Advancements in Industry 4.0 technologies, which combine big data analytics, robotics, and intelligent decision systems to enable new ways to increase automation in the industrial sector, have undergone significant transformations. In this research, a Hybrid Attention-Gated Recurrent Unit (At-GRU) model, combined with Sand Cat Optimization (SCO), is proposed to enhance fault identification and predictive maintenance capabilities. The model utilized multivariate sensor data from cyber-physical and IoT-enabled robotic platforms to learn operational patterns and predict failures with enhanced reliability. The At-GRU provides deeper temporal feature extraction, thereby improving classification performance. The robustness of the proposed model is validated through analysis of a benchmark dataset for industrial robots, and the results demonstrate that the proposed model exhibits impressive predictive capacity, surpassing other prediction methods and predictive maintenance approaches. Additionally, the performance evaluation indicates a lower computational cost due to the lightweight gating architecture of GRU, combined with attention. The robotic motion is further optimized by the SCO algorithm, which reduces energy usage, execution delay, and trajectory deviations while ensuring smooth operation. Overall, the proposed work offers an intelligent and scalable solution for next-generation industrial automation systems. Furthermore, the proposed model demonstrates the real-world applicability and significant benefits of incorporating hybrid artificial intelligence models into real-time robot control applications for smart manufacturing environments.

The flexible job shop scheduling problem with automated guided vehicles (FJSP–AGV) couples production and transport decisions, making scheduling and energy management computationally challenging. Conventional genetic algorithms apply a single decoder throughout the search and thus cannot adapt when instance characteristics or battery constraints change. We propose a portfolio–island decoding genetic algorithm (PID-NSGA-II) that shifts the focus from modifying evolutionary operators to learning which decoding strategies work best. Five heterogeneous decoders run in parallel on separate islands, and an upper-confidence-bound multi-armed bandit measures each island’s contribution to makespan improvement and adaptively reallocates population resources, automatically balancing exploration and exploitation. The framework is tested under two settings—pure makespan minimization and energy-aware scheduling with AGV battery considerations. Experiments on benchmark datasets show that PID-NSGA-II consistently improves solution quality and stability compared with single-decoder genetic algorithms, with greater gains when energy constraints are present. Adaptive learning of decoders delivers more robust scheduling decisions for complex FJSP-AGV environments and provides a scalable platform for smart manufacturing applications, achieving up to 25 % makespan reduction and substantial improvements in AGV battery levels across small, medium and large problem instances.

Abstract Modelling complex manufacturing processes presents significant challenges related to accuracy and explainability. Physics-based models, while interpretable and generalizable, often suffer from reduced accuracy due to simplifications and incomplete system understanding. On the other hand, purely data-driven models are typically more accurate but lack transparency, limiting their trust and adoption in critical manufacturing applications. Existing hybrid approaches attempt to address these issues but often retain black-box AI components that compromise interpretability. In this study, we propose a novel hybrid modelling framework that intrinsically integrates physics-based models with explainable AI, to correct for modelling inaccuracies. This approach offers both high accuracy and transparent, traceable decision-making. Its effectiveness is demonstrated through a case study predicting the real-time position of cutting tools from accelerometer signals during ultra-precision diamond turning.

Application of a forward design-based multi-attribute decision-making method in quality assessment in pharmaceutical tablet manufacturing.

AIGC-empowered smart manufacturing: Prospects and challenges

Human-Robot collaboration in industrialized construction manufacturing 5.0: A bibliometric mapping of smart production research

Artificial intelligence (AI) is increasingly adopted in manufacturing for tasks such as automated inspection, predictive maintenance, and condition monitoring. However, the opaque, black-box nature of many AI models remains a major barrier to industrial trust, acceptance, and regulatory compliance. This study investigates how explainable artificial intelligence (XAI) techniques can be used to systematically open and interpret the internal reasoning of AI systems commonly deployed in manufacturing, rather than to optimise or compare model performance. A unified explainability-centred framework is proposed and applied across three representative manufacturing use cases encompassing heterogeneous data modalities and learning paradigms: vision-based classification of casting defects, vision-based localisation of metal surface defects, and unsupervised acoustic anomaly detection for machine condition monitoring. Diverse models are intentionally employed as representative black-box decision-makers to evaluate whether XAI methods can provide consistent, physically meaningful explanations independent of model architecture, task formulation, or supervision strategy. A range of established XAI techniques, including Grad-CAM, Integrated Gradients, Saliency Maps, Occlusion Sensitivity, and SHAP, are applied to expose model attention, feature relevance, and decision drivers across visual and acoustic domains. The results demonstrate that XAI enables alignment between model behaviour and physically interpretable defect and fault mechanisms, supporting transparent, auditable, and human-interpretable decision-making. By positioning explainability as a core operational requirement rather than a post hoc visual aid, this work contributes a cross-modal framework for trustworthy AI in manufacturing, aligned with Industry 5.0 principles, human-in-the-loop oversight, and emerging expectations for transparent and accountable industrial AI systems.

