Arya News - The two systems — AI-GeoMapping and AI-OreSeeking — were released by the China Geological Survey of the Ministry of Natural Resources at the 28th China Mining Conference and Exhibition in Tianjin, as the country pushes to secure critical mineral supplies.
BEIJING – China unveiled two geological mapping and mineral exploration systems powered by artificial intelligence technologies on Friday, marking a shift from decades of expert-driven fieldwork to more accurate, comprehensive, and efficient data-and-algorithm-based predictions.
The two systems — AI-GeoMapping and AI-OreSeeking — were released by the China Geological Survey of the Ministry of Natural Resources at the 28th China Mining Conference and Exhibition in Tianjin, as the country pushes to secure critical mineral supplies.
Geological mapping serves as a road map for the subsurface mineral realm, revealing how strata, structures, and rocks are distributed. It is the most fundamental and critical prerequisite work for discovering mineralization clues and delineating prospecting target areas, according to Li Fengdan, an official with the China Geological Survey who is responsible for AI-GeoMapping.
After a decade of development work, the system has evolved from breakthroughs in core AI algorithms and specialized software platforms to a full-process intelligent ecosystem, Li added.
Li said that the system addresses long-standing field challenges, such as inaccessible rugged terrain, limited rock exposure under forest, swamp, or desert cover, and the complexities of identifying geological units across extensive survey areas.
“It identifies geological features with 90 percent accuracy and lifts data processing efficiency by more than half compared with conventional methods,” Li said, noting that the system fuses satellite imagery, airborne geophysical surveys, and ground observations through AI.
The research team overcame key technical hurdles, including AI-driven geological object recognition based on field route deep learning, as well as multimodal fusion of geological, geophysical, geochemical, and remote sensing data.
The system has been tested on nearly 100 regional geological survey sheets at a 1:50,000 scale in places such as Yunnan, Anhui, and Qinghai provinces, and has been promoted to countries like Morocco, Laos, and Rwanda.
“With multi-language support built in, the system will offer a Chinese solution for AI-powered geological surveying that can benefit the global community,” Li added.
Where mapping reveals the geological picture, AI-OreSeeking goes further — integrating not just mapped geological data, but also geophysical, geochemical, and remote sensing datasets to predict where mineral deposits may lie.
Wang Bin, an official with the China Geological Survey who is responsible for the prospecting system, said the system was developed over three years to meet the demand for rapid prediction and evaluation of mineral resources. He likened the platform to an “intelligent brain” — a unified system that consolidates what was once a fragmented workflow spread across various stand-alone software packages.
The platform comprises five core modules and more than 200 algorithms, covering the full chain from data management and interpretation through predictive evaluation, map production and report generation. It is built around a large prospecting model combined with knowledge graph constraints, fusing millions of geological knowledge entries to ensure predictions conform to established mineralization principles.
With each drill hole costing millions of yuan, Wang said the system can effectively narrow down favorable exploration targets before expensive field verification, reducing the cost of blind field exploration. “It shifts mineral exploration from expert experience-driven to data- and knowledge-driven prediction, opening a new chapter in intelligent mineral exploration,” Wang added.
In trial runs across the Qinling Mountains, Jiaodong Peninsula, and Wuyi Mountains metallogenic belts, the system delineated a series of gold, copper, iron, and lead-zinc exploration targets. Taking the Zhaishang-Liba mining district in the West Qinling as an example, the system completed data processing and prediction for 32 survey sheets at a 1:50,000 scale in just five days, newly delineating two exploration targets that were subsequently verified by drilling — work that would typically take months using traditional methods.
Wang said that trial applications across more than 100 projects in over 10 provincial-level regions have shown that the system can capture weak mineralization-related anomalies that are difficult for humans to detect, identifying new favorable exploration targets.
“The overall efficiency has improved by more than 60 percent, pushing mineral exploration toward being comprehensive, fast, and precise,” he said, highlighting its particular value for deep mineral exploration.
The two systems come as China steps up efforts to apply AI to its new round of strategic mineral exploration, a campaign led by the Ministry of Natural Resources aimed at securing breakthroughs in domestic reserves of gold, copper, lithium, and other critical minerals. The goal is to identify a new batch of developable mineral sites by 2030 to ensure steady reserve growth at home and strengthen national energy and resource security.