Classification of Structural & Architectural Maintenance by K-means Clustering : A Case Study of Purple Line Metro Project

Abstract

This research aims at clustering of the type and number of repair workgroups arising from the passenger's service using K-Mean Clustering Analysis using passenger boarding and alighting at the station and equipment damage as per actual structural and architectural works of the purple metro line in January 2020. The data obtained are compared with regard to the rate of equipment damage per million passengers boarding and alighting at the station. The K-mean clustering analysis results in 3 clusters, namely 1) One station with a high number of passengers, 2) Six stations with moderate amount of passengers, and 3) Nine stations with a low amount of passengers. According to the cluster analysis, it was found that there are frequent calls for repair and maintenance work on the drain cover and the handrail installed in the station cluster with a high number of passengers which accounts for 7.58 and 3.79 times per one million passengers, respectively. On the other hand, there are frequent calls for repair and maintenance work on service compartments in stations, station service shaft, roller shutter doors, and warning signs installed in the station cluster with a low number of passengers which accounts 14.06, 4.68, 8.89, and 7.52 times per one million passengers, respectively.