As the world continues its journey through the energy transition, the need for more efficient tools to track down sources of key metals and minerals only increases. Artificial intelligence offers huge potential for mineral exploration, not only providing increased speed and efficiency but also offering a minimal environmental impact compared with traditional methods. Nicholas Kenny speaks to Yair Frastai, co-founder and CEO at VerAI Discoveries, to learn more about the benefits that AI can provide when it comes to critical mineral discovery.
How AI can be leveraged to expedite discovery of critical minerals. (Credit: Dominik Vanyi on Unsplash)
The path to net zero is thorny one, with many challenges along the way. Despite comments from the president of Cop28, Sultan Ahmed Al Jaber, during an online event on 21 November, where he declared there was “no science out there, or no scenario out there, that says that the phase-out of fossil fuel is what’s going to achieve 1.5°C”, we’ve long known that humanity will need to greatly reduce its use of fossil fuels in the coming years.
Of course, Al Jaber would quickly backtrack on his claims, putting the ensuing global outrage down to “misrepresentation and misinterpretation”, but the damage had already been done. In some ways, however, the conversation around Cop28 was energised by the controversy, with global leaders such as UN secretary general António Guterres firmly coming out with calls for a full phase out of fossil fuels, along with a concrete plan for doing so.
And as we phase out fossil fuels, our need for the metals and minerals key to the energy transition will only increase. Current supply gaps for lithium, nickel, graphite, cobalt, neodymium and copper could all imperil the world’s progress to net zero by 2050, according to the Energy Transitions Commission in a July 2023 report. Annual investment in these key metals averaged at just over $45bn a year for the past two decades, compared with the $70bn required annually through to 2030 to expand supply.
Beyond increasing investment, however, it’s clear that the mining industry will also need to step up its game to expand its supply of these metals – for example, it still has to find sufficient deposits of them. The mining industry spends a sizable fortune on exploration – in 2021, Canada alone spent over $800m to hunt down new sources of metals and minerals, while Australia and the US forked out $531.3m and $345.2m, respectively. In 2022, the global total came to $4.055bn. It goes without saying, then, that mineral exploration is big business.
Traditional forms of exploration have largely already discovered any ‘easy’ sources of critical minerals, leaving the vast majority of remaining ore deposits concealed under less-explored terrain. New developments in this area, however, seek to improve and speed up modern mineral exploration by introducing AI to the mix.
Adapt to a new paradigm
It would be far from the first time that AI has been tasked with discovering the solution to a problem. Most notably, AI is already being used in healthcare services across the world to analyse MRI images and read brain scans, capable of reading these images and identifying patterns faster than a human can, supporting clinicians and radiologists in making assessments. This can speed up the time it takes for patients to be diagnosed and treated, allowing healthcare services to screen greater numbers of people more quickly, while providing them with the help they need.
“The benefits of AI are not unique to the challenge of discovering mineral deposits,” says Yair Frastai, co-founder and CEO of VerAI Discoveries, a leading AI-based mineral asset generator based in Boston, US, which has developed AI solutions that aim to improve the probability of discovering economic mineral deposits. “There are several industries that depend on their ability to make accurate, faster and cheaper discoveries – at vast scale. Sectors such as medicine, pharma, insurance, finance and homeland security intelligence that, until recently, have been based only on expert knowledge, which is great when it’s working.” What happens, however, when the discovery challenge becomes significantly harder – whether that’s due to increased scarcity or to time constraints, or the lack of an expert hypothesis to find it?
Here, Frastai uses the Covid-19 pandemic as an example, seeing it as a more tangible metaphor for a general audience. With Covid, the need to discover a cure as quickly as possible was of paramount importance, leading to pharmaceutical companies and scientists embracing new technology and eventually driving the development of revolutionary mRNA vaccines over the finish line. “There is no way that a solution can be found so quickly if you’re not deploying new types of technology that bring a different paradigm to the game,” he adds.
