JV Article: Micromine’s exploration AI boosts human input without replacing it 

Artificial Intelligence (AI) has dominated news headlines this year as companies harnessing its power disrupt markets.   

But what does AI mean for the mining industry – should geologists prepare to be replaced by new sets of technologies that perform tasks that have historically been done by humans? Or, conversely, what exactly can AI do to help industry professionals do their jobs more efficiently?  

These were the points presented for discussion and debate at a panel hosted by mining technology company Micromine at the AME Roundup conference in Vancouver in January. Anthony Vaccaro, president of The Northern Miner Group, moderated the session.  

Panelists Warren Black, senior geologist and geostatistician at APEX Geoscience and B2Gold (TSX: BTO; NYSE-A: BTG) evaluation geologist Martin Nunez, assessed Micromine’s Origin Grade Copilot. The AI tool uses deep learning neural networks to identify patterns, guide exploratory data analysis and validate geological interpretations faster than traditional methods. 

“You should be engaging with curiosity rather than fear,” B2Gold’s Nunez said. “It’s not coming to get you – it’s here to help you – it’s a tool in the toolbox.” 

Micromine Origin Grade Copilot helps speed up data verification, shortening the time to reach a conclusion in resource evaluations when, especially in the mining industry, time is money, he said. The $4.78-billion market cap B2Gold operates producing assets in Mali, Namibia, and the Philippines and is developing the Back River Project in Nunavut, Canada, with first gold expected in the second quarter of 2025. 

3D visualizations 

Apex’s Black said using Micromine Origin Grade Copilot brings more certainty in prospectivity modelling, and for resource estimation helps find recovery densities. 

“There’s a lot of really interesting ways to use machine learning and AI to guide that process and in decision making,” Black said. “AI takes a lot of high-quality data to generate a tool that creates predictions. Humans are better at finding novel patterns - if done together and used together it is powerful.”  

Panelist Manel Molina, Micromine’s exploration technology adviser, said the dominant skepticism in mining about AI a few years ago has subsided as the market gains more understanding. 

Molina said AI enhances geologists’ and engineers' skills, guiding and informing decision making without replacing human expertise.  

“You can build expertise on AI, and it allows technology companies to improve their products, so it's a win-win.”  

Next generation 

The skepticism keeps the industry, which is under a lot of pressure to deliver, competitive while geologists and geoscientists should instead see AI as next generation technology for mining, Nunez said. 

The mining industry can reimagine AI as a complement rather than a replacement for geologists’ and engineers’ skill sets, built to inform decision making without replacing human expertise, the panelists said. Nunez uses AI to cross check data provided, which gives an objective opinion without bias. 

“In resource estimation the question is – is there mineral continuity here? That’s something that AI can do for you, not based on any bias,” Nunez said. “Every AI tool has its own place. Being able to verify how many ounces are in the ground goes a lot further to speed up the process and ultimately helps confirm the resource is in the ground.” 

Nunez also said confirming the resource earlier means fewer holes are drilled, leaving less environmental impact and shortening timelines across the board from community engagement to government permitting to access the critical minerals needed to drive technology faster.  

The B2Gold evaluation geologist reminisced about working in exploration camps out of canvas tents with no internet or cell phone access, as part of a small team or even alone, spending painstaking hours poring over data manually.   

“It used to take days or weeks to do a lithology model, now it takes one and a half hours,” he said.     

In exploration, precision and efficiency are paramount. Companies need to reach a high level of confidence before starting activities. Juniors blowing budgets putting drills in the ground based on what they think they know is there – and turning out to be mistaken – can devastate finances.   

Increased knowledge 

In university, Warren Black’s master’s thesis work focused on using machine learning to create a data set for prospectivity modelling.  

Black also clarified three categories of tools. There are predictive tools, such as Micromine Origin Grade Copilot, and tools to help with dimension reductions, such as principal component analysis “that declutter 3D views and create new variables, making it easier to interpret a data set.” 

And there is clustering, which pulls apart a data set into populations that represent different geochemical processes. Choosing a tool depends on the objective, Black said.“You ask the tool the question – it’s not driving the narrative,” he said.   

Mastery and expertise   

Molina acknowledges adoption challenges, but emphasizes AI enables better understanding, helps identify where to drill next, and where and where not to allocate resources. He also said the tools deepen understanding of the geological history of a deposit through ‘geo modelling’ and validate confidence in estimates through resource modelling.  

“Micromine Origin Grade Copilot has a security of data,” Molina said. “It's a tool that allows geologists to build block models faster. It speeds up decision making by validating the data they produced. A geologist that is an expert in AI, and builds mastery and expertise, is going to be twice or three times more productive.” 

Nunez emphasized that it is the geologist who drives the proverbial machine but pointed out that markets are not patient about ”doing things the old-fashioned way.”  

“AI gives geologists another tool. It allows you to see if you’ve missed something, and if you are using the best tools, you can improve your chances of success significantly,” he said.  

“It's precisely because the stakes are so high that you should be using AI.”  

The preceding Joint Venture Article is PROMOTED CONTENT sponsored by Micromine and produced in co-operation with The Northern Miner. Visit: https://experience.micromine.com/grade-copilot for more information or to request a demonstration.