[Reading Science] AI, the 'Golden Hand' in Mineral Exploration and Alternative Material Discovery

AI Accelerates Exploration Period and Research Speed Significantly
Active Research by MS and Korea Institute of Geoscience and Mineral Resources

Securing minerals, especially rare minerals and rare earth elements, is essential for national competitiveness. Amid the US-China conflict, the fact that resources can become weapons is a key reason for exploring undiscovered mineral resources and securing alternative materials.

Alternative to Lithium Discovered Through AI and Quantum Computers. Photo by Microsoft

In the past, oil was used as a resource weapon, but recently, minerals needed for manufacturing advanced semiconductors and batteries have taken that place. Certain rare earth elements are expensive and their mining inevitably causes environmental damage. Finding new deposits also requires long periods of time and effort. In this situation, scientific advancements offer new alternatives.

Lithium, whose price has recently dropped significantly, is a core mineral in the battery era. With the battery industry being nationally important, securing a stable supply of lithium is essential. In the past, the focus was on purchasing lithium, but going forward, it may be necessary to find materials that can replace lithium or to discover lithium mines independently.

Recently, Microsoft (MS) found a material that can reduce the amount of lithium used in batteries by using quantum computers and artificial intelligence (AI), which is considered highly significant. MS discovered a way to make batteries using 70% less lithium than before in just a few weeks. Scientists at the Department of Energy’s Pacific Northwest National Laboratory (PNNL), who collaborated with MS on the research, were also surprised by the unexpectedly rapid results. Brian Abramson, PNNL’s Chief Digital Officer, welcomed the development, saying, "Technological advancements have opened opportunities to accelerate scientific discoveries." This achievement utilized the quantum computer cloud service ‘Azure Quantum Elements,’ which focuses on discovering new materials. Azure Quantum Elements hit a ‘home run’ just seven months after its launch.

When Satya Nadella, MS CEO, shared this achievement on the social networking service ‘X’ (formerly Twitter), Elon Musk, Tesla CEO, responded with a comment saying, "Interesting." Raphael Gomez Bombarelli, a professor at MIT, welcomed the news, saying, "Using AI allows physicists to quickly process calculations that used to take decades."

In Korea, efforts to explore minerals using AI are active. The Korea Institute of Geoscience and Mineral Resources (KIGAM, Director Lee Pyeong-gu) has recognized the importance of AI in future mineral exploration and has been strengthening research. KIGAM has worked with the Electronics and Telecommunications Research Institute (ETRI) to utilize AI, and results are beginning to appear.

Lee Pyung-gu, Director of the Korea Institute of Geoscience and Mineral Resources, is explaining lithium exploration using AI on the 11th. Photo by Baek Jong-min

Director Lee explained, "We focused on researching technology to explore key minerals using AI last year and achieved results." He introduced a case where even KIGAM was surprised by AI’s performance. He said, "The lithium mine at the Salton Sea in the US, which became a hot topic late last year, was also detected by AI developed by our researchers." The lithium reserves in this lake are estimated at 18 million tons, valued at about 708 trillion won. This scale ranks second in the world. The researchers were also surprised when the US Department of Energy officially announced the lithium deposit potential at the Salton Sea while they were still assessing it.

Schematic diagram of the application of the AI-based critical mineral prediction platform developed by the Korea Institute of Geoscience and Mineral Resources. The Institute successfully selected 11 abandoned mines with high lithium deposit potential from a mineralogical perspective by training AI with smart mining technology using geochemical data containing sample information from 25,600 river sediment points across South Korea, along with geological maps, fault and dike maps.

KIGAM has applied AI domestically and has identified 11 candidate sites for lithium mines. After additional verification, the institute plans to announce the results. The goal of the institute is for Korea to become a lithium powerhouse.

Director Lee said, "We are conducting machine learning on algorithms that pinpoint mines using AI. By combining secured data and existing research results through big data and AI, we expect to identify lithium mines that must be checked by the end of this year." He also expressed his ambition, saying, "We will strive to become the world’s best in technology for finding mines through AI."

The competitor Director Lee pointed out is the AI-based mining exploration company ‘Kobold Metals.’ This company is supported by Amazon founder Jeff Bezos and MS founder Bill Gates, among others.