As the industrial digital transformation progresses deeper, new scenarios and applications have placed higher demands for industrial network infrastructure, encompassing ubiquitous connectivity, deterministic transmission, integration of control networks and computing resources, open intelligence, and secure controllability. Industrial 5G stand-alone non-public network (SNPN), leveraging its differentiated advantages in asset exclusivity, resource specialization, flexible deployment, autonomous management, and security reliability, are expected to serve as a powerful supplement to traditional industrial networks and operator 5G virtual/hybrid private networks. Based on the core requirements of industrial scenarios, this paper elaborates on the system architecture and functional characteristics of the industrial 5G SNPN for building a new industrial network exclusive digital foundation, clarifying its key capabilities in industrial application support, computing network collaboration, and network security protection. Through the construction of an integrated digital foundation of “connection management-computing & network collaboration-security protection”, the industrial 5G SNPN provides universal connectivity, guaranteed transmission, and autonomous controllability as the underlying support for the new industrial network, facilitating the transition from automation to intelligence and providing strong support for the digital transformation of industries.
Single Pair Ethernet (SPE) and its evolution into ethernet advanced physical layer technologies enable long-distance data transmission and intrinsic safety power delivery over a single twisted pair cable, breaking down the traditional physical barriers between OT and IT. This paper provides an in-depth analysis from four dimensions: policy drivers, technical standards, network architecture transformation, and typical unmanned application scenarios. It demonstrates how SPE, by establishing an “end-to-end” full-IP architecture, provides critical physical layer support for the process industry in achieving predictive maintenance, unmanned inspections in hazardous environments, and the foundational structure for digital twin models.
With the rapid advancement of integrated communication-sensing technology in future 6G networks, the deep integration of communication and sensing functions not only enhances system performance but also introduces multidimensional security challenges. From a security perspective, this paper analyzes the composite security risks of integrated communication-sensing systems at the physical layer, privacy protection, and in traditional communication versus 6G new scenarios. Specifically, physical layer security faces threats such as channel state spoofing, environmental reflection manipulation, and joint attacks on communication and sensing signals. Privacy protection requires addressing risks of user location information leakage, behavioral trajectory exposure, and high-dimensional sensing data breaches. Simultaneously, the system must still meet fundamental security requirements including identity authentication, data integrity, and low-latency interaction. To address these security needs, key technological advancements are summarized, including lightweight RF fingerprint authentication, deep learning-based anomaly detection, perception data protection combining differential privacy and federated learning, resource regulation and joint scheduling mechanisms driven by deep reinforcement learning, as well as blockchain-based distributed trust management and lightweight edge collaboration strategies. By integrating these technologies, system-level security safeguards can be established across multiple dimensions including physical layer, data processing, resource management, and cross-domain collaboration, providing theoretical support and technical references for secure and trustworthy operation of integrated communication-sensing systems in high-dynamic, high-density, and strongly real-time environments.
Under the global wave of industrial digitalization and intelligence, traditional manufacturing enterprises face systematic challenges in their digital transformation, including heterogeneous systems, data silos, difficulties in cross-entity collaboration, and high costs associated with large-scale replication. New industrial network technologies, serving as next-generation industrial infrastructure that integrates advanced communication, computing, and control technologies, provide a critical pathway for constructing a unified digital foundation to support enterprise-wide transformation. This paper systematically explores how to drive the construction of group-level data platforms and intelligent platforms by establishing a “cloud-edge-device” collaborative industrial network architecture. The research focuses on analyzing the application mechanisms of advanced technologies within the core model of “capability precipitation and rapid replication.” By integrating scenarios such as intelligent application promotion, data closed-loop processes, and integrated group operations, it elaborates on viable pathways for technology-enabled business transformation. Furthermore, this paper provides an in-depth analysis of the core challenges encountered during the transformation process—including data governance, technology integration, organizational change, and security controllability—along with corresponding coping strategies. This case study demonstrates that leveraging new industrial networks as the foundation and guiding the overall strategy with a unified architecture represents an effective paradigm for group-based manufacturing enterprises to advance from local optimization to holistic synergy, and from technology application to model innovation.
At present, the integration of Industrial Computing and Network Convergences and industrial artificial intelligence still faces such challenges as poor versatility of solutions, which leads to high investment and difficulties in promotion and application. Most integrations are still at the system adaptation stage, and the lack of deep integration with business processes and management control mechanisms leads to fragmented management and insufficient collaboration. By examining the current state of technological convergence, this paper analyzes the problems of the mismatch between supply and demand, insufficient depth of integration, potential risks in system security and the standard development is still in the initial stage. Meanwhile, this paper looks forward to the future development trends in terms of technology, application and ecology, and puts forward targeted suggestions from three dimensions: technological tackling, guidance and support, and ecological construction. With continuous breakthroughs in technological innovation and deepening ecological synergy, integrated development will gradually overcome current bottlenecks, delivering multiple benefits to the industrial sector—significant improvements in production efficiency, substantial reductions in energy consumption, and marked enhancements in self-reliance. This will propel China’s industrial system toward smarter, greener, and more integrated.
