Honghong Song

ORCID: 0009-0000-3136-2598
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Infrastructure Maintenance and Monitoring
  • BIM and Construction Integration
  • Infrastructure Resilience and Vulnerability Analysis
  • Manufacturing Process and Optimization
  • Structural Integrity and Reliability Analysis
  • Concrete Corrosion and Durability
  • Geotechnical Engineering and Underground Structures
  • Design Education and Practice
  • Digital Transformation in Industry
  • Evacuation and Crowd Dynamics
  • Structural Health Monitoring Techniques
  • 3D Surveying and Cultural Heritage

Dalian Maritime University
2023-2025

Cardiff University
2023

Digital twin (DT) has been moving progressively from concept to practice for bridge operation and maintenance (O&M), but its issues of data synchronization fault tolerance remain problematic. This paper investigates the time delay DT services according communication computation complexity, revealing distinct impact their sequence, proposes an AIoT-informed framework solve above issues. The information hierarchy two-way can be leveraged minimize complexity in framework. Meanwhile, flow...

10.1016/j.autcon.2023.104835 article EN cc-by-nc-nd Automation in Construction 2023-03-22

Modern structural design must balance criteria with increasing objectives like cost minimization, carbon reduction, and stakeholder interests. However, this multi-domain knowledge exists in unstructured forms, such as text, formulas, tables, converting it into machine-readable structured within a unified framework remains challenging. This paper proposes an ontology-based modeling mapping approach to transform from specifications, cost, emissions knowledge. enables self-containing compliance...

10.1080/17452007.2025.2471078 article EN cc-by Architectural Engineering and Design Management 2025-03-03

Photographs of the damage collected at bridge sites are essential reference data for repair decisions. Bridge recommendations depend heavily on manual assessments by experts, which time-consuming and labour-intensive. To address this challenge, a multimodal deep-learning-based method is proposed in study, can automatically generate from image. In deep-learning model, CNN used to extract image features, Transformer efficiently capture relationships between visual features images text...

10.2139/ssrn.4804304 preprint EN 2024-01-01
Coming Soon ...