The manufacturing industry is undergoing a profound transformation driven by rapid technological advances and the growing integration of intelligence into industrial systems. As manufacturing continues to evolve beyond traditional automation, unprecedented opportunities are emerging to enhance autonomy, adaptability, resilience, and sustainability throughout the product lifecycle. Meanwhile, these developments bring new scientific and engineering challenges associated with the design, deployment, and management of increasingly intelligent manufacturing systems. Against this backdrop, this Special Issue brings together fourteen contributions that capture recent progress and emerging perspectives across a broad spectrum of intelligent manufacturing research.

A defining characteristic of next-generation manufacturing is the emergence of autonomous intelligence, in which manufacturing entities are expected not merely to execute predefined tasks but also to perceive, reason, learn, and collaborate within dynamic environments. Zhang et al. presented a comprehensive overview of embodied intelligence technologies, including data-driven situational awareness, learning-based control, and generative actuation design, while discussing the implications for digital twins, foundation models, multimodal intelligence, and human-centered manufacturing environments. Complementing this perspective, Shen and He reviewed the historical evolution of manufacturing agents, tracing their development from rule-based distributed agents to reinforcement learning-based agents and, more recently, cognitive autonomous agents, which further identifies major challenges associated with interoperability, scalability, resilience, and human-AI collaboration in future manufacturing systems. Extending the discussion from individual intelligent entities to system-level interactions, Camarinha-Matos examined the evolution of collaborative manufacturing networks and their significance within both technology-oriented Industry 4.0 paradigm and human-centric vision of Industry 5.0. Particular attention is devoted to value co-creation, human-in-the-loop systems, and collaborative intelligence. Collectively, these studies portray a future manufacturing landscape characterized by increasing autonomy, interconnectedness, and closer integration between humans and intelligent systems.

Bridging the physical and digital worlds has become a central topic in intelligent manufacturing, driving the development of increasingly sophisticated cyber representations of manufacturing assets, processes, and systems. Addressing this challenge, Kusiak systematically presented a fifteen-step framework that spans modeling, testing, validation, deployment, maintenance, and continuous evolution, offering practical guidance for the implementation of digital twins in manufacturing environments. Furthermore, Chen et al. introduced a novel hypergraph-based neural modeling framework capable of capturing both macro-level system behaviors and micro-level dynamic interactions. The proposed approach achieved enhanced modeling fidelity and operational effectiveness, illustrating the growing convergence of machine learning and digital twin technologies. Taken together, these contributions strengthen both the theoretical foundations and practical realization of digital twins, reinforcing their position as a key enabler of data-driven manufacturing.

Beyond system representation, intelligent manufacturing ultimately requires effective decision-making mechanisms to adapt to uncertainty, complexity, and continuously evolving production requirements. Lennartson investigated the relative strengths of model-based and model-free reinforcement learning under dynamic disturbances and reveals the superior data efficiency and robustness offered by model-based approaches. Focusing on engineering design optimization, Yuk and Park employed reinforcement learning techniques to optimize lattice unit-cell geometries, enabling simultaneous improvements in structural stiffness and lightweight performance. Meanwhile, Yan et al. surveyed recent advances in machine learning, deep learning, Internet of Things (IoT)-enabled services, and generative design technologies for cyber manufacturing applications, providing a comprehensive perspective on future opportunities in intelligent design, process planning, and manufacturing services. Together, these studies illustrate the expanding capability of learning-based optimization approaches to support adaptive and intelligent decision-making across multiple stages of the manufacturing lifecycle.

The emergence of foundation models is reshaping the landscape of industrial artificial intelligence, enabling more generalizable, context-aware, and knowledge-intensive solutions for manufacturing applications. At the same time, the adoption of such technologies raises important questions regarding trustworthiness, explainability, and domain-specific adaptation. Zhao et al. proposed an industrial large model framework for the defect detection that emphasizes interpretability and reliability while demonstrating strong performance across multiple benchmark datasets. Ren et al. developed a domain-adapted large language model tailored for prognostics and health management tasks, including anomaly detection, fault diagnosis, maintenance decision-making, and knowledge mining. In parallel, Wang et al. integrated causal reasoning and equipment structural knowledge through knowledge graphs to improve maintenance recommendation quality, illustrating the growing importance of knowledge-enhanced AI in industrial settings. Collectively, these contributions signal a transition from narrowly specialized intelligent tools toward more versatile, explainable, and knowledge-aware industrial AI systems.

The ultimate value of manufacturing intelligence lies in its ability to improve operational outcomes, including product quality, process reliability, and resource efficiency. For this purpose, Zhao et al. developed a generalized neural operator framework capable of predicting machining deformation fields across varying geometries, addressing long-standing challenges related to model transferability and scalability. Feng et al. proposed an innovative surface defect detection approach that combines super-resolution reconstruction and deep neural networks while incorporating expert knowledge into the learning process. In addition, Wu et al. demonstrated how IoT infrastructures, together with temporal convolutional and graph attention networks, can support real-time energy anomaly detection and diagnosis, thereby enhancing both energy efficiency and operational safety. These contributions exemplify the growing impact of advanced data analytics, intelligent sensing, and predictive technologies in enabling more reliable, sustainable, and high-performance manufacturing operations.

Collectively, the articles assembled in this Special Issue reflect the ongoing convergence of physical systems, computational intelligence, and human expertise. Spanning topics from embodied AI and digital twins to reinforcement learning, foundation models, predictive analytics, and sustainable operations, they reveal how intelligent manufacturing is evolving from isolated technological advances toward integrated, adaptive, and interconnected industrial eco-systems. Beyond advancing their respective research domains, these contributions also underscore the importance of interdisciplinary collaboration in addressing the complex challenges that accompany the next generation of manufacturing systems.

We sincerely thank all authors, reviewers, and editors for their valuable contributions to this Special Issue. We hope that the ideas, methodologies, and findings presented in this collection will inspire further exploration and foster collaboration across disciplines and industrial sectors. As intelligent manufacturing continues to advance, deeper integration among artificial intelligence, digital engineering, human-centered design, and industrial practice is expected to unlock new levels of autonomy, resilience, sustainability, and societal impact. We look forward to witnessing and contributing to this exciting transformation.