The Utilization of Python Models in Real-Time Data Processing for Oil and Gas Monitoring Systems

Abstract

This research aims to utilize Python models for real-time data processing in monitoring systems for the oil and gas industry, leveraging pressure and temperature sensors. The Internet of Things (IoT) is used to collect and process sensor data directly. Using Python methods, the data is analyzed to detect patterns, anomalies, and predict potential damage. Anomaly detection using the K-Nearest Neighbors (KNN) and Isolation Forest algorithms successfully detected operational anomalies with high accuracy. KNN achieved an accuracy of 90%, while Isolation Forest produced better results with an anomaly detection accuracy of 92%. Furthermore, the equipment failure prediction model built using the Logistic Regression and Random Forest algorithms showed good predictive ability. Logistic Regression reached an accuracy of 90%, with a precision of 89% and a recall of 86%. Meanwhile, Random Forest provided better prediction results with an accuracy of 92%, a precision of 90%, and a recall of 88%. The results of this research indicate that the application of Python models for IoT-based real-time data processing can significantly improve the efficiency of monitoring systems and accelerate problem detection in the field, with excellent detection and prediction accuracy.