OIL THEFT DETECTION AND CONTROL SYSTEM USING UNSUPERVISED MACHINE LEARNING ALGORITHM: A CASE STUDY OF COASTAL AREAS IN NIGER DELTA

This project addresses the critical issue of oil theft in the coastal areas of Nigeria's Niger Delta region through the development of an innovative detection and control system using unsupervised machine learning algorithms. Oil theft has resulted in significant economic losses, environmental degradation, and security challenges, with traditional detection methods proving inadequate. The study employs a comprehensive methodology involving data collection from multiple sources, including pipeline flow data, sensor readings, and historical records. The collected data undergoes rigorous preprocessing before being analyzed using a combination of clustering (DBSCAN) and anomaly detection (Isolation Forest) algorithms. The implemented system demonstrates superior performance in detecting potential oil theft activities compared to traditional methods, with improved accuracy and reduced false positives. Integration with existing infrastructure enhances its practicality in real-world scenarios. The project's findings underscore the potential of unsupervised machine learning in addressing complex industrial challenges, offering a more proactive and accurate approach to oil theft prevention. This research contributes to the sustainable development of the Niger Delta region by potentially reducing economic losses, minimizing environmental damage, and enhancing security measures in the oil and gas sector. Future work includes integration with advanced technologies, adaptation to evolving threats, and exploration of scalability to other regions. Despite limitations such as data availability challenges and environmental factors, this project provides a solid foundation for ongoing development of advanced oil theft detection systems, contributing both to academic understanding and practical efforts in combating oil theft.