Abstract The increasing adoption of digital technologies in operations has stimulated growing academic interest in their potential to enhance organizational sustainability. Nevertheless, prior research remains fragmented across technological, operational, and sustainability perspectives, limiting the development of an integrated understanding of how digitalization strategically enables sustainable operations. This study aims to consolidate this fragmented body of knowledge by developing an integrative conceptual framework for understanding digital technology adoption in operations as a driver of sustainability. Methodologically, the study applies a combined bibliometric and systematic content analysis of peer-reviewed articles retrieved from major academic databases. Bibliometric techniques are employed to examine publication trends, leading journals, influential countries, institutional affiliations, keyword co-occurrences, thematic clusters, and co-authorship networks. Content analysis is used to synthesize theoretical and empirical evidence within the identified clusters. The results indicate sharp growth in publications over the past decade, with dominant contributions from China, the USA, and European countries, and the emergence of strong collaborative networks among leading universities and research centers. Keyword co-occurrence and cluster analyses reveal four major research streams: Industry 4.0 and smart manufacturing, data-driven decision support, digital supply chain transparency, and enabling the circular economy. Content analysis further demonstrates that digital technologies, including artificial intelligence, big data analytics, blockchain, and the Internet of Things, contribute to environmental, social, and economic sustainability through efficiency gains, resource optimization, transparency improvements, the facilitation of circularity, and the development of operational resilience. Beyond mapping publication patterns, the study advances an integrative conceptual framework that positions digital technologies as strategic enablers of sustainability’s operationalization. The study concludes that the benefits of sustainability from digitalization are contingent upon the strategic integration of digital technologies within operational decision-making processes. Managerially, the findings guide firms in prioritizing digital investments to achieve sustainability objectives, while policy-wise, they inform the formulation of digitalization strategies for sustainable industrial development. Limitations arise from the reliance on secondary data and database coverage, suggesting that future research should empirically validate the proposed framework across industries and institutional contexts.

This research investigates the integration of Smart Production Systems (SPS) within the framework of Industry 5.0, emphasizing how such integration redefines operational efficiency and human–machine collaboration. The study aims to identify the contributions of smart technologies to productivity, sustainability, and human value in modern production systems. A Systematic Literature Review (SLR) was conducted following PRISMA guidelines, drawing from international databases including Elsevier, Springer, IEEE Xplore, Wiley, Taylor & Francis, ACM, and SAGE, as well as national sources. Publications from 2023–2025 were screened using keywords such as “Industry 5.0,” “Smart Production Systems,” “Human–Machine Collaboration,” and “Operational Efficiency.” Thematic analysis categorized findings into four dimensions: operational efficiency, human–machine collaboration, industrial sustainability, and socio-ethical aspects. Results indicate that SPS integration significantly enhances operational efficiency while fostering adaptive and creative collaboration between humans and machines. The combination of Artificial Intelligence (AI), Cyber-Physical Systems (CPS), and human creativity establishes a resilient, sustainable, and innovative production paradigm. Successful implementation of Industry 5.0 requires harmonizing technological advancement, human skills, and ethical principles. Practically, the study offers insights for industry stakeholders and policymakers in designing human-centered digital transformation strategies, strengthening supply chain resilience, workplace safety, and innovation. This research contributes conceptually by highlighting ethical and sustainable human–machine interactions in future production systems.

