Case Study Template

Case Study

Cokebusters + University of Strathclyde

(Project 2) To streamline analysis procedure and integrate autonomous data interpretation of our Ultrasonicsmart inspection 'pigs' in the field.

Mechanical decoking and in-line inspection

Cokebusters

Cokebusters was incorporated in 2005, bringing together expertise in mechanical decoking (the first in Europe) and intelligent inspection pig. The business primarily operates in the downstream oil and gas sector.

What was the need?

The Challenge

Affected by machine learning technology, the mechanical decoking, descaling, and associated in-line inspection market is becoming increasingly competitive. Different inspection technologies are being rapidly developed and introduced to the market. The project aims to help Cokebusters improve its competitiveness and productivity in this trend. Technically, to introduce new inspection capabilities to Cokebuster’s smart pigs, enabling the company to expand its services. Opportunities exist to add/enhance detection algorithms. The first challenge is to determine the best option from a commercial and technical standpoint. At this stage, machine learning techniques are employed to achieve a higher data utilisation rate, improve the probability of detection and accuracy of wall thickness measurements. The machine learning-based tool will be tested and matured using the onsite test facilities until it covers the inspection needs and is ready to be released into the field.

What did we do?

The Solution

An AI-based module has been developed as part of this KTP project to assist inspectors with pipeline interrogation. Multiple Machine Learning models will be integrated into Cokebusters’ technical platform to improve the overall inspection quality in terms of:

Enhance data processing capabilities, significantly reduce data handling time
Improve the overall Probability of Detection (PoD)

What changed?

The Impacts and Benefits

Impacts for the Company

The most significant impact will be the reduction in inspection costs and handling time associated with ultrasonic data. This impact will significantly enhance the efficiency and accuracy of pipeline inspections, enabling us to deliver high-quality, faster services to our customers.

Impacts for the Academic Team

This is a valuable practical implementation experience for the academic team at the University of Strathclyde, as the machine learning-based module will be deployed in the real world and optimised to adapt to various physical conditions – The core idea of knowledge transferring.

The project's outcomes and achievements in advancing NDT will be published in leading venues and presented at international conferences for official announcements and promotion.

Impacts for the KTP Associate

As a KTP associate, I aim to help the company integrate new product features to maintain its competitive advantages. As a result, the company will receive constant development. From my personal development perspective, I gained knowledge of product design, business models, and R&D experience. In the long term, this role will provide me with valuable insights into product innovation, help me identify market gaps, and facilitate a smooth transition from academia to industry.

The People

Meet the Team

Hongyu You

KTP Associate

Professor Gordon Dobie

Knowledge Base Supervisor

Nick Bettley

Company Supervisor