Against the backdrop of the accelerated development of industrial intelligence, industrial data has become a key production factor. However, its low quality, heterogeneity, fragmentation, and semantic disconnection have restricted cross-system applications. This paper proposes a technical architecture for industrial data fusion and interoperability, systematically studying three key technologies: data standardization, data quality processing, and data annotation. It constructs a standardized promotion path, introduces a large model-driven dynamic self-closed-loop quality processing system, and establishes a scenario-oriented annotation evaluation model to achieve controllable and quantifiable assessment of the annotation process. The related research provides methods and technical support for the construction of a high-quality industrial data application environment.
As the first step to promote the formation of a national integrated data market, the practical exploration of data registration is in full swing and has differentiated into two major development paths: the combined registration path integrating data property rights and intellectual property rights, and the independent registration path for data property rights. A comprehensive review of current local data registration practices reveals that data registration has not fulfilled its expected role, and the root cause lies in the lack of a unified national registration system. In view of this, coordinated efforts should be made in three aspects: improving the institutional framework, enhancing the technical system, and optimizing the organizational structure, to establish a data registration system that meets the development needs of data elements and the construction goals of a national integrated data market.
Based on the technology acceptance model, this study explores the driving factors and mechanisms of digital leadership transformation in the era of artificial intelligence through theoretical analysis and empirical research. Research has found that leaders exhibit significant hierarchical differences in their acceptance of technology. Senior managers are more concerned with the strategic value of technology, mid-level managers focus on the feasibility of technology implementation, and grassroots managers rely more on organizational support. Through structural equation modeling verification, it was found that technological infrastructure and industry competitive pressure are the core driving factors, while the impact of changes in customer demand did not reach a significant level. Research has shown that enhancing perceived usefulness and perceived ease of use can effectively improve the efficiency of digital leadership transformation, providing theoretical support and implementation framework for the dynamic evolution of digital leadership.
Based on the current development of cybersecurity situation, this article examines the evolution trends of cybersecurity industry policies in the Asia-Pacific region. By defining conceptual boundaries and selecting typical countries and regions, such as the United States, Russia, Australia, South Korea, Japan, and ASEAN nations for horizontal and vertical comparisons, the research provides a summarized analysis from perspectives including policy differences, overarching strategies, and specialized sectors, clarifying the developmental trajectory of cybersecurity industry policies across the Asia-Pacific region. In light of both international and domestic security dynamics, targeted recommendations are proposed to advance China’s cybersecurity industry and effectively address external challenges, focusing on areas such as talent development, industrial foundation, and international cooperation.
Traditional artificial neural network training typically focuses on closed, static, independent and identically distributed data, and performs a single task after completing offline training. However, when the data distribution continuously changes with the environment, the model will forget the knowledge learned from previous tasks, a phenomenon known as “catastrophic forgetting”. As an emerging learning paradigm, continual learning aims to endow models with the ability to continuously learn, accumulate, and consolidate knowledge from data streams with constantly changing distributions. This enables artificial neural networks to achieve a “stability-plasticity” balance, thereby overcoming catastrophic forgetting. Through in-depth analysis of the key characteristics of current continual learning algorithms, a real-world robotic physical verification platform was established. The effectiveness of continual learning algorithms was verified in the scenario of robotic physical object grasping. Experimental results show that when the Contrastive Correlation Preserving Replay (CCPR) algorithm is applied to the robotic physical object grasping task, the average accuracy of the grasping task increases by 26.67%, better assisting the robot in performing the target task.
This study explores the application of artificial intelligence in the construction of smart wind farms, empowering smart wind power and addressing issues such as operational efficiency and equipment fault diagnosis. A comprehensive intelligent application process is established, covering data collection, governance, feature engineering, modeling, and application. Through the implementation of this process, wind farms have achieved smartization throughout their entire lifecycle, including planning, operation, maintenance, and decommissioning. A case experiment on short-term power prediction verifies the effectiveness of simulated annealing optimized long short-term memory and the engineering applicability of the process.The innovative full lifecycle intelligent management paradigm proposed in this research provides technical support for the smart transformation of the wind power.
Despite the substantial body of international research developed over decades, scholarly investigation into deceptive and misleading behaviors, commonly known as dark patterns, within the context of Chinese internet applications is still limited. This gap has resulted in a lack of theoretical underpinning for corresponding governance frameworks. To inform relevant research and policy-making, this paper reviews scholarly progress in regulating dark patterns, clarifies their conceptual definition and manifestations within digital governance, and examines their adverse effects on users and society through real-world case studies. It concludes by focusing on the key regulatory challenges and potential pathways forward.