ANALYSIS OF FACTORS AFFECTING DOWNTIME IN RC BUILDINGS CONSIDERING CUMULATIVE DETERIORATION USING MACHINE LEARNING

Abstract

While current seismic design primarily aims to ensure life safety during major earthquakes, growing societal demand has shifted the focus toward resilience performance, emphasizing early functional recovery and reduced downtime. A critical issue is to evaluate the impact of both the delay time before repair initiation and the repair time of structural and non-structural components on downtime. This study investigates how cumulative deterioration caused by repeated seismic events during the service life of RC buildings influences downtime after a major earthquake. Furthermore, a machine learning-based importance analysis is conducted to identify key factors influencing downtime, supporting effective recovery strategy development.