Maintenance optimisation based on historical data: A case study of a food manufacturing company in Brazil

Tools

QIU, Huixuan (2019) Maintenance optimisation based on historical data: A case study of a food manufacturing company in Brazil. [Dissertation (University of Nottingham only)]

[img] PDF - Registered users only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (4MB)

Abstract

This study is a company-based project. It presents an approach to generate an optimal maintenance planning for a specific machine via combining Failure Mode and Effective Analysis (FMEA) and K-means clustering algorithm based on historical failure data of the machine for the aim of enhancing maintenance management and maintenance optimisation.

An introduction of maintenance planning and problem description of the case company are discussed in the beginning of the paper. The review of Maintenance Optimisation Model, Reliability Theory and FMEA method will be present next as the background of the study. The paper applies both qualitative and quantitative method, thus the main body contains two parts, where firstly presents case study analysis about description of how FMEA method to be applied in the case and evaluation of effectiveness of FMEA procedure for the company, secondly conducts 7 experiments regarding to K-means clustering algorithm on historical failure data in R to further support generating maintenance plan. One of the experiments is selected as the optimal clustering result, then the 140 failure records are clustered into 4 groups which have different criticality level and the engineer within the company finally allocates variable type of maintenance actions to each failure based on their criticality in order to reduce the occurrence of failure while company eventually achieve the goal of improved availability for machine operation via the new optimal maintenance plan.

The research gap for this study and future work are pointed out at the end of the paper.

Item Type: Dissertation (University of Nottingham only)
Depositing User: QIU, Huixuan
Date Deposited: 02 Dec 2022 15:56
Last Modified: 02 Dec 2022 15:56
URI: https://eprints.nottingham.ac.uk/id/eprint/58106

Actions (Archive Staff Only)

Edit View Edit View