Roller element bearing remaining useful life prediction: ML hybrid approach

A roller element remaining life predictor

Problem Statement

Bearing degradation is sometimes severe and is the major cause of machinery failure. This project aims at predicting the remaining useful life of the ball bearing by applying ANN and various ML algorithms. Further optimisation of the neural network was achieved by applying genetic algorithm, thus, significantly reducing the error.

Technology used

Python, Pandas, Keras, Genetic Algorithm, Sklearn

Project link

Major Report