Chinese researchers turn to AI to locate rare earth deposits in the Himalayas

The machine has now reached an accuracy of 96 percent but the location of the reserves might make mining difficult.
Ameya Paleja
Stock image of the Himalayas
Stock image of the Himalayas


Researchers at the China University of Geosciences in Wuhan have used artificial intelligence (AI) to locate deposits of rare earth minerals and found a large potential reserve in the Tibetan plateau in the Himalayas, the South China Morning Post reported.

In the past, China held a dominant position in mining bulk minerals such as copper, iron, aluminum, and coal, which fueled its industrial and urban growth. However, the changing face of technology now demands using rare earth minerals for multiple applications ranging from energy to defense applications.

Since rare earth resources have been in countries other than China, the latter's dominance has been diminishing. Reserves found in Inner Mongolia have become a major production zone for China. Still, the accidental discovery of lithium in some rock samples from Tibet nearly a decade ago provided hope that could tilt the balance in China's favor once again.

Turning to AI

Geologists in China have long studied the Himalayan belt for minerals but only found granite in places, including Mount Everest. Two years ago, a team of researchers led by Zuo Renguang at the China University of Geosciences built an AI-based system to scan raw satellite data to locate new rare earth deposits.

The AI was trained on a limited data set to identify light-colored granite that could contain rare-earth minerals such as niobium and tantalum alongside lithium, a vital component for making electric vehicles.

Initially, the machine had an accuracy of 60 percent. Still, the team worked on improving the accuracy of its algorithms by adding information about the chemical composition of rocks, their magnetic and electrical properties, and geological maps of the region, which increased its accuracy rate to 96 percent.

Chinese researchers turn to AI to locate rare earth deposits in the Himalayas
AI was used to scan through raw data to locate rare earth reserves

Mining in the Himalayas

The mineral reserves spotted by the machine are estimated to be at least the size of the site in Mongolia, if not bigger. However, mining in the Himalayas is not as straightforward as in Inner Mongolia.

For one, the reserves are located in the Tibetan belt of the country, where it has vowed to safeguard the environment. The Himalayan belt extends into countries like India, Nepal, and Bhutan and holds strategic importance.

Activities like mining contribute to economic growth and attract more people, but some areas are disputed territories and could escalate geopolitical tensions.

From China's point of view, the regions are also remote and will need further investments in infrastructure to make them accessible while also clearing waste from the operations, the SCMP report said. In an area with limited water resources, ill-managed activities could have severe consequences.

Chinese researchers aren't the only ones turning to AI to find lithium, nickel, cobalt, and copper deposits. A Berkeley-based mining company, KoBold, has adopted this strategy and operates at 60 sites across three continents.

Venture capitalist firm Andreessen Horowitz funds the company. A recent round of funding found support from Bill Gates VC firm Breakthrough Energy Ventures and reached a one-billion-dollar valuation, Fortune reported.

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board