Machine Learning Workflow to Predict Remaining Useful Life (RUL) of Equipment

• Main Author: Mohd Fauzi, Muhammad Farhan Asyraf
• Format: Final Year Project
• Language: English
• Published: IRC 2019
• Subjects: Q Science (General)
• Online Access: http://utpedia.utp.edu.my/20925/1/Muhammad%20Farhan%20Asyraf%20Mohd%20Fauzi_22963.pdf; http://utpedia.utp.edu.my/20925/

Today, modern industrial equipment is very complex as it involves sophisticated assets and systems. Thus, machine equipment optimization and safety have become operators' main concerns in the quest for maintaining optimum operational efficiency, asset availability, safety and cost-effective. Due to its complexity of the internal structure of the equipment, engineers are often faced with large amounts of information called multivariate datasets which are hard to understand by human nature. This led to difficulty in achieving high accuracy prediction of the equipment and decision-making is hard to achieve. Thus, an organization unable to decide whether to purchase new equipment or provide maintenance strategies. Hence, the purpose of this research is to develop a machine learning workflow model of the integration between Alteryx tools to do prediction of RUL using “real world” multivariate dataset in Oil and Gas industry, and Microsoft Power BI to visualize the result of the prediction for a better insight. One of the most popular machine learning approaches is employed in the prediction workflow which is the Artificial Neural Network (ANN) algorithm, due to its capability to learn from a large volume of data points and high prediction accuracy. Performances of the accuracy of prediction workflow were measured using root mean squared error (RMSE).