Design and Implementation of PLC-Based Automation for Smart Manufacturing Systems

Production management is dominant and essential in Taiwan because it confirms and certifies the effective use of timely delivery, resources, and high-quality output in an expert-driven and highly competitive manufacturing sector. With Taiwan being a world-famous player in industries, especially for precision machinery, semiconductors, and electronics. A modern difficulty in production management in Taiwan is the severe deficiency among significant manufacturing organizations, particularly in semiconductors, where over 30,000 posts remain empty, containing roles in maintenance, production, and quality control. For the valuation of the above problems, we consider the following production management systems in Taiwanese enterprises, such as the lean manufacturing system, smart production system, the automated inventory control system, the internet of things integrated manufacturing system, and the sustainable production management system. Therefore, we construct the procedure of the linguistic bipolar complex fuzzy soft multiattribute border approximation area comparison model and linguistic bipolar complex fuzzy soft multi-attribute decisionmaking model based on the proposed operators. For the assessment of the above problem, we resolve some numerical examples based on the above two models and also derive the activity of the comparative analysis between proposed and existing ranking values to enhance the efficiency and rationality of the derived models.

The external color of smoked sausages is a critical indicator of quality and uniformity in processing. Commercial colorimeters are unsuitable for high-throughput sorting due to the challenges posed by the sausage’s curved cylindrical surface and the need for an inline application. This study introduces a novel non-contact sensing module (LEDs at 45°, fiber optic collection at 0°) to acquire spectral data (400–700 nm) and derive CIE LAB. First, a handheld prototype validated the accuracy of the sensing module against a benchtop spectrophotometer. It successfully categorized five color grades (‘Over light’, ‘Light’, ‘Standard’, ‘Dark’, and ‘Over dark’) with a clear distribution on the a*-L* diagram. This established acceptable color boundary conditions (44.2 < L* ≤ 61.3, 14.1 < a* < 23.9). Second, three sensing modules were integrated around a conveyor belt at 120° intervals, forming the core of an automated inline sorting system. Blind field tests (n = 150) achieved high sorting accuracies of 95.3–97.3% with an efficient inspection time of less than 2 s per sausage. This work realizes the standardization, digitalization, and automation of food color inspection, demonstrating strong potential for smart manufacturing in the processed meat industry.

Continuous monitoring of the cutting tool insert’s condition is essential to enhance product quality and efficient machining process, by reducing the machine downtime. But the available tool condition monitoring approaches are often limited by coolant induced visibility loss in the cutting zone that reduces the feature reliability. This study proposes a transfer learning based deep learning method where the machining vibration signals are converted into visual representations and classified using ResNet 18, MobileNet V2, SqueezeNet, ShuffleNet, DenseNet 201, and EfficientNet B0 pretrained convolutional neural networks. This combination enables the model to learn deep wear profiles from vibration data without the manual feature extraction. Also, this method enhances signal strength, making it highly suitable for smart, scalable, and real world manufacturing environments. The effects of the proposed pretrained network hyperparameters, such as mini batch size, solver type, learning rate, and filter size, were studied and EfficientNet B0 was identified as the best performing network with a classification accuracy of 89.23% for tool condition monitoring tasks.

ABSTRACT Integrated Sensing and Communication (ISAC) systems harmoniously combine environmental detection and data transmission within a unified platform, enabling transformative applications in autonomous driving, smart manufacturing, and telemedicine. However, current implementations typically rely on mechanical beam steering and discrete bulky components to fulfill desired functions, limiting steering speed, reliability, and system scalability. Here, we demonstrate an ISAC system based on an integrated photonic‐assisted phased array for the first time. Operating as a transmitter, the array achieves rapid beam steering (23 ) over a wide range through on‐chip time‐delay manipulation. Moreover, the array directly converts optical signals into broadband (27 GHz) millimeter‐wave signals, eliminating the need for external radio‐frequency components. By employing an orthogonal frequency division multiplexing‐frequency modulated continuous‐wave (OFDM‐FMCW) hybrid signal as an integrated waveform, the proposed ISAC system supports the dual functions of sensing and communication without compromising performances in either domain, achieving efficient multiplexing of time and spectrum resources. As a proof‐of‐concept experiment, high‐speed communication (20 Gbps) and high‐precision (3 cm) ranging across different steering angles are successfully demonstrated within a shared spectrum and hardware platform. The proposed scheme shows significant advantages in steering speed, operating bandwidth, and reliability, delivering a potential solution for next‐generation ISAC networks.

