Monterey, CA; Naval Postgraduate School
TEXT MINING AND ANALYSIS ON NAVAL AVIATION MAINTENANCE ACTION FORMS
Abstract
dc:description.abstractNaval aviation generates extensive maintenance data, much of it recorded manually in unstructured free-text fields on Maintenance Action Forms (MAFs) and in structured oil analysis databases. These records contain valuable but underused information relevant to component health and failure modes. This thesis demonstrates the use of machine learning to classify maintenance discrepancy data and evaluates the feasibility of using Navy Oil Analysis Program (NOAP) oil analysis data to predict V-22 Proprotor Gearbox (PRGB) "oil-wetted" failures. We applied Latent Dirichlet Allocation (LDA) to free-text maintenance discrepancy records to classify each PRGB removal as an oil-wetted failure or a non-failure. Principal Component Analysis (PCA) reduced the dimensionality of Atomic Emission Spectroscopy (AES) wear metal data and LaserNet Fines (LNF) particle count data, and separate Cox Proportional Hazards (CPH) survival models were fit to each data source. The methodology is sound and replicable across platforms and components. However, the V-22 PRGB is a difficult subject for this type of analysis due to limited oil analysis history. We recommend adapting this approach to another Type/Model/Series (T/M/S) aircraft to definitively prove its predictive power before investing further in V-22-specific data collection.
Degree
thesis:*- Name thesis:degree_name
- Master of Science in Operations Research
- Level thesis:degree_level
- Masters
- Discipline thesis:degree_discipline
- Operations Research
- Department dc:contributor.department
- Operations Research (OR)
- Grantor dc:publisher
- Monterey, CA; Naval Postgraduate School
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bobay, Joseph M.
- Advisor dc:contributor.advisor
-
- Yoshida, Ruriko
- Committee member dc:contributor.committeemember
-
- Huang, Jefferson
Rights
dc:rights- Statement dc:rights
-
- This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.
- Distribution Statement A. Approved for public release: Distribution is unlimited.
Identifiers
dc:identifier.*- Dc Identifier Other
- NPS-26-N121-A
- Dc Identifier Programcode
- 360, Operations Analysis
- Thesis Thesis Identifier
- 42079