Lotte Construction announced on the 30th that it has developed and filed a patent for a tracking system that visualizes cracks occurring on the back side of earth retaining temporary structures based on artificial intelligence (AI).
The crack tracking system for the back side of earth retaining temporary structures allows construction site workers to capture site videos with personal recording devices and upload them to a platform, where an AI model analyzes the videos to visualize crack information. Through history management of the extracted crack data, the system enables comparison and analysis of crack progression status.
Typically, earth retaining temporary structures are installed at construction sites during underground excavation to prevent soil collapse or groundwater intrusion. However, these structures often collapse due to subsidence and cracks on the back side.
Monitoring of subsidence and cracks on the back side of temporary structures has rarely been conducted, resulting in situations where the condition is only visually assessed after major disasters such as ground subsidence occur.
In response, Lotte Construction developed the industry's first system that uses AI-based video analysis technology to analyze images of the road surface behind earth retaining temporary structures through deep learning, providing risk alerts to managers.
The company explained that by introducing the system, the time required compared to the traditional method of measuring crack progression with a tape measure is significantly reduced, and it becomes possible to warn of risks and take preemptive measures based on the documented crack history.
Lotte Construction plans to collaborate with specialized IT companies to apply the system across all sites in the future.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
