Session: 01-01-01 Offshore Platforms-1

Submission Number: 179680

Multi-View RGB-D Fusion for Structure Position Estimation in Offshore Wind Turbine Installation 

To meet the continuously increasing global energy demand while reducing environmental and climate impacts, sustainable energy sources, especially wind power, are expected to play an increasingly important role. In many regions, offshore wind resources are more abundant than those onshore. However, offshore wind turbine (OWT) installation remains challenging due to harsh marine conditions and the need for specialized heavy lifting equipment. OWTs are typically installed in a segmented manner using crane vessels, where accurate knowledge of the relative position between the suspended and fixed components is crucial. Reliable position estimation enhances operational safety, improves installation efficiency, and reduces equipment costs.

Traditional position measurements based on GNSS and IMU sensors provide accurate results but require physical contact and additional installation steps, which makes them less suitable for offshore operations. Vision based methods have made significant progress in object position estimation, but their accuracy is often affected by illumination and environmental conditions. To address these challenges, this study presents an enhanced visual monitoring method that fuses RGB and depth information from multiple viewpoints. Compared with a single view approach, where a single camera placed at a distant side view was used to track the relative motion between the suspended and fixed structures, the proposed system introduces an additional upward facing camera located near the base of the fixed component. This new viewpoint captures the lifting process from the barge to the top alignment phase with improved precision. By combining observations from both views, the proposed multi view fusion framework achieves more accurate and reliable position estimation for OWT installation under complex visual environments.

Presenting Author: Hervin Vongi Juan Luy Tsinghua Shenzhen International Graduate School

Presenting Author Biography: Hervin Vongi Juan Luy is currently pursuing a Ph.D. in Ocean Engineering at Tsinghua Shenzhen International Graduate School, Tsinghua University. He received his M.S. degree in Engineering Fluid Mechanics for the Offshore, Coastal and Built Environment from Imperial College London and his B.Eng. degree in Mechanical Engineering from the University of Sheffield. His research focuses on structural monitoring for offshore operations using computer vision and depth sensing. His recent work aims to develop multi-view visual fusion techniques for accurate pose estimation during offshore wind turbine installation.

Authors:

Hervin Vongi Juan Luy Tsinghua Shenzhen International Graduate School
Jing Dong Shanghai investigation, design & research institute Co.,Ltd.
Zhenzhong Hu Tsinghua Shenzhen International Graduate School
Zhengru Ren Shanghai Jiao Tong University