Predictive Equipment Maintenance using Machine Learning has been an aspirational goal ever since I started exploring Machine Learning and its use cases in O&G. Beginning of this year, I presented my ideas, methodologies, and results on Condition Monitoring in Hydraulic Systems at the International Conference on Developments in eSystems Engineering (DESE), 2023.
My paper titled "Prediction of Component Level Degradation in a Hydraulic Rig using Machine Learning Methods" predicts failures in a Hydraulic Rig constituting of 4 major pieces of equipment - Pumps, Valves, Coolers, and Accumulators. It was a multi-class and multi-target (5 Target Variables) classification problem. The paper presents a comparative analysis of the performance of Machine Learning models in noise-free and noisy test sets.
My domain experience in Instrumentation Maintenance further propelled me to explore failure prediction in the absence of labeled historical data, a rarity in the industry. This paper has further reinforced my opinion that Domain expertise is non-negotiable for identifying and developing use cases of Machine Learning, especially in the O&G industry.
My paper can be accessed at: https://lnkd.in/dUTRhQdk.
Please let me know your thoughts.