the workstations was based on open-source technologies instead of proprietary software and protocols, to support interoperability and, to a considerable extent, code reusability.

This study addresses the need for advanced machine learning-based process monitoring in smart manufacturing. A methodology is developed for near-real-time part quality prediction based on process-related data obtained from a CNC turning center. Instead of the manual feature extraction methods typically employed in signal processing, a novel one-dimensional convolutional architecture allows the trained model to autonomously extract pertinent features directly from the raw signals. Several signal channels are utilized, including vibrations, motor speeds, and motor torques. Three quality indicators—average roughness, peak-to-valley roughness, and diameter deviation—are monitored using a single model, resulting in a compact and efficient classifier. Training data are obtained via a small number of experiments designed to induce variability in the quality metrics by varying feed, cutting speed, and depth of cut. A sliding window technique augments the dataset and allows the model to seamlessly operate over the entire process. This is further facilitated by the model’s ability to distinguish between cutting and non-cutting phases. The base model is evaluated via k-fold cross validation and achieves average F1 scores above 0.97 for all outputs. Consistent performance is exhibited by additional instances trained under various combinations of design parameters, validating the robustness of the proposed methodology.

This paper presents the integration of a Natural Language Processing (NLP) approach with an industrial robot to enable voice control for online programming. This approach initially involves real-time conversion of voice commands into text, utilizing the Microsoft Speech Software Development Kit (SDK). The generated text is processed to extract numerical data corresponding to desired positions and orientations. Subsequently, a machine learning binary classification model is trained to distinguish between robot related and unrelated commands. The recognized commands undergo further keyword-based classification based on a preselected vocabulary of supported robot operations, so that finally the inverse kinematics problem can be solved and the relevant robot program can be generated. A custom graphical user interface (GUI) facilitates user interaction. Two modes are supported, the first for issuing multiple individual commands and the second for programming typical pick-and-place industrial tasks. The obtained results using a 6-DOF Staubli RX90L robot verify the approach's feasibility.

Industry 4.0 approaches in manufacturing rely on extensive data to create digital twins. Modern CNC machine tools can provide such data via built-in systems. However, legacy CNC machines, still prevalent in SMEs, lack these features. This paper focuses on enhancing legacy CNC machines towards process monitoring and integration into a cloud manufacturing ecosystem. The proposed methodology, demonstrated on a specific turning center, begins with instrumenting the machine and using a DAQ device to capture and synchronize analog signals e.g. for motor speed, torque and vibration. Next, a software application developed following a Service Oriented Architecture, operates on a single producer service to accumulate sensor data, group it into buffers, and distribute it in near-real time to other software components. These include a process monitoring service using a CNN to predict quality parameters, a bridge service to transmit streaming data via Server Sent Events and a web-based visualization interface.

Remanufacturing has emerged as a pivotal practice in the pursuit of sustainable manufacturing but it is facing challenges, including decision complexity, ergonomic concerns, and assembly and disassembly plan efficiency. This paper introduces a framework for advancing remanufacturing processes emphasizing Industry 4.0 (I4.0) concepts with special focus on Virtual Reality (VR) as exemplified by three applications. The first one leverages an armband motion sensor for intuitive programming of a robot, demonstrating potential enhancements in programming robotic tasks involved. The second one introduces VR for automated ergonomic assessment in remanufacturing tasks involving the human, using standard protocols. The third one explores VR-based evaluation and guidance of collaborative assembly plans by human and robot.

In the framework of the H2SusBuild project, a prototype hybrid energy system combining Renewable Energy Sources (RES) and hydrogen (H2), as energy storage material and as a green fuel, has been designed, installed and is currently in operation in Lavrion, Greece. The system comprises thin-film PV panels, wind turbines, a water electrolysis unit, a H2 compressor, a micro-CHP unit (PEM fuel cell) and a H2 burner. The aim of this system is to cover the energy needs of a ≈500 m2 office building and to render it a self-sustained, zero CO2 emission installation. The entire site is operated through an advanced Energy Management and Control System (EMCS), responsible for monitoring magnitudes of interest (for example power, energy, stored H2 volume etc.) and managing the synergistic operation of all components based on a custom developed control algorithm. This paper’s aim is to present the results of system operation and evaluate the performance of key components based on data collected during system operation under real conditions. The results show good efficiencies for all major system components and that the implemented control algorithm achieves the zero CO2 emissions goal.

Background:Despite recent guidelines appropriate lipid-lowering treatment (LLT) remains suboptimal in everyday clinical practice. Aims:We aimed to describe clinical practice of use of LLT for at least high CV risk populations in a Hellenic real-world setting and assess how this relates to the European Society of Cardiology treatment guidelines. Methods:We analyzed data from a retrospective cohort study of the National Registry of patients with dyslipidemia between 1/7/2017 and 30/6/2019 who were at least of high CV risk and filled a dual or triple lipid-lowering treatment (dLLT, tLLT) prescription. The primary outcomes of interest of this analysis were to report on the patterns of LLT use in at least high CV risk patients. Results:A total of 994,255 (45.4% of Greeks on LLT) were of at least high CV risk and 120,490 (5.5%) were on dLLT or tLLT. The percentage of patients with reported statin intolerance ranged from 2 to 10%. While persistence was reported to be satisfactory (>85% for both dLLT or tLLT), adherence was low (ranging between 14 and 34% for dLLT). In 6-month intervals, the percentage of patients achieving a low-density lipoprotein cholesterol (LDL-C) target below 100 md/dL ranged from 20% to 23% for dLLT and 34%-37% for tLLT. Conclusions:The prevalence of at least high CV risk patients among patients receiving LLT in Greece is substantial. Despite the high persistence and probably due to the low adherence to treatment, LDL-C remains above targets in more than two thirds of patients.

