Artificial neural network approach for condition-based maintenance

TL;DR

This paper proposes using artificial neural networks to analyze equipment condition signals for improved condition-based maintenance, aiming to optimize maintenance scheduling and reduce costs.

Contribution

It introduces an ANN-based approach to accurately estimate failure times from condition signals, enhancing maintenance decision-making.

Findings

ANN effectively captures equipment condition signals

Improved estimation of failure times

Potential for optimized maintenance scheduling

Abstract

In this research, computerized maintenance management will be investigated. The rise of maintenance cost forced the research community to look for more effective ways to schedule maintenance operations. Using computerized models to come up with optimal maintenance policy has led to better equipment utilization and lower costs. This research adopts Condition-Based Maintenance model where the maintenance decision is generated based on equipment conditions. Artificial Neural Network technique is proposed to capture and analyze equipment condition signals which lead to higher level of knowledge gathering. This knowledge is used to accurately estimate equipment failure time. Based on these estimations, an optimal maintenance management policy can be achieved.