IET Collaborative Intelligent Manufacturing最新文献

This paper presents the development and validation of a digital twin system for a bakery production line, integrating real-time sensor data, physics-based process models and advanced predictive analytics with CNN + LSTM neural networks. The proposed architecture combines logistic growth, moisture evaporation and heat transfer equations with deep learning for accurate prediction and early detection of baking defects. Simulation and pilot implementation results demonstrate that the digital twin reproduces dough volume dynamics with an error below 3%, predicts humidity within ± 2% and stabilises oven temperature in a narrow range (± 1.2°C). The intelligent system enabled a 77% reduction in unplanned equipment downtime, decreased alarm events by over 60% and reduced the share of defective products from 8% to 2%. These outcomes highlight the practical impact and scalability of the hybrid digital twin framework for improving product quality, minimising losses and enhancing process reliability in food manufacturing.

This study addresses an integrated job shop scheduling problem with material handling robots (JSPMHR), aiming to minimise the maximum completion time (makespan). First, a mathematical model is developed to formulate the JSPMHR. Second, three meta-heuristics, genetic algorithm (GA), particle swarm optimisation (PSO) and artificial bee colony (ABC), are improved to solve the concerned problems. Based on the problem-specific features, seven local search strategies are designed to improve the convergence speed. Third, two reinforcement learning algorithms, that is, Q-learning and Sarsa, are employed to assist meta-heuristics in selecting the premium local search strategies during iterations. Finally, comprehensive experiments are conducted to evaluate the performance of the proposed algorithms by solving 82 benchmark instances. The proposed GA with Q-learning shows the strongest competitiveness among all compared algorithms solving JSPMHR.

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.

The historical development of mechatronics is discussed to identify the challenges in applying existing mechatronic principles to accelerate engineering innovations in the digital era. There are emerging needs in advancing the theory of mechatronics to (1) expand design principles to promote innovations and knowledge transfer, (2) integrate rapidly developed artificial intelligence (AI), human robots interactions (HRIs) and digital technologies to design and operate complex systems and (3) develop systematic approaches to enhance a system's scalability, adaptability and sustainability. Mechatronic systems are characterised in terms of their novelties and innovations, the importance of innovative thinking in mechatronic designs is thoroughly examined and the theory of inventive problem solving (TRIZ) is incorporated to facilitate mechatronic innovations. Five system engineering (SE) methods are introduced to stimulate innovations at different development phases of mechatronic systems. Finally, project-based mechatronic design (PBMD) is proposed as a methodological framework to integrate these design methods in promoting innovations in a mechatronic design.

In this paper, we address a stochastic variant of the well-known permutation flowshop scheduling problem, where the processing times of the jobs in the machines are assumed to be random variables. The objective considered is to minimise the expected makespan. This problem is substantially harder than its deterministic counterpart, as—except for a few special cases involving rather restrictive hypotheses—there is no closed formula to obtain the expected makespan of a given solution. Therefore, for most cases, it is necessary to estimate its expected value by sampling and averaging the results across a large number of replications. Furthermore, it has been shown that the number of replications required to obtain a reasonable estimate increases with the variability of the instance. In practice, this imposes extremely high computational costs for evaluating a solution, making it difficult to apply local search methods to instances of realistic size. Our proposal is to embed a machine learning technique (more specifically, gradient boosted trees or GBT) into a GRASP (greedy randomised adaptive search procedure) to compute a probabilistic threshold so it is possible to discard solutions with low probability of improving the actual best solution. The computational experience carried out shows that (1) the GBT is able to provide rather accurate estimates of the expected makespan even with a modest training effort and that its accuracy is not essentially influenced by the variability of the scenario and (2) the proposed procedure is able to produce the same quality of results as using the full sample of each solution, reducing the number of evaluated solutions by roughly 15%.

The fifth industrial revolution, known as Industry 5.0 (I5.0), has a vision for a new, resilient, socio-centred and competitive industry. The new approach provides a vision to enhance human-machine interaction (HMI) and assist operators efficiently. This study investigates the integration of human-centric principles within Industry 5.0, specifically in production engineering, with a central focus on the collaboration between humans and machines. Through an extensive literature review, the research identifies emerging trends and significant gaps in the current body of knowledge, especially regarding the development of intuitive and flexible interfaces, ethical frameworks in automated systems, and the management of cognitive load within manufacturing environments. The findings reveal considerable gaps in understanding the practical application of HMI across various industrial settings, emphasising the need for production technologies that enhance capabilities and advance a sustainable, ethical, and human-centred manufacturing landscape. This research contributes to the ongoing discourse on I5.0 by advocating for frameworks that prioritise human-centric values alongside technological innovation.