The quality control of discretely manufactured parts typically involves defect recognition activities, which are time-consuming, repetitive tasks that must be performed by highly trained and/or experienced personnel. However, in the context of the fourth industrial revolution, the pertinent goal is to automate such procedures in order to improve their accuracy and consistency, while at the same time enabling their application in near real-time. In this light, the present paper examines the applicability of popular deep neural network types, which are widely employed for object detection tasks, in recognizing surface defects of parts that are produced through a die-casting process. The data used to train the networks belong to two different datasets consisting of images that contain various types of surface defects and for two different types of parts. The first dataset is freely available and concerns pump impellers, while the second dataset has been created during the present study and concerns an automotive part. For the first dataset, Faster R-CNN and YOLOv5 detection networks were employed yielding satisfactory detection of the various surface defects, with mean average precision (mAP) equal to 0.77 and 0.65, respectively. Subsequently, using transfer learning, two additional detection networks of the same type were trained for application on the second dataset, which included considerably fewer images, achieving sufficient detection capabilities. Specifically, Faster R-CNN achieved mAP equal to 0.70, outperforming the corresponding mAP of YOLOv5 that equalled 0.60. At the same time, experiments were carried out on four different computational resources so as to investigate their performance in terms of inference times and consumed power and draw conclusions regarding the feasibility of making predictions in real time. The results show that total inference time varied from 0.82 to 6.61 s per image, depending on the computational resource used, indicating that this methodology can be integrated in a real-life industrial manufacturing system.

The importance of the economy being up to date with the latest developments, such as Industry 4.0, is more evident than ever before. Successful implementation of Industry 4.0 principles requires close cooperation of industry and state authorities with universities. A paradigm of such cooperation is described in this paper stemming from university partners with partly overlapping and partly complementary areas of expertise in manufacturing. Specific areas that are targeted include Additive Manufacturing, cloud computing and control, Virtual Reality, Digital Twins, and Artificial Intelligence. The manufacturing system domains that are served pertaining to process planning and optimization, process and system monitoring, and innovative / precision manufacturing. The described collaborative research and training framework involves a combination of pertinent targeted individual exploratory innovation projects as well as a synthetic multifaceted common research project. Based on these, the research and innovation project knowledge will be transferred to the industry by building a Cluster of Excellence, i.e., a network consisting of academic and industrial stakeholders.

In the last decades there has been an ever-increasing trend of accelerated scientific research in the biotechnology sector and the COVID-19 pandemic has further emphasized the need for high-throughput yet flexible processes. Even though automation has gained significant popularity to deal with this problem, the focus has been placed at laboratory level and for specific type of processes only. The aim of these automation applications is to offset the increasing costs of clinical trials, automate tedious laboratory work, run experiments in parallel and make scientific testing efficient and programmable. This paper investigates the application of robotic systems at a larger scale that will allow the automated execution of even custom defined biotechnology experiments so that these can be offered as a service to any interested stakeholder. A digital model of the proposed robotic biotechnology workcell is developed and the necessary methodology is investigated to showcase the feasibility of this approach. The goal is to use the digital model for evaluating different design considerations, such as equipment layout and robot types, as well as for planning and simulating the robot operation under various offline programming strategies. To achieve this goal, the Damped Least Squares Method is used for inverse kinematics control and a robotic arm trajectory planner is developed parametrically using characteristic intermediate points so as to create a motion plan that can be used for any robotic manipulator and any experiment within a family of facility setups. Two experiments are simulated using a SCARA robot and an articulated arm robot, each with an alternative cell layout to show the flexibility and robustness of the proposed approach. The obtained results show that the trajectory planner can consistently generate appropriate motion plans.

This paper aims at identifying the main challenges associated with the adoption of blockchain and smart contract technologies as potential solutions for cybersecurity in an Industrial Internet of Things environment. Initially, an introduction to IIoT, blockchain and smart contracts is made, highlighting their core concepts as well as explaining their characteristics and properties. Subsequently, based on a literature review, the various benefits and challenges for their implementation are identified. The latter are presented through a hierarchical approach that first categorizes them in architecture and privacy related challenges, with each category being further divided into more specific challenges to emphasize the complex nature of the problem. Some typical use cases are also mentioned to showcase the feasibility of these technologies. Finally, a conceptual framework for IIoT cybersecurity is proposed in terms of future work directions.

One of the aims of the 4th industrial revolution is to seamlessly connect equipment and personnel to enable a greater level of collaboration, which in turn will result in higher operational efficiency and improved decision making strategies. In this context, this work presents a deep learning based approach for automated activity recognition of two human workers who are carrying out typical manual assembly and part handling tasks in a production floor environment. A typical smartphone is employed as a wearable sensor recording acceleration in 3 directions, total acceleration, brightness, geomagnetic field strength, orientation in space and sound wave intensity and is attached to the wrist of the two workers, who subsequently perform a large number of experiments examining 5 activities of interest. For the classification of these activities, two types of neural networks are employed. Initially, the experimental signals are fed to the variational autoencoder neural network so as to extract the appropriate features and subsequently these features are added to the last layer of a long short-term memory (LSTM) time series network, which is trained to classify the 5 activities of interest. An additional goal of the present paper is the study of long-term signals (LTS), that is time series of much longer duration than those of the 5 aforementioned activities, including enough noise due to the existence of time periods corresponding to unrelated activities. Ultimately, a new methodology is proposed combining both trained neural networks mentioned above, in order to identify the desired activities within the LTS in a real time application.