In a world hungry for battery metals, bots are mining more than data

Artificial intelligence has made a splash in recent months by writing poems, holding conversations and creating original artwork, and the technology is also taking on a growing part of the mining industry.

AI, using computers to receive, synthesize and make inferences from data, could substantially cut the number of years it takes to develop a project. It could also help solve the mining sector's challenge of rising demand for energy transition metals alongside a dwindling reserve base across several commodities.

Coming up against those issues and recent cost inflation, the mining sector has been increasing its use of AI as it looks to increase production and improve efficiency. The number of mining companies deploying AI rose to 66% in 2022 from 57% in 2021, according to a survey by heavy industry technology services provider Axora for its 2022-2023 innovation forecast.

"[AI] is the key to the acceleration of the development of the battery economy. You still need these metals developed," James Cormier-Chisholm, president of Eureka Maps Inc., said in an interview. "To get a mine up for base metals usually takes 15 years. What I'm proposing with this type of methodology is to accelerate it to, probably, about three years."

Mining companies budgeted $13 billion for exploration in 2022, up 16% from the year prior but far below exploration budgets from the 2010s, according to an analysis by S&P Global Commodity Insights. Broader application of AI in the sector could make those dollars go further.

AI could deliver efficiency and productivity improvements by automating tasks such as mapping, surveying and analyzing geological data, Brian Sathianathan, co-founder of AI solutions provider Iterate.ai, said in an interview. Miners could also use AI processes to optimize production processes, detect operational hazards, sample materials, analyze accidents, perform predictive maintenance and more.

"In recent years, the cost of AI technology has been decreasing, which has made it more accessible to organizations in the mining industry," Sathianathan said. "Additionally, the development of new AI applications and algorithms specifically tailored to the needs of the mining industry has made it easier for organizations to incorporate AI into their operations."

In 2019, the market for AI services in the mining industry was $76 million, but it is expected to swell at a compound annual growth rate of more than 23% through 2024, according to an early 2021 report from analytics firm GlobalData PLC. Freeport-McMoRan Inc., for example, has deployed artificial intelligence processes to better understand its ore bodies and to increase production from its operations, including at its Bagdad copper mine in Arizona.

"AI remains at an early deployment stage but had the largest year-on-year increase [of technologies currently being applied by organizations surveyed]," the Axora report noted. "It also maintained its position as the technology representing the biggest growth opportunity, with a notable year-on-year improvement in outlook in the next three to five years."

In a 2017 report, consultant McKinsey estimated that by 2035, smart mining, including autonomous mining and data analysis and digital technologies such as AI, would save between $290 billion and $390 billion annually for producers of copper, iron ore, natural gas, coal and crude oil.

Mapping new resources

Nova Scotia-based Eureka Maps has used an AI data mining algorithm to identify battery metals such as copper and nickel, claiming a 95% accuracy rate. Currently, the odds of a conventional mineral exploration program leading to an operating mine are about 1 in 1,000, according to the Prospectors & Developers Association of Canada.

The company combined that process with a list of endangered species identified by the International Union for Conservation of Nature to simultaneously find new metal deposits and potentially threatened species, thus better informing miners where it makes the most economical or legal sense to mine.

More than two decades ago, while working as the main reviewer for the Federal Government of Canada on the Voisey's Bay nickel mining project in Canada, now owned by Vale SA, Cormier-Chisholm found the team was overwhelmed by geological data as they made plans for the operation. So, the executive looked for tools to better comprehend the information.

"I had these two standing problems: too much data and that geology tends to be more of a theory-based exercise," Cormier-Chisholm said. "I was thinking at the time, why not just use an algorithm to get a clearer picture of what's happening there?"

Industries such as the insurance and banking sectors have been deploying data mining for decades to increase profits, Cormier-Chisholm noted, and those that dig for metals and other materials from the earth have a similar opportunity.

"I'm actually using nothing different from what they're applying. I'm using exactly the same algorithms the banking industry applies to, say, credit cards to identify a good credit risk or a bad credit risk," Cormier-Chisholm said. "Everything's the same, but it's being applied to an industry where people are still operating on a post-World War II model of how they explore."

The company provides its services to both mining companies and funds focused on environmental, social and governance issues, which want to know as much as possible about the environmental risks posed by potential investments.

In August, the Defense Advanced Research Projects Agency teamed up with the U.S. Geological Survey to offer prizes for those with ideas on using machine learning and AI tools to speed up its critical mineral assessments.

"[The Defense Advanced Research Projects Agency] is known for fostering innovation by creating problem-focused research communities, and we're excited to put the results of this competition to immediate use," Sarah Ryker, associate director for energy and mineral resources at the USGS, said in an August statement. "The United States is under-mapped, and the Bipartisan Infrastructure Law provides a historic opportunity to catch up — if we can precisely target our investments in new mapping."

Challenges to deployment

The mining industry faces obstacles to implementing AI, Sathianathan said. That includes a lack of skilled personnel to operate and maintain AI systems and a perception among the workforce that automating tasks in the mines introduces a threat to job security. Trouble hiring workers has been an ongoing issue since most of the jobs are in far-flung, rural locations, and the mining sector has a historical reputation of contributing to climate change and otherwise harming the environment.

"The adoption of AI technology often requires the retraining of existing workers in order to prepare them for working with AI systems," Sathianathan said. "This can be a time-consuming and expensive process, and it may not be feasible for all companies in the mining industry."

Other concerns include AI technology making important decisions involving the environment or public safety and potential regulatory hurdles to deploying the technology. Overcoming such challenges could be a game-changer for an industry expected to experience significant growth in demand, largely thanks to a transition toward cleaner energy resources.

S&P Global Commodity Insights produces content for distribution on S&P Capital IQ Pro.