The cover image is based on the article A Novel DQN-Based Hybrid Algorithm for Integrated Scheduling and Machine Maintenance in Dynamic Flexible Job Shops by Wenchao Yi et al., https://doi.org/10.1049/cim2.70028.

This paper proposes a novel diagnostic approach for detecting inter-turn short-circuit faults in induction motors, combining the root-Prony method with fuzzy logic. Traditional techniques, such as the periodogram, have limitations in detecting low-magnitude harmonics and providing high-frequency resolution. To address these challenges, complex high-resolution methods, such as MUSIC and ESPRIT, have been developed. In this study, the root-Prony method is selected for its adaptability and low computational burden as it does not rely on space decomposition, making it faster than MUSIC. The proposed approach focuses on analysing the stator current signal within a specific frequency range near the fundamental rotor slot harmonics. By reducing the number of processed samples, computation time is further decreased. The integration of fuzzy logic enables intelligent decision-making regarding the condition of the stator circuit by considering harmonic magnitudes under different load torque values for accurate diagnosis. Experimental tests were conducted on an induction motor initially powered directly from an electrical network supplying symmetrical sinusoidal three-phase voltages. To demonstrate the robustness of the proposed method in noisy environments, additional tests were performed with the motor powered by a converter. In such scenarios, the conventional periodogram-based technique was unable to detect the desired harmonics due to the high harmonic content in the stator current signals. The test results confirm the superior effectiveness of the root-Prony method over the classical periodogram technique in estimating the frequencies and amplitudes of the targeted harmonics. The integration of the root-Prony method with fuzzy logic offers an advanced, efficient and reliable solution for fault diagnosis in induction motors.

This study aims to predict hard disk drives (HDDs) that pass initial testing but fail during reliability testing, using historical data from 8968 records with 218 features, such as head position and flying height of the read/write head. Since reliability testing is time-intensive, early failure prediction can significantly accelerate problem detection and resolution. The research focuses on detecting fly height modulation, a key symptom of HDD failure, and introduces an adaptive machine learning (ML) framework integrating AutoML for optimised model selection and hyperparameter tuning with MLOps for deployment, monitoring and continuous updates. Building on a previously proposed dual-stage classification framework that combines novelty detection and supervised learning, the proposed framework addresses the inefficiencies of manual hyperparameter tuning inherent in the earlier methods. The proposed framework achieves 92% accuracy in novelty detection and 100% in supervised learning, outperforming prior approaches. This integration of AutoML and MLOps offers a scalable, robust solution for early failure prediction, enabling real-time adaptability with minimal human intervention. Future work will focus on enhancing computational efficiency and responsiveness to data shifts and drifts, advancing data-driven decision-making in reliability testing.

Efficiency, security and transparency are critical for public health issues such as counterfeit vaccines, cold chain integrity and data authenticity. This study introduces a decentralised smart contract-driven blockchain framework that incorporates federated learning and edge computing to overcome the limitations of traditional centralised systems. Through decentralised decision-making and real-time data management, the proposed framework facilitates traceability, reduces vaccine wastage and ensures robust cold chain monitoring. The framework applied hybrid evaluation using Blockchain Impact Modelling (BIM) and Multi-Criteria Decision Analysis (MCDA) in order to show prominent improvement in scalability, security and operational efficiency. The experimental results confirm that the framework can detect temperature anomalies 76% faster, increase transaction throughput by 36.1% and improve breach detection to 99.7%. Smart contracts enable automated and accurate decision-making, whereas FL ensures privacy-preserving analytics among stakeholders. EC integration reduces latency and computational overhead, allowing real-time responses. This new architecture provides a strong foundation for vaccine supply chain management, addressing global challenges while adapting to different regulatory landscapes. This framework marks a new benchmark in securing healthcare logistics and extending its applicability to broader pharmaceutical supply chains.