Hancock Iron Ore Pty Ltd (formerly Roy Hill) is one of Australia’s leading iron ore producers, operating an integrated mine, rail, and port network in the Pilbara region of Western Australia. The Digital Innovation team partners closely with Operations to embed advanced technologies, particularly AI, into the heart of the business, delivering measurable value across productivity, sustainability, and safety.

TrackDefectX is an AI-powered track defect detection system that is revolutionising the monitoring and maintenance of rail infrastructure. Designed entirely in-house, the solution automates the detection and classification of critical rail defects, including Rolling Contact Fatigue (RCF), spalling, squats, gauge corner checking, and wheel burns, at a level of speed and accuracy beyond human capability.

Using high-resolution imagery captured by the business’s Rail Surface Imaging System (RSIS), TrackDefectX applies advanced deep learning models to detect defects and immediately deliver actionable insights to maintenance teams. The result: safer rail operations, lower costs, extended asset life, and fewer unplanned disruptions.

TrackDefectX’s core is advanced AI and deep learning for rail track defect detection, with unique features such as context-aware segmentation, multi-task learning, and incremental learning. What sets TrackDefectX apart is its originality and technical sophistication: a world-first application of advanced AI techniques.

The system has delivered $475,000 in annual savings to date, and by identifying defects early, TrackDefectX has increased rail lifespan by more than 10%, significantly lowering replacement frequency and capital expenditure. TrackDefectX is a novel combination of advanced AI techniques that delivers unmatched accuracy and scalability.

For more information about Hancock Iron Ore, visit hancockironore.com.au.