China intensifying the integration of AI into mineral resource exploration

China is accelerating the integration of artificial intelligence into mineral resource exploration, with experts at a recent forum urging a shift from experience-driven geological surveys to data-driven, intelligent systems to safeguard the country's supply of strategic minerals critical for high-tech industries.

At the Forum on AI and Evaluation & Exploration of Strategic Mineral Resources held recently in Beijing as part of the 28th China Association for Science and Technology Annual Conference, industrial insiders showcased a range of AI-powered tools — from autonomous drilling rigs and heavy-lift drones to geological mapping robots and deep learning models for mineral prediction.

Wang Guofa, a member of the Chinese Academy of Engineering and chief scientist at China Coal Technology and Engineering Group, said that China has a large consumption of mineral resources, yet its proven reserves represent a small percentage of the world total. At least 12 strategic minerals, including iron, copper, nickel, cobalt and lithium, have high import dependency ratios.

Coal, however, remains the notable exception, a position Wang attributed in part to the rapid advancement of intelligent mining technologies.

China has built 1,066 intelligent coal mines, he said, with Shanxi and Shaanxi provinces as well as the Inner Mongolia and the Xinjiang Uygur autonomous regions now producing more than 80 percent of the country's coal output.

"Traditional mining development has relied on a linear path of resource reserves, capital investment and scale expansion," Wang said.

"The future must follow an exponential model driven by data, knowledge accumulation and algorithm iteration."

Wang said China has built 1,066 intelligent coal mines, with Shanxi and Shaanxi provinces as well as the Inner Mongolia and the Xinjiang Uygur autonomous regions now producing more than 80 percent of the country's coal output.

He highlighted efforts to develop a "super mine" ecosystem empowered by digital and intelligent technologies, including a 40-million-metric-ton open-pit coal mine in Xinjiang and unmanned electric haul truck fleets operating in Inner Mongolia.

Beyond coal, the push for intelligent mining is extending to deep metal deposits that are essential for new energy vehicles, advanced electronics and defense industries.

Zhao Xingdong, a professor at Northeastern University in Shenyang, Liaoning province, said China processes abundant deep-seated metal mineral resources, with more than 110 metal mines now operating at depths exceeding 1,000 meters. At such depths, extreme temperatures, high ground stress and corrosive conditions make traditional mining design methods inadequate.

Zhao's team has pioneered advanced ground pressure control techniques, featuring a pre-controlled roof method employing downward fan-shaped drilling. This innovation cuts development and excavation workloads by 40 percent and raises ore recovery rates to above 95 percent, delivering higher productivity without compromising safety.

Song Mingchun, a professor at Hebei GEO University in Shijiazhuang, Hebei province, presented findings from the Jiaodong gold mining district in Shandong province, China's largest gold concentration area with proven reserves of about 6,000 tons — roughly one-third of the national total.

His team used three-dimensional geological modeling to identify a stepwise mineralization pattern, where gold deposits concentrate at structural transitions in fault zones. The model has guided deep exploration breakthroughs and led to a prediction that total gold resources in the region could exceed 10,000 tons.

Wang Liguan, a professor at Central South University in Changsha, Hunan province, said his team developed a digital geological logging system incorporating more than 320 built-in validation rules to ensure data reliability from field collection onward.

"Without authentic and reliable data, AI is just talking nonsense," Wang Liguan said.

Wang Gongwen, a professor at China University of Geosciences (Beijing), called for stronger interdisciplinary standards, noting that geological terms and mining engineering concepts often fail to be converted into formats usable by AI systems.

"The industrial application of AI is the direction forward," Wang Gongwen said. "What matters most is improving quality and efficiency. Meanwhile, AI efficiency compensates for the limitations of human cognition."