A Comparative Study of Manual Machine Learning and AutoML Approaches for Predicting Remaining Useful Life in NASA's Turbofan Engine Degradation

Aug 5, 2026

Upload by Skyler

Predicting the remaining useful life (RUL) of machinery, including NASA’s turbofan engines, is essential for maintenance and reliability. Manual machine learning methods often depend on domain expertise and extensive feature engineering, while AutoML aims to reduce effort by automating model selection, hyper-parameter tuning, and feature engineering. This study compares manual ML and AutoML for RUL prediction using the C-MAPSS dataset, evaluating accuracy, interpretability, and usability to identify when each approach is preferable in aerospace contexts.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages68

Reading Time2 h 51 min

AccessFree to download

Introduction

Background

Physics-based Technique

Data-driven Technique

Hybrid Technique

Machine Learning

Automated Machine Learning

Dataset Description

The Turbofan Engine

The Cost of Turbofan Engine

The Function of Turbofan Engine

The C-MAPSS Dataset

Proposed Methodology

A Comparative Study of Manual Machine Learning and AutoML Approachesfor Predicting Remaining Useful Life in NASA's Turbofan Engine Degradation A Thesis Presented to The Faculty of the Computer Science ProgramCalifornia State University Channel Islands In(Partial)Fulfillmentof the Requirements for the DegreeMasters of Science in Computer Science by Shuhui(Melinda),Yu December,2023 Master Thesis by Shuhui Yu ◎2023Shuhui YuALLRIGHTS RESERVED APPROVED FOR THE COMPUTER SCIENCE PROGRAM Dec 14,2023 Advisor:Dr.Michael Soltys Date Tason lsaacs Jason lsacs (Dec14,202316:53CST) Dr.Jason Isaacs Date Keun Som20231502P5)L Date Mr.Kevin Scrivnor APPROVED FOR THE UNIVERSITY 12/18/2023 Master Thesis by Shuhui Yu Non-Exclusive Distribution License In order for California State University Channel Islands(CSUCI)to reproduce,translate anddistribute your submission worldwide through the CSUCI Institutional Repository,your agreement tothe following terms is necessary.The author(s)retain any copyright currently on the item as well asthe ability to submit the item to publishers or other repositories. By signing and submitting this license,you (the author(s)or copyright owner)grants to CSUCI thenonexclusive right to reproduce,translate(as defined below),and/or distribute your submission(including the abstract)worldwide in print and electronic fomat and in any medium,including but notlimited to audio or video. You agree that CSUCI may,without changing the content,translate the submission to any mediumor format for the purpose of preservation. You also agree that CSUCI may keep more than one copy of this submission for purposes ofsecurity,backup and preservation. You represent that the submission is your original work,and that you have the right to grant therights contained in this license.You also represent that your submission does not,to the best ofyour knowledge,infringe upon anyone's copyright.You also represent and warrant that thesubmission contains no libelous or other unlawful matter and makes no improper invasion of theprivacy of any other person. If the submission contains material for which you do not hold copyright,you represent that you haveobtained the unrestricted permission of the copyright owner to grant CSUCl the rights required bythis license,and that such third party owned material is clearly identified and acknowledged withinthe text or content of the submission.You take fullresponsibility to obtain permission to use anymaterial that is not your own.This permission must be granted to you before you sign this form. IF THE SUBMISSION IS BASED UPON WORK THAT HAS BEEN SPONSORED OR SUPPORTEDBY AN AGENCY OR ORGANIZATION OTHER THAN CSUCI,YOU REPRESENT THAT YOUHAVE FULFILLED ANY RIGHT OF REVIEW OR OTHER OBLIGATIONS REQUIRED BY SUCHCONTRACT OR AGREEMENT. The CSUCI Institutional Repository will clearly identify your name(s)as the author(s)or owner(s)ofthe submission,and will not make any alteration,other than as allowed by this license,to yoursubmission AComparative Study of Manual Machine Laarning and AutoML Approachesfor Pradicting Remaining Usaful Lito in NASA's Turbofan Engine Dagradation Title of Item Manual Machine Learning.AutoML,remaining useful lite.turbofan,LSTM 3 to 5 keywords or phrases to describe the item Author(s)Name (Print) heluy Author(s)Signature Date 12/11/2023 A Comparative Study of Manual Machine Learning and AutoMLApproaches for # Predicting Remaining Useful Life in NASA's Turbofan Engine Degradation byShuhui(Melinda),Yu Computer Science ProgramCalifornia State University Channel Islands Abstract Predicting the remaining useful life (RUL)of machinery,such as NASA's Turbofanengines,is crucial for maintenance and reliability.Traditionally,this task has relied on manualmachine learning(ML)approaches that require domain expertise and extensive featureengineering.However,the emergence of AutoML(Automated Machine

FAQs

Why is remaining useful life (RUL) prediction important in this study?

RUL prediction supports maintenance planning and improves system reliability for machinery such as NASA’s turbofan engines.

What key difference does the thesis highlight between manual ML and AutoML approaches?

Manual ML typically requires domain expertise and substantial feature engineering, while AutoML automates steps like model selection, hyper-parameter tuning, and feature engineering.

Which dataset is used to evaluate the methods and what aspects are compared?

The C-MAPSS dataset is used, and comparisons focus on predictive accuracy, model interpretability, and ease of use.

Aug 5, 2026

Upload by Skyler