Abstract
Predictive maintenance shows significant impact in maintaining and monitoring the utility systems at commercial buildings. Here in this research paper, the predictive maintenance that is in line with Industry 4.0 is applied. The chilled water system is highlighted in this research and the main goal is to reach out the most occurred faults in one of the chilled water system components, which is the chiller. An engineering management framework was used through three parts, set up, machine learning, and quality control. One of the most accurate machine learning algorithms, which is the decision tree algorithm, is applied in this research and prediction model showed high prediction accuracy. During an empirical period, the most occurred faults were refrigeration leak, low condenser flow, and high condenser temperature.