2026-08-30

The paper, titled "Recovering layer information for precise component segmentation from 2D drawings", has been published in Advanced Engineering Informatics.

2D drawings offer a cost-effective data source for BIM reconstruction of existing buildings, yet automated component segmentation remains challenging due to missing layer information in raster drawings. To address the problems of imprecise boundaries in existing methods and the semantic ambiguity caused by neglecting line-symbol relationships, this study proposes a two-module framework: first, a Graph Convolutional Network (GCN)-based method that incorporates line-symbol proximities into a heterogeneous graph for layer recovery; and second, a rule-based method that reconstructs vectorized component contours from the recovered layers, while resolving incomplete contours caused by dashed lines and layer recovery errors. On residential building drawings, the framework achieves an F1 score of 84.3% for layer recovery and a mean IoU of 80.1% for component segmentation, with improved boundary consistency over CNN-based baselines. Further experiments on office building drawings and paper drawings demonstrate the framework's applicability. A preliminary end-to-end BIM modeling test shows that the generated contours require less boundary post-processing than CNN-based masks, providing a practical foundation for BIM reconstruction of existing buildings.

Note: Advanced Engineering Informatics is a Q1-ranked top journal in the field of engineering and technology. Its 2026 impact factor was 11.5. The first author of the paper is Lu Sun, a doctoral student from the Hong Kong University of Science and Technology. Yu Yantao from the Hong Kong University of Science and Technology is the corresponding author. The research results were funded by the Innovation and Technology Fund Seed Research Project, the RGC Theme-based Research Scheme, the Guangdong NSF, and the Guangdong Basic and Applied Basic Research Foundation.

2026-08-25

The research group's latest software achievement, "Smart Marine Disaster Early Warning Platform V1.0," has been granted a Computer Software Copyright Registration Certificate (Registration No.: 2026SR0865277).

The Smart Marine Disaster Early Warning Platform is an integrated visualization application software designed for marine environmental data integration, disaster process display, thematic simulation analysis, risk assessment, and marine engineering safety evaluation. Within a unified web portal, the platform organizes four core business modules—environmental digital twin, typhoon disasters, pollution dispersion, and marine engineering analysis—enabling users to view multi-source data, replay temporal processes, query specific locations, analyze result curves, conduct structural engineering assessments, and visualize specialized scenarios.

The platform employs an interactive interface combining map scenes, 3D scenes, sidebars for business functions, timelines, statistical charts, and analytical pop-ups, allowing users to seamlessly perform environmental monitoring, case switching, process playback, result queries, engineering analysis, and data export around a single operational object.

The development team also includes doctoral student Li Yilin, postdoctoral researcher Liu Yi, Associate Professors Li Sunwei, Chen Shengli, Li Binbin, and Jing Lu. This research was supported by the Guangdong, Hong Kong and Macao Team Project of the Guangdong Regional Joint Foundation.

2026-08-21

The 2026 National Conference on Solid Mechanics was successfully held in Tianjin from August 21 to 24. Organized by the Solid Mechanics Committee of the Chinese Society of Theoretical and Applied Mechanics, the conference focused on the theme "Frontiers and Challenges in Solid Mechanics" and featured various academic exchange formats, including plenary sessions, specialized symposia, and poster presentations. With one main session and 29 sub-sessions, the event brought together over 3,000 participants, including more than 20 academicians and experts in solid mechanics, making it the largest comprehensive academic conference in China's mechanics community in 2026.

On August 23, during the special session MS28 titled "Mechanical Problems in Marine and Ship Engineering," I delivered an invited lecture entitled "Digital Twin Method for Simulation Analysis and State Inversion of Offshore Structures under Complex Marine Environments." The presentation discussed simulation analysis, state inversion, and digital twin methodologies for offshore structures in complex marine conditions. It introduced a research framework integrating environmental data, structural models, and monitoring information to analyze structural mechanical states, and shared insights on applying digital twins for intelligent perception and safe operation and maintenance of marine engineering structures. Dr. Jiang Chunhao, a Ph.D. student from our research group, presented an oral report titled "Virtual Sensing Method for Offshore Wind Turbine Structures Based on Subspace Spectral Graph Networks" and also displayed related research findings through a poster presentation. Participation in this conference strengthened academic exchanges with domestic peers and further broadened our research perspectives in digital twins, intelligent sensing, and safety operations for marine engineering structures.

Note: The Chinese Society of Theoretical and Applied Mechanics, established in February 1957, is a national academic organization under the China Association for Science and Technology, dedicated to advancing the development of mechanics and promoting academic communication.

2026-08-19

The paper "A digital twin-based framework for simulation and monitoring analysis of floating wind turbine structures," published by the research group, has surpassed 100 citations on Google Scholar. It is worth noting that as of this year alone, this article has been cited 38 times.

The research has attracted significant attention from the academic community, primarily because it addresses key technical challenges in offshore wind power development. To tackle the difficulties of managing multi-source and heterogeneous data in the design and operation of floating wind turbines, the study proposes a unified framework based on digital twin technology. By leveraging three-dimensional models to achieve real-time synchronization and inversion of sensor data, it effectively resolves challenges related to data integration and simulation analysis throughout the turbine's entire lifecycle. Meanwhile, the proposed technical system features a clear pathway, has been validated through engineering practice, and offers strong methodological reference value and broad applicability across various scenarios.

The research group has also published several papers related to digital twin technology. Among them, the paper "The Application and Development of Digital Twin in the Marine Domain" focuses on digital twin technology for ship applications, providing a detailed discussion on concepts, application frameworks, key technologies, and future development in the field of marine digital twins, and has been cited 9 times on Google Scholar. The paper "Real-time Monitoring and Analysis for Digital Twin of Floating Wind Turbine" proposes a monitoring and analysis framework suitable for floating wind turbine digital twins, which has been cited 3 times on Google Scholar. Meanwhile, "Research Prospect of BIM-Based Information Technologies for Engineering Management" explores the future development of digital twins in engineering from the perspective of building information technology, and has been cited 43 times on China National Knowledge Infrastructure (CNKI).

2026-08-17

Recently, the journal "Ocean" (published by Tsinghua University) has been officially included in the EuroPub database (the European Academic Publishing Center Database). The journal is edited by Professor Torgeir Moan from the Norwegian University of Science and Technology, Academician Wang Zhonglin from the Chinese Academy of Sciences, and Academician Zhang Jianmin from Tsinghua University.

The EuroPub database (https://europub.co.uk/) is a well-known comprehensive and multidisciplinary open-access academic journal index database covering various fields such as natural sciences, engineering technology, and social sciences. Currently, it includes over 27,000 active and authoritative journal index records and more than 700,000 articles. Most of these journals are also indexed by Web of Science and Scopus. This inclusion in the EuroPub database marks a solid step forward for the Ocean journal in terms of international dissemination and academic influence, and also builds a broader international academic exchange platform for global marine research results.

Ocean focuses on the latest research progress in marine science, marine engineering, and marine technology, covering areas such as marine biology, marine earth science, marine automation and robotics technology, and underwater engineering. We sincerely invite submissions of articles with leading-edge, innovative, and practical content related to the marine field, to jointly promote the development of marine disciplines and technological innovation!

Submission link: https://mc03.manuscriptcentral.com/ocean.