To date, VerAI has focused its operations on countries with well-established and regulated mining operations in North and South America, having successfully used its AI solutions to discover ore deposits containing lithium, nickel, cobalt, copper, gold and silver in Ontario, Arizona, Nevada, Mexico, Chile and Peru. Focusing on these places, however, can present a challenge, as many in the industry believe that all the mining opportunities for sizable new discoveries have already dried up. But this is far from the truth, Frastai claims.
“The only thing that has really been exhausted is the low-hanging fruit – everything that was outcropping has already been well developed,” he notes. However, countries like Australia, or US and Canadian states like Arizona, Nevada and Ontario, possess huge areas covered by a veneer of regolith sediments or younger rock, with any valuable ore buried beneath rather than outcropping at the surface.
Currently, most of the existing mineral resources that have been discovered and extracted have come from outcropping or near outcropping geology, according to Geoscience Australia, the Australian government’s agency for geoscientific research. This kind of geology covers some 20% of Australia’s total land area, with the remaining 80% covered by a few hundred metres of the aforementioned regolith, creating huge opportunity for mineral exploration.
“The traditional industry today is based on prospecting and expertise that goes back centuries,” says Frastai. “But it can no longer find manifestations [of valuable ore] on the surface of the covered areas, so you need a different way to approach the problem.”
This is where AI comes in. Rather than searching on the physical surface, AI instead can help sort through huge swathes of information from the subsurface and identify relevant data within it. VerAI’s solution, says Frastai, is tailored to deal with this exact challenge, setting it apart from how others in the sector approach the problem when approaching massive amounts of data.
“We are highly focused on the geophysical data,” he explains. “And this allows us to eventually see patterns that [traditional exploration] experts cannot detect using simple and limited hypothesis.” This, of course, can pose a challenge when pitching AI to a traditional industry like mining, which can be slow to embrace new technologies and ways of doing things – or, as Frastai puts it, “release itself from hypotheses and practices that belong to the 19th century”.
VerAI’s CEO, however, remains optimistic, seeing both the challenge for mineral discovery and exploration and the opportunity that AI offers as unique. “If you think about medicine and the complexity of the human body, AI is dealing with discovery targets that are changing and evolving all the time,” he says. “In our case, the ore bodies we are going after formed millions of years ago, and it’s not moving, and it’s not changing – at least in our lifetime.”
This perspective is shaped by experience, coming after more than 25 years working in intelligence and homeland security. Over the past 15 years, he witnessed a huge cultural shift as machines and algorithms gradually became more and more integrated throughout intelligence, replacing the expert knowledge in many areas, including decisionmaking. Similarly, he makes comparisons between mining and the fintech sector. “We called it ‘finance’ before it was called ‘fintech’,” he jokes. “There is no reason why mining will not be called ‘minetech’ as well. This is already happening.”
Delve into the data
VerAI has trained its AI systems to search for minerals like lithium, cobalt, nickel, copper, zinc, gold, silver and molybdenum, and Frastai highlights copper to serve as the case in point when describing how these systems function. “Copper is well-known and there are several types of mineral copper out there that people are trying to target,” he says, by way of explanation. “One of the very large mineralisation types that people would very much like to find are called ‘porphyry copper deposits’ (PCDs).”
What VerAI’s systems found notable about PCDs it that examples with the exact same mineralisation type can have very different signatures depending on its geological area. “It’s not because the mineral system is so different, [it’s] that the relationship with the hosting geological setting impacts the pattern that the algorithm is eventually able to find,” Frastai notes.
VerAI is well-versed in how to adapt to these challenges, Frastai stresses, but he brings these issues up to highlight that the solution is not as simple as merely inputting data and receiving the results you’re looking for. Geological knowledge is vital to set the right conditions for a successful search. “You can’t approach this problem without relevant geoscience knowledge that allow you to put things in the right context and to define your discovery problem upfront.” Essentially, VerAI develops models or profiles based on existing economic deposits, and then uses that large and diverse library of profiles to search through datasets to identify locations that have the same pattern that could have been missed by traditional exploration methods. Of course, human expertise is always required to provide oversight for these findings, but this technology provides a great opportunity when conducted properly.
