Piping Circuit Development using K-Prototype Clustering

Abstract

The oil and gas industry is one of Indonesia's vital industries, and it contributes the most to the country's foreign exchange. Piping is an important piece of equipment in the oil and gas production facilities; therefore, the piping inspection plan should be well prepared. An integral part of inspection plan development is a piping circuit; it allows an inspector to manage the necessary inspections, calculations, and better recordkeeping. A problem faced in piping circuit development is the need for relatively many working hours and variability results. Although this problem is often encountered, piping circuit development generated by manual work is still common in practice. To overcome the issues in the piping circuit development, therefore a k-prototype algorithm was introduced. A k-prototype algorithm was used to accommodate the shortcomings in grouping objects with features comprised of mixed categorical and numerical data. This study concludes that the k-prototype algorithm is a promising clustering technique that can reduce the time spent developing the piping circuit and eliminating the resulting variability.