Designing Machine Learning Model for Predictive Maintenance of Railway Vehicle
Hafid Galih Pratama Putra, Suhono Harso Supangkat, I Gusti Bagus Baskara Nugraha, Fadhil Hidayat, PT Kereta
Abstract
Indonesia’s geographical layout makes it challenging to create a single united railroad network with many branches. Instead, the railroad operation is divided by location, mainly between large islands such as Java and Sumatra. Therefore, distributing appropriate resources for each operational area is hard to manage properly. Kereta Api Indonesia, which provides rail transportation services, needs to address and change its various business processes to achieve Digital Transformation. Among them, asset maintenance is crucial in railroad operations. This paper will focus on designing machine learning models using actual data available from railway vehicles such as generator trains to predict their condition for maintenance by utilizing classifying algorithms for machine learning, which can help automate routine checks. As a preliminary research paper, testing and evaluation of the machine learning model are not available yet.