Unlocking critical resources through fully autonomous mass mining

When most people think about autonomous underground mining operations, it’s usually from an equipment or systems perspective i.e., an area of a mine – generally in a production setting – which operates without the physical presence of humans. Operations like these have been possible for some time and would be classed as level three ‘conditionally autonomous’ on the GMG’s Mining Automation Maturity Model. However, for an operation to be classed as ‘highly’ or ‘fully’ autonomous (level four or five as defined by the GMG) requires an operation to be run exclusive of people and their decision making.

Systems with this capability are also classed as level six ‘highly automated’ under the GMG’s maturity model for short interval control (SIC) systems (see figure 1). These technologies enable a centralized coordinator to optimize equipment and activities against functions like maintenance, mine planning, and logistics, providing improvements in accuracy, effectiveness, and speed of operation-wide activity coordination.

For the majority of mines, operations and systems like these are still a pipedream. But, for mass mining operations, such as block and panel caves, fully autonomous production is a tangible possibility today; technologies that can provide significant advantages already exist, and when the right solutions and providers are brought together, those advantages can be achieved.

Continuous mining suits full autonomy

Block caves are generally large-scale, long-life assets. The high speed, high volume, repetitive nature of production activities – for instance, loading at drawpoints and ramp haulage with trucks – means they lend themselves well to highly or fully autonomous operation, as does the relatively stable operational footprint (although this will develop and expand over time). This stability means that an autonomous production system can be set up and tuned for optimal performance as well as integration with other mine systems and functions, for instance, planning and scheduling.

In practice, a system like this might comprise a fleet of autonomous loaders and trucks which are dynamically dispatched to different drawpoints and ore passes depending upon the real-time cave state and its compliance with the mine plan. It might also include auxiliary equipment, such as water sprays which are operated autonomously when dust levels exceed a certain threshold, or breakers that intervene when oversized material is detected at a drawpoint.

In comparison, operations which use mining methods such as sub-level stoping, tend to be more dynamic and, as such, when autonomy is introduced, the focus would likely be on a smaller area of the operation with a view to quick setup and operational consistency. A truck loop is a good example: autonomous trucks shuttle between the loaders and the ore pass, on a predefined route with human supervision.

AutoMine® recently achieved a safety milestone of over seven-million lost-time-injury (LTI) free working hours globally. Osborne said this success can partly be attributed to the wraparound Lifecycle Services that Sandvik provides for it.

Automated decision-making elevates autonomy

In February 2023, Sandvik completed the acquisition of industrial mathematics specialist, Polymathian. The company is now part of Deswik, whose integrated suite of software solutions enables effective planning, scheduling and execution of mass mining operations, while unlocking higher levels of autonomy through automated and optimized decision making.

He added that this capability is exclusive to level six SIC systems, as defined by the GMG maturity model – at the time of writing, Sandvik was the only vendor whose technologies provide this.

ORB was developed in 2013 by Polymathian to help a copper-gold mine in Australia to solve a scheduling bottleneck. In that case, cave productivity was constrained by the throughput of the operation’s loaders.

Following that success, ORB was deployed at Rio Tinto’s Argyle diamond mine in Western Australia which was experiencing issues with convergence and wet muck. There, the issue wasn’t just optimizing throughput, but making sure the right material was mined at the right time to deliver a blend that kept the primary crusher up and running. The mine saw a 7% improvement in tonnage over shift-changes. Since then, ORB has also been rolled out to various locations globally including Rio Tinto’s Oyu Tolgoi copper-gold mine in Mongolia where it’s used alongside AutoMine® for fully autonomous dispatching and to manage cave draw.Since its inception, ORB has received a number of updates and its applications have been expanded to encompass sub-level caving and other mining methods, as well as cave blending capabilities. The tool can add value in three ways: by improving compliance, improving throughput and through data capture and cave monitoring; each can be performed independently or run simultaneously for greater value.

Enabling future caving operations

In time, these integrations will be expanded to encompass other industry-leading solutions and technologies from Sandvik’s portfolio. For example, Deswik.OPS, which is a collaborative short-term planning and shift execution tool for monitoring and managing compliance used by mines like Oyu Tolgoi, could be integrated with AutoMine® and ORB to dynamically manage cave execution and development.

For example, linking these systems and the OPS Operator App would provide automated capture of production data as well as machine, operator and location status information. This would allow highly effective replanning across all parts of caving operations, not just production. Further out, although equally important, mine development planning solutions, such as Deswik.Spatial, Deswik.Planning and Deswik.Caving, could also be added.

Another common occurrence in caving operations are hang ups – oversized materials blocking a drive or drawpoint which need to be broken up or blasted before production can continue. By integrating Newtrax technologies with AutoMine® and ORB, the closest breaker could be quickly identified and autonomously dispatched to resolve the issue.