Improving Industrial Maintenance: Leveraging ML and NLP for better Breakdown Log Analysis and Automated Workflows
2024 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 HE credits
Student thesis
Abstract [en]
The reason for this thesis is the study of solutions to improve the workflow of companies dealing with machine breakdown logs. The current ways of working with these logs is tedious and not very efficient as it requires manual work, task duplication and highly trained operators.
Through an iterative process of problem analysis and solution deployment, the key points of the study are reached, leading to the proposed solutions.
The study demonstrates the potential of using NLP and LLMs for the simplification and explanation of robot breakdown logs as well as the usefulness of automated alert generation based on benchmark breakdown data and the potential improvement of workflow through the automation of the maintenance work order creation.
The thesis has resulted in contributions from the practical as well as the scientific per-spective. The thesis contributes to the practical field by solving concrete problems from the two case companies, and thus providing real solutions that are directly applicable and can be extrapolated to be used by other companies and in other fields. The thesis contributes to the research in the field by creating a new method of using AI technologies to optimize maintenance workflows.
Place, publisher, year, edition, pages
2024. , p. 62
Keywords [en]
Maintenance, Maintenance Logs, Machine Learning, NLP, CMMS
National Category
Robotics and automation
Identifiers
URN: urn:nbn:se:hv:diva-22200Local ID: EXR600OAI: oai:DiVA.org:hv-22200DiVA, id: diva2:1887145
Subject / course
Robotics
Educational program
Master i robotik och automation
Supervisors
Examiners
2024-08-232024-08-072025-09-30Bibliographically approved