The persistent challenge of air leakage in smart factories continues to impose significant costs and operational inefficiencies. Conventional solutions, such as infrared detectors, suffer from drawbacks, demanding additional manpower for detection and incurring monetary losses during equipment downtime. Addressing the urgent need for early air leakage detection in manufacturing plants amid the ongoing digital transformation, this paper introduces an end-to-end framework that jointly handles class imbalance and provides uncertainty-aware predictions. At its core, we propose a novel unsupervised-enhanced data sampling method (UEDSM) to preserve data structure while alleviating imbalance, integrated with a dropout-enabled neural network (ALDNet) that applies Monte Carlo Dropout for robust inference. The effectiveness of our method is validated through a comprehensive series of experiments, incorporating real-time physical monitoring of two air compressors within a manufacturing plant. Beyond minimizing resource wastage and human intervention, our solution achieves over 95% accuracy and an F1-score above 80%, enabling reliable leakage detection several minutes in advance. These results highlight the practical viability of our approach for deployment in edge environments, contributing to improved efficiency, reduced resource wastage, and enhanced resilience in smart manufacturing.

Biomechanical energy harvesting based on the triboelectric effect offers great potential in the fields of wearable electronics and smart textile manufacturing processes. While triboelectric nanogenerators (TENGs) show immense promise for self-powered wearable electronics, their transition from laboratory prototypes to commercially available products has been hindered by a lack of scalable, cost-effective manufacturing methods. Here, we address this critical industrial gap by introducing an entirely textile-based TENG production platform that leverages high-speed automated embroidery. This study presents a scalable industrial method that uses interlocking embroidery of silver-core PTFE and nylon sheaths to produce self-powered triboelectric sensors and harvesters. This approach overcomes significant challenges in sensor technology, such as high production costs and complex fabrication processes, by reducing manufacturing overhead and enabling flexible design capabilities. A 3.5 × 4.5 cm2 embroidered TENG achieves a high sensitivity of 6.3 V/kPa, a response time of ∼70 ms, stability over 30,000 cycles, abrasion resistance of up to 50,000 cycles, and machine washability of more than 50 washes. Moreover, it delivers an open-circuit voltage of ∼326 V, a short-circuit current of ∼12.8 μA, and a maximum peak power density of 652.36 mW/m2 at 33 MΩ, enough to illuminate 120 LEDs and charge capacitors to power low-power electronic devices. The self-powered sensors, integrated into the outer sole of a shoe and a doormat, demonstrate human motion sensing to analyze human body movement, pressure distribution, tapping activity, sports sensing, and gait metrics. These sensors transmitted the acquired bio signals wirelessly to an individual's smartphone, which, in turn, connected to a cloud IoT platform─highlighting their potential for mass production of next-generation smart textiles and wearable technology.

To address the high incidence of work-related musculoskeletal disorders (WMSDs) at manual edge-banding workstations in furniture factories, and in an effort to tackle the existing research challenges of poor cumulative risk quantification and inconsistent evaluations, this paper proposes a three-stage system for continuous, automated, non-invasive WMSD risk monitoring. First, MediaPipe 0.10.11 is used to extract 33 key joint coordinates, compute seven types of joint angles, and resolve missing joint data, ensuring biomechanical data integrity for subsequent analysis. Second, joint angles are converted into graded parameters via RULA, REBA, and OWAS criteria, enabling automatic calculation of posture risk scores and grades. Third, an Adaptive Pooling Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM) dual-branch hybrid model based on the Efficient Channel Attention (ECA) mechanism is built, which takes nine-dimensional features as the input to predict expert-rated fatigue states. For validation, 32 experienced female workers performed manual edge-banding tasks, with smartphones capturing videos of the eight work steps to ensure authentic and representative data. The results show the following findings: (1) system ratings strongly correlate with expert evaluations, verifying its validity for posture risk assessment; (2) the hybrid model successfully captures the complex mapping of expert-derived fatigue patterns, outperforming standalone CNN and LSTM models in fatigue prediction—by integrating CNN-based spatial feature extraction and LSTM-based temporal analysis—and accurately maps fatigue indexes while generating intervention recommendations. This study addresses the limitations of traditional manual evaluations (e.g., subjectivity, poor temporal resolution, and inability to capture cumulative risk), providing an engineered solution for WMSD prevention at these workstations and serving as a technical reference for occupational health management in labor-intensive industries.

