A STUDY ON THE CONSTRUCTION OF CLOUD-BASED TUNNEL FACE AI DIAGNOSTIC SYSTEM

Abstract

This paper proposes an AI diagnostic system built on a cloud service. In mountain tunnel construction, the condition of the face, weathering, cracking, and the amount of spring water are observed during excavation. Support selection is conducted using the evaluation results of the ground. When observing the face, field engineers evaluate it. However, it is expected that an objective diagnostic system will be constructed from the viewpoint of supporting the evaluation. We evaluate the accuracy rate between engineers and AI for each observation item.