The Hong Kong-Zhuhai-Macao Bridge is a cross-sea bridge connecting Hong Kong, Zhuhai, and Macao, spanning 55 kilometers. The bridge features a "bridge-island-tunnel" design, including a 6.7-kilometer undersea tunnel and two artificial islands, with a total investment of approximately 120 billion yuan.

As the world's longest cross-sea bridge, it not only shortens the travel time between the three regions but also plays a significant role in promoting regional economic integration and strengthening economic cooperation. This bridge is not only a significant achievement in China's transportation infrastructure construction but also one of the world's largest and most technologically complex cross-sea bridges.

In a study published in China Ocean Engineering researchers from the Hong Kong-Zhuhai-Macao Bridge Authority and Wuhan Maritime Communication Research Institute introduced an AI-driven real-time crack detection system.

This new method uses an attention-based deep learning model that can quickly and accurately segment and identify structural cracks. Traditional crack detection methods often face issues of low efficiency and inaccuracy that this technology has greatly improved. Researchers have successfully combined machine vision with deep learning algorithms to create a system capable of analyzing bridge surface images and identifying cracks with high precision.

The research team also used underwater robots to collect detailed data on the underwater structures of the Hong Kong-Zhuhai-Macao Bridge. These underwater images and sensor data were used to construct a custom dataset, further validating the model's effectiveness in real-world scenarios.

This technology will completely change the way we monitor and maintain cross-sea bridges. It provides a new perspective for the construction of real-time and accurate automated crack detection systems, which is crucial for intelligent maintenance and operation.

The implementation of this technology can save millions of yuan in maintenance costs and significantly improve the safety and service life of cross-sea bridges.

More information: Yong-chuan Zhou et al, A Rapid Crack Detection Technique Based on Attention for Intelligent M&O of Cross-Sea Bridge, China Ocean Engineering (2024). DOI: 10.1007/s13344-024-0068-0

Provided by TranSpread