Hitachi supports research into predictive maintenance for mining machines

HCME is collaborating with a PhD student on a research project to predict the remaining life of critical components on mining machines. The initiative will help engineers plan maintenance before parts need replacing to improve the availability, reliability and safety of dump trucks that operate in some of the world’s toughest environments. It will also help to significantly reduce operational downtime and life-cycle costs.

Providing complex data

HCME’s digital solutions team for mining operations has agreed to supply the complex data to Malihe Goli, a PhD candidate at the Geo-Resources Section of the Department of Geoscience and Engineering at TU Delft. This data will help her to build a robust model that will capture degradation trends in components such as pumps, cylinders and brakes. The project is jointly supervised by the Geo-Resources Section and the Intelligent Sustainable Prognostics Group within the Faculty of Aerospace Engineering at TU Delft.

“Our mining machines have sensors installed on key components, allowing us to gather detailed information on indicators like temperatures and pressures,” says Daan van Berkel, Manager Mining Projects and Sustainable Mining for HCME.

Predictive maintenance strategies

Condition monitoring data from HCME will help Malihe, who is also a Control and Automation Engineer, improve her model and provide more accurate estimates of when a component might fail. This information can then be used to inform predictive maintenance strategies.

“We will be able to plan when a truck needs to come into the workshop with more precision, and order any parts that may be required ahead of time,” explains Daan. “Moreover, addressing potential problems before they occur reduces the risk of a major issue that could also damage other parts and put a machine out of action for weeks.”

HCME started supplying the data in January 2025 and Malihe’s PhD thesis – “Predictive Maintenance for Mining Trucks” – is scheduled to run for another two years. “As well as providing useful information, we are also very happy to share our domain expertise with Malihe and her colleagues at TU Delft,” adds Daan.

According to Malihe, the support from HCME has been fundamental to the progress of this research. “Access to large-scale, real-world datasets – including detailed failure records, maintenance logs, and sensor measurements – has enabled the development of accurate, data-driven models for component degradation.

“I sincerely appreciate their ongoing collaboration and the valuable technical insights they share into component behaviour, which have been instrumental in guiding model development and interpretation.”