A Clustering-based Approach for Reorganizing Bus Route on Bus Rapid Transit System

Abstract

The passengers of transit-oriented public transportation have experienced growth. Reorganizing public transportation routes must be proposed in order to adapt to the increasing number of passengers. This research aims to reorganize bus routes by clustering bus stops on every route of transit-oriented public transportation route in order to meet the customer demand and suit the planned routes with the actual condition in its operation. Data sets for this research were provided by TransJakarta, a bus rapid transit (BRT) system in Jakarta, Indonesia. Using K-Means clustering algorithms, it is found that the optimal cluster is 15 clusters with several bus stops contained in it that adjusted and balanced with the number of passengers on each bus stops so that there is no inefficiency of utility or overcapacity. Elbow method is used to determine the optimal number of clusters in this research. This research can provide insight to transit-oriented public transportation in Indonesia by performing reorganizing bus routes with the new clusters of bus routes, it is expected to be able to meet the demands of each bus stops and load balancing can occur on every bus route that leads to increase the customer satisfaction and the number of public transportation users continues to increase.