Hancock Iron Ore scales Azure AI to extend rail lifespan 10%

Hancock Iron Ore needed to improve reliability across mining operations where rail disruptions, inspection volume, and delayed defect insights could affect output, cost, safety, and maintenance planning.

Hancock Iron Ore built a repeatable AI-assisted operating model using Azure AI, Microsoft Foundry, Microsoft Fabric, Azure Databricks, and Microsoft 365 Copilot.

Enterprise assistant RoyBot handles over 70,000 messages a month for 2,000-plus users, while TrackDefectX saves $475,000 a year in rail grinding costs and extends rail lifespan more than 10%, providing evidence of AI at scale across the business.

For mining organizations, reliability is a business imperative, and increasingly, so is the ability to turn operational data into trusted, scalable intelligence. For these companies, rail infrastructure serves as a critical link between production and shipment. Every disruption affects output, costs, and operations, and with the amount of operational and inspection data generated by mining companies today, it can be difficult to access the right information in time to prevent failure.

“For Hancock Iron Ore, AI is unlocking practical and tangible benefits, providing holistic data and insights that is solving real operational challenges.”

Kate Flanagan, Chief Technology Officer, Hancock Iron Ore

Turning operational data into AI-assisted decisions

Hancock Iron Ore, an Australian mining company, implemented an AI-assisted approach to help teams act earlier on maintenance, inspection, and operational decisions. Microsoft Foundry, Microsoft Fabric, Azure Databricks, and Microsoft 365 Copilot all support AI application development, unified data and analytics, agentic solutions, and insight delivery.

That Microsoft platform now powers its own AI apps and agents: DeviationManagerX and CrusherIQ in supply chain, SkyInspectX in asset inspection, TrackDefectX in monitoring railways. Its AI assistant, RoyBot, drives the platform. “For Hancock Iron Ore, AI is unlocking practical and tangible benefits, providing holistic data and insights that is solving real operational challenges. Our executive chairman, Mrs Gina Rinehart, has long championed a culture of improvement across her businesses to achieve safer, more efficient operations, and this has put us at the forefront of technological innovation within the mining industry,” says Kate Flanagan, Chief Technology Officer, Hancock Iron Ore.

Insights now surface within workflows, instead of from across sources, helping maintainers identify the most meaningful defects and respond faster. This approach helps Hancock Iron Ore operationalize AI, moving from analysis to real-world decisions. Teams deploy AI while maintaining governance, visibility, and operational control. ”The challenge was safely getting insight into the tools our teams already use. Microsoft helped us move from experiment to production, enabling experts from across the business to work alongside the data innovation team, questioning what was possible then making it a reality,” explains Chad Burrows, Head of Digital Innovation, Hancock Iron Ore.

The anchor: an enterprise AI assistant

RoyBot, Hancock Iron Ore’s enterprise AI assistant, exemplifies frontier-firm scale. Grounded across the company’s knowledge bases and systems, it anchors content generation and research throughout the business. RoyBot handles 70,000-plus messages a month for 2,000-plus users and earned an ABA100 award for AI innovation.

Delivering measurable business impact

TrackDefectX turns this approach into hard numbers. Its computer vision detects and classifies defects across hundreds of kilometers of railway track, shifting maintenance from reactive to proactive. This has extended rail lifespan by more than 10% and generated $475,000 in annual rail grinding cost savings, while reducing risk and prioritizing safety issues. 1,900-plus staff have now trained through hands-on AI labs.

“Manually inspecting around 350,000 track images a month was incredibly time consuming to locate defects. Having reliable and trusted outputs from TrackDefectX means we can focus on the data that matters and get ahead of problems instead of reacting to them,” says Munish Kumar, Specialist Track Engineer, Hancock Iron Ore.

Creating a foundation for industrial AI at scale

By combining operational expertise with Microsoft’s AI and data platform, Hancock Iron Ore is applying AI to reliability, maintenance, and emerging needs while keeping governance and control at the center.

Hancock Iron Ore’s breakthrough is a repeatable model for scaling AI from the rail corridor to the back office. As adoption expands, Hancock Iron Ore is building the organizational muscle required to scale AI responsibly, across training and workflows. For leaders focused on reliability, efficiency, and growth, the lesson is clear: organizations create measurable value when intelligence becomes part of everyday operations rather than a separate analytical exercise.

Discover more about Hancock Iron Ore on LinkedIn and YouTube.