How AI Is Transforming Coal Mining: Inside China’s First National Mining AI Model Competition

On January 6, the first national coal‑industry mining AI model competition finals and award ceremony were held in Beijing, jointly organized by the China Coal Society, China University of Mining (Beijing) and Huawei Technologies.

Key government and industry leaders, including Deputy Directors Xing Huaibin (Ministry of Science and Technology), Ren Lixin (National Energy Administration), and Academician Yuan Liang, attended the event.

The competition focused on innovative mining intelligence scenarios, inviting nationwide AI application proposals to achieve the goal of "fewer people, unmanned, safe and efficient" mining, and to build a platform for technology exchange, talent cultivation, and collaborative opportunities.

Yuan Liang emphasized that the competition is a proactive exploration that can drive innovation in the traditional coal sector.

Ren Lixin described the contest as a youthful force empowering the coal industry with modern technology, noting that 2023 is a pivotal year for advancing mining intelligence to its second‑stage goals.

Xing Huaibin highlighted that intelligent mines are a crucial AI application scenario, and hopes the emerging products and technologies will be effectively transformed to support industrial upgrading.

Fifteen teams reached the finals, with one special prize, two first prizes, four second prizes and eight third prizes. The special‑prize winner, IMP‑Lab‑1 from China University of Mining (Beijing), used Huawei’s mining AI model with machine‑vision and robotic picking to replace manual coal‑gangue detection, improving efficiency and equipment lifespan.

First‑prize teams included the "Sea‑Team" from China University of Mining (Beijing), which achieved unstructured road‑edge detection and multi‑scale target detection