Multiclass classification for predicting Remaining Useful Life (RUL) of the turbofan engine

For complex equipment and systems or their subsystems maintaining good health is very important. Several approaches were tried by the researchers applied by the industry professionals to predict the health of equipment, to take corrective measures. In this context, data analysis and machine learning are the latest tools that use historical data of equipment to predict the health or remaining useful life of the equipment and classify it into three categories based on predicted health namely, good, poor and average. Random forest algorithm was employed for the prediction of the health of a turbofan engine from data obtained using a simulation model C- MAPPS. The data contained attributes containing three different operating conditions along with 28 sensors to record the conditions in which the engine is running. The model gives an accuracy of 81% to predict the correct health condition of the turbofan engine.