Year: 2024
Type: article
Abstract: Structural monitoring of quay walls, where various events occur due to unexpected high waves, vessels, and heavy equipment, is essential. However, real-time events cannot be constantly monitored by on-site personnel. To resolve the aforementioned issues, this study proposes an innovative AI-powered, cloud-based wireless sensor system that incorporates a high-sensitivity accelerometer with an ultra-low noise level of 0.003 mg, designed to monitor the low response amplitude of massive quay walls. The sensor can be activated by a scheduled trigger or a long-rangefinder. Vessel detection is performed utilizing the AI-based object detection method, Faster R-CNN, which employs ResNet as the backbone network. The detected anchor box’s position and dimensions are subsequently processed to confirm the presence of a berthing vessel. The collected data are then transmitted wirelessly to a proposed cloud server through LTE communication in real-time. The developed system was installed on a caisson-type quay wall in Korea, where acceleration, tilt, temperature, and camera image data were analyzed to assess its performance for real-time event monitoring. The results demonstrated that the safety of quay walls can be automatically managed by monitoring events during berthing and mooring with the proposed system.
DOI: 10.3390/jmse12111902
Authors: Junsik Shin, Junyoung Park, Junyoung Park, Jongbin Won, Jongwoong Park, Jongwoong Park, Jiyoung Min
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,955
Awards: G53423507, G6286322158, G7095576870
OpenAlex ID: W4403734022