Engine Health Status Prediction Based on Oil Analysis With Augmented Machine Learning Algorithms

Tribology is the very efficient and strong tool in machine operations analysis. In the article author presented how the artificial intelligence algorithms could be applied to help in engine oil test results analysis. Based on the real-life turbofan engine oil sample test results dataset, the novel methodology of the machine learning algorithm implementation was presented. In order to take advantage of the artificial intelligence in engine oil test results interpretation, the augmented engine oil dataset was generated with additional predictors. Research case study was conducted for both original engine dataset as well as the augmented one. For the scientific purposed, various machine learning performance metrics were calculated, what allowed to precisely compare the results achieved for the original dataset and the one generated on the basis of the proposed novel method. The greatest achievement of the article was the presentation of the new methodology implementation in the real-life turbofan engine health status prediction. Presented methodology implemented into the aircraft (engine) maintenance management computer system allows to automate engine health status analysis and improve engine maintenance management.