Oil and gas pipeline failure prediction system using long range ultrasonic transducers and Euclidean-Support Vector Machines classification approach

► This paper presents an oil and gas pipeline failure prediction system using LRUT in conjunction with Euclidean-SVM approach. ► The conventional NDT pipeline failure prediction systems are deployed at pre-determined intervals. ► LRUT pipeline failure prediction system with Euclidean-SVM provides continuous monitoring for pipelines and makes decision without human errors and misinterpretations. ► In contrast to the conventional SVM, Euclidean-SVM is less dependent on the choice of the kernel function and parameters. ► Experimental results show that Euclidean-SVM has more consistent and better classification performance as compared to the conventional SVM.