Cloud-Based Internet-of-Things System for Long-Term Bridge Bearing Monitoring Using Computer Vision

Bearings play a crucial role in mitigating loads, maintaining stability, and transferring forces between superstructures and substructures. However, bearing failures caused by external factors can compromise structural safety. Therefore, continuous monitoring of bearing displacement is essential, yet current inspection methods are labor-intensive and unsuitable for long-term management. To address this, researchers have proposed systems such as Linear Variable Differential Transformers (LVDTs) and computer vision-based monitoring methods to track bearing displacement over time. However, reliance on external power sources and complex installation processes has limited their widespread application. This paper proposes an automated monitoring system integrating low-power IoT sensors, computer vision, and cloud computing. The system features an event-driven power mechanism to minimize energy consumption and utilizes vision-based displacement measurement techniques, providing both portability and efficiency. Applied in a real-world setting for nine months, the system successfully enabled the long-term monitoring of bridge bearings. The results demonstrate its effectiveness in overcoming traditional limitations and highlight its potential in supporting automated, data-driven assessments of structural stability.