A REVIEW ON: DETECTING ANOMALIES IN OIL PIPELINES USING MACHINE LEARNING

Abstract

The oil and gas industry is very critical in the global economy. However, it is prone to a lot of operational anomalies that may result in severe environmental and economic consequences. Oil pipelines are very relevant to the transportation of crude oil and petroleum products to the processing plants and refineries. Pipelines suffer from numerous anomalies such as cracks, corrosions, and leakages, which have resulted in large-scale environmental and economic losses. Traditional detection methods for oil pipeline anomalies are normally expensive and time-consuming, hence less effective. Recently, ML techniques have appeared as a promising solution for the detection of oil pipeline flow anomalies. This paper provides a comprehensive review of methodologies and techniques used in oil pipeline anomaly detection using ML. We discuss several anomaly detection ML algorithms, their data sources, and feature extraction techniques, as well as the challenges in implementing those technologies. Key findings emphasize that constant evolution of ML applications is required to guarantee reliable safety regarding oil pipelines. The paper concludes by recommending areas of future research and possible improvements of the current methodologies.

Keywords: Machine Learning, Anomaly Detection, Oil Pipelines, Data Analytics, Predictive Maintenance

DOI: 10.7176/JETP/14-3-01

Publication date: September 30^th 2024