Predictive maintenance has proven a cost-effective maintenance management method for critical equipment in many verticals. The semi-conductor industry could also benefit. Most semiconductor fabrication plants are equipped with extensive diagnostic and quality control sensors, that could be used to monitor the condition of assets and ultimately mitigate unscheduled downtime by identifying root causes of mechanical problems early before they can develop into mechanical failures. Machine Learning is the process of building a scientific model after discovering knowledge from a data set. It is the complex computation process of automatic pattern recognition and intelligent decision making based on training sample data. Machine learning algorithm can gather facts about a situation through sensors or human input and compare this information to stored data and decide what the information signifies. We present here the results of applying machine learning to a predictive maintenance dataset to identify future vibration-related failures. The results of predicted future failures act as an aid for engineers in their decision-making process regarding asset maintenance.
|Number of pages||10|
|Journal||Journal of Computer Engineering & Information Technology,|
|Publication status||Accepted/In press - 11 Apr 2019|
- machine Learning
- Neural Network
King, R., & Curran, K. (Accepted/In press). Predictive Maintenance for Vibration-Related failures in the Semi-Conductor Industry. Journal of Computer Engineering & Information Technology, 8(1), 1. . https://doi.org/10.4172/2324-9307.1000215