Abstract Today, terms such as Sustainable Production, Industrial Cyber-Physical Systems, Cyber-Physical Production Systems (CPPS), Software-Defined Manufacturing, Smart Manufacturing, Industry 4.0, Industry 5.0, System of Systems, Internet of Things, Human-in-the-Loop, and Digital Twins are widely used. These concepts emphasize key characteristics of modern and future production systems, including heterogeneity, structural and behavioral complexity, intelligence, autonomy, reconfigurability, and human centrism. They also highlight the growing importance of reliable and up-to-date risk assessment, safety, and reliability measures, given the significant environmental, economic, and social demands. However, current industrial risk analysis methods lag behind the rising technical sophistication of such systems. It remains unclear whether existing methods can capture complex failure scenarios of dynamic, AI-driven systems with advanced software architectures.This paper discusses the main challenges facing safety engineers in industrial automation. We provide a classification and overview of available risk and reliability analysis methods and metrics, supported by a systematic review of 95 papers up to October 2025. The review addresses questions such as: which CPPS aspects must be considered, which methods are applicable, what are their advantages and limitations, and how can methods be combined? The findings reveal the need to extend classical approaches toward dynamic risk assessment, probabilistic model checking, AI-based techniques, digital twins, and intelligent fault injection. The study provides both a comprehensive overview of current risk and reliability assessment methods for CPPS and a roadmap for advancing their future development.

ABSTRACT The accelerating adoption of generative artificial intelligence (AI) is reshaping sustainable product design, yet current research remains fragmented across computational design, multi‐objective optimisation, and smart manufacturing. This systematic review addresses this fragmentation by analysing 59 peer‐reviewed studies (2010–2025) using PRISMA guidelines, advanced bibliometric mapping, and structural topic modelling to uncover how these domains converge to create superior sustainability outcomes. The study develops the Technology Convergence Framework, a unified theoretical model that integrates Advanced Computational Methods, Multi‐Objective Optimisation, and Smart Manufacturing into an interconnected system capable of delivering emergent performance improvements. Findings show that when these domains operate synergistically—supported by mechanisms such as infrastructural maturation, empirical validation feedback loops, and standardisation‐driven diffusion—manufacturers achieve 30%–65% gains in energy efficiency, waste reduction, and material optimisation, far exceeding improvements achieved through isolated technological efforts. The framework further incorporates human‐AI collaboration principles aligned with Industry 5.0, emphasising the critical role of human judgement, contextual reasoning, and ethical oversight in complementing AI‐driven decision systems. By bridging methodological, technological, and operational gaps, this review provides a holistic roadmap for transitioning from fragmented innovation to integrated sustainable product realisation, offering both scholars and industry leaders a coherent foundation for advancing next‐generation sustainable manufacturing ecosystems.

Technological empowerment and resilience play a crucial role in enhancing the capacity of organizations to maintain sustainability and respond effectively to crises and disruptions. This study aimed to examine the growth trend and thematic structure of international research in the field of smart technologies and sustainable production during the period from 2000 to 2024 through a systematic review of 290 scientific research articles indexed in global databases. The findings indicate that until 2018, the growth of publications was very slow, with an annual average of fewer than six articles. However, since 2019, a remarkable acceleration has been observed, with more than 65% of the articles (190 papers) published within the last five years. In terms of quality, the proportion of articles published in Q1 journals in recent years has increased to over 48% (compared to about 16% during the first decade of the study period). Moreover, the journals Journal of Cleaner Production, Sustainability, and International Journal of Production Research showed the highest frequency of publications, and the co-word network of keywords reflects a focus on areas such as sustainable development, Industry 4.0, the Internet of Things (IoT), and artificial intelligence (AI). Structurally, keyword clustering demonstrates the synergy between advanced smart technologies and sustainability objectives in industry. The statistical results of the meta-analysis showed that the Z-effect index was 1.82 (lower than the critical value of 2.69), indicating the stability of the findings. Furthermore, trend analysis reveals that the focus of studies has shifted from theoretical and feasibility issues toward the practical application of IoT and AI in industry and supply chains, with more than 35% of all articles dedicated to these topics. An examination of the cluster distribution of frequently used keywords in the fields of “smart manufacturing,” “smart technologies,” and “sustainable production” indicates that the largest share of articles in recent years falls under the clusters of “smart manufacturing and Industry 4.0” (12%) and “technological innovation and additive manufacturing” (12%). These findings highlight scientific maturity, an increase in international impact, and the growing attention of researchers to technological and resilient approaches in advancing sustainable production and development.