An opportunity to inform and collaborate with communities
While identifying unknown deposits of critical minerals for the energy transition is key to VerAI’s work, it’s also important to the company that as humanity mines more and more of these materials that we bear in mind the cost to the environment and to communities. “The way to do it is to do responsible exploration and responsible mining in our backyard,” Frastai stresses.
Historically, more than 99% of exploration projects fail to become mines, according to KoBold Metals, another leading player in AI mineral exploration. This rate of failing to put the drilling machine on the right spot, from the start, creates a serious environmental impact, as a high-footprint field operation. This is where VerAI’s technology and its approach to exploration has a huge advantage, because it works remotely, searching in the data space where its physical footprint is non-existent. It can conduct its exploration and discovery processes without needing to interfere with the environment and local communities, avoiding any unnecessary tension before the minerals have even been located.
Going even further, it also offers the opportunity for local communities to lead any mineral development projects on their land, by engaging with VerAI. “We can highlight the specific and highpotential locations for different minerals in their land, without creating any damage and allowing them to drive the process instead of being reactive to exploration results conducted by others who not always align with the community’s best interest.”
The communities, Frastai continues, aren’t necessarily opposed to mining operations in their area. “They just want to do it right. And they want to make sure that their culture and cultural assets will remain protected,” he says. “If those communities have the information they need upfront, instead of being constantly alert and the last one informed when something is happening, I think we’ll find that there’s no real conflict here.”
In 2024, he expects VerAI to continue its efforts to engage with those communities that would like to take the lead exploring their own in areas, and he believes this will be key to sustainable exploration. As much as the world needs critical minerals for its net zero future, it doesn’t make much sense if you have to despoil the land and create conflict with its local owners in the process. “People often talk about ESG in the exploration industry, but it’s very hard to transform this into something tangible,” Frastai notes. “But VerAI’s technology and our approach to this, I think, creates great opportunity.”
A fresh set of AIs
It should be said, too, that VerAI are far from newcomers to the market, having been in the mineral exploration business for over ten years. At the same time, this still grants the company a fresh set of eyes – or AIs – according to Frastai. “VerAI’s founders have a very strong background coming from intelligence, the art of discovery […] We are bringing a unique approach and methodology into mineral exploration, but it’s not a walk in the park either – creating a systemic and systematic platform that dramatically improves the probability of discovery is a huge challenge.”
However, in some ways the mining industry is still operating within its comfort zone. There are many demands around the need for critical minerals, but the supply gaps mentioned earlier have only had an impact further down the chain. “You need to ask yourself why Tesla is buying land in Nevada for lithium,” Frastai adds. “Downstream, they are dealing with manufacturing batteries and electric cars, and then they go all the way upstream to buy land? It’s not because they’re great miners or they know exploration – it’s because they are frustrated that they are not able to secure the raw materials they need.”
VerAI doesn’t intend to set itself up as a miner, either, preferring to keep its focus on the problems at hand. “The advantage of our technology is that it tells us where to put the drills, but not how to drill,” summarises Frastai. The company instead intends to stay lean, serving as a project generator that will have a stake in the assets it generates, but it is not a miner, a developer or an explorer in and of itself. “By doing this, it’s allowed us as a business to scale and grow very rapidly, focusing on our competitive advantage of targeting better, faster and cheaper than the traditional industry.”
Mineral exploration was long overdue an attempt at disruption, and while VerAI and many of the other players in the AI exploration space are smaller fry, big fish in the mining industry such as Rio Tinto have also begin investing in this technology. And with the demands for critical minerals only set to grow in the coming years and decades, it’s not hard to see why – if all the easily locatable deposits have already been identified, it’s inevitable that AI and other new technologies will need to be brought in to track down new sources. The human mind, after all, is not well suited for finding a needle in a haystack – AI, however, can pinpoint it just fine.
This article first appeared in World Mining Frontiers magazine.