Semi-Autonomous Labeling of Unstructured Maintenance Log Data for Diagnostic Root Cause Analysis

Published

September 7, 2016

Author(s)

Michael E. Sharp, Thurston B. Sexton, Michael P. Brundage

Abstract

To facilitate root cause analysis in the manufacturing in- dustry, maintenance technicians often ll out \maintenance tickets" to track issues and corresponding corrective actions. A database of these maintenance-logs can thus provide problem descriptions, causes, and treatments for the facility at large. However, when similar issues occur, di erent technicians rarely describe the problem in an identical man- ner. This leads to description inconsistencies within the database, which makes it dicult to categorize issues or learn from similar cause-e ect relationships. One way to solve this problem is via natural language processing (NLP) techniques to properly tag similar ticket descriptions, allowing for more formalized statistical learning of patterns in the main- tenance data as a special type of short-text data. This paper showcases a proof-of-concept pipeline for merging multiple machine learning (ML) and NLP techniques to cluster and tag maintenance data, as part of a broader research thrust to extract structured data from largely un- structured natural-language diagnostic databases. The accuracy of the proposed method is tested on real data from a small manufacturer.

Proceedings Title

APMS 2017 International Conference Advances in Production Management Systems (APMS 2017)

Conference Dates

September 3-7, 2017

Conference Location

Hamburg

Pub Type

Conferences

Download Paper

Local Download

Citation

Sharp, M. , Sexton, T. and Brundage, M. (2016), Semi-Autonomous Labeling of Unstructured Maintenance Log Data for Diagnostic Root Cause Analysis, APMS 2017 International Conference Advances in Production Management Systems (APMS 2017), Hamburg, -1, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=923144 (Accessed December 13, 2024)