➤ Augment Technologies said it has successfully used machine learning to improve safety and ore recovery at major gold and copper mines in Australia and Zambia.

➤ The next stage is to move downstream to ensure geological modeling benefits are realized in plants across the mining industry.

➤ AI will empower those already working on mine sites to tap into their experience and do things quicker, rather than "automate jobs out of existence."

Augment Technologies CEO Brice Gower. Source: Augment Technologies Pty. Ltd.

Australian mining technology group Augment Technologies Pty. Ltd. and Swedish digital reality firm Hexagon AB have teamed up on a solution that could boost ore recovery rates by up to 5% and create tens of millions of dollars in additional revenue per year and per mine, Augment said Feb. 7 in announcing the deal.

Augment's AI-driven Muckpile Block Model tracks material movement during blasting with high accuracy and resolution, which the company said could help miners minimize ore loss and dilution. The tool can also use remotely operated drones to measure blast movement, reducing ground risks for miners.

The Muckpile Block Model, which Augment developed while working with Canadian miner First Quantum Minerals Ltd.'s large open pit copper porphyry mines in Zambia, is now part of standard operating procedure at Australian gold major Northern Star Resources Ltd.'s West Australian open pit mines.

Brice Gower co-founded Augment in 2018 with Greg Hardwich, who was previously operations manager across Australia and South Asia for oilfield services giant Schlumberger Ltd.

S&P Global Commodity Insights spoke with Gower about the genesis of Augment's technology and the many changes AI is poised to bring to the mining sector. This interview has been edited for clarity and length.

S&P Global Commodity Insights: What did you set out to solve by turning to AI for mining?

Brice Gower: What we're solving are issues around blast movement. We took an approach that addresses a misalignment with academia and mining operations in that things need to be practical.

Our core technology is the Muckpile Block Model, which models how blasting changes the grades and locations of all the material that the geologists have estimated at a mine site. It is practical because it's open, transparent and can be used by any geological modeling program, without people needing a PhD to operate it.

Miners can never recover 100% of their resources and reserves. It's just not economic, but our technology gives them the best estimate that they've ever had of what is actually sitting in the pit, ready to be mined, and helps avoid the losses normally incurred from the chaos of the blasting process.

What does the deal with Hexagon mean for your company and for blast monitoring in the industry?

Miners used to have to place a blast vector instrument in the ground and then find it wherever they land after the blast to measure the blast movement. Then Blast Movement Technologies Pty. Ltd. (BMT), which Hexagon acquired in 2020, revolutionized blast measurement balls with internet-of-things technology by using radio frequency signals so they can be found easily and precisely. Those balls are now the industry standard.

This deal with Hexagon is a step change because we are spring-boarding off BMT's technology, and we provide such a higher level of precision dataset with the Muckpile Block Model that even Hexagon agrees that it's the future of blast monitoring. Where we're changing the game for the better is that we use drone-driven topographic surveys, which reduces the need to actually walk on that broken ground, which is a safety risk, and there have been incidents out there.

So we're combining BMT's blast movement balls as the industry standard with our software-driven, physics engine AI program.

What is the best application of this technology?

We're quite focused on gold at the moment because it's the biggest value-add, but we're also looking at copper because the volumes they move also represent a big value proposition.

Complex base metal mines suffer from a different problem called mixing, where the geochemistry starts to get very confusing and their plants are very sensitive to that. So our technology helps miners understand what no other solution was able to help them with before — ore bodies that have high selectivity or complex geochemistry.

Due to the way that blasting quite literally destroys the precision of all geological modeling that happens beforehand, very little of that geological precision could flow downstream through the rest of the mining operation and ultimately into the plant. So where we're going next is downstream to connect into fleet management systems and assist the full reconciliation workflow and even some of the processing plant actions.

How do you see the impact of AI in terms of complementing or replacing human expertise in mining?

We called the company Augment because we see AI technology increasing the capability of our experts. There's a very real job shortage at the moment of experts like geologists, and both our current software and our future projects are designed to empower the professionals that are already working for the mine and give them more capability or enable them to do things much quicker than they could before. And it's all focused on this value of "augment" — none of our solutions are meant to automate a role away. They're meant to empower our professionals and make use of their experience.

I honestly could not give you any AI product in mining that actually removes jobs. There are definitely examples where technology aggregates jobs up to one higher-level role, but that's how this technology is changing — it's augmenting people who are experienced, driven and motivated to do more, be more efficient and effective. It's quite a fallacy to think that jobs are going to evaporate.

A role for a human is one that is empathetic, instinctual or hospitable, which are key human elements that you can't take away. The job for a machine is precise, monotonous and dangerous. You shouldn't ever want a person to do a dangerous job. If a job is monotonous, no one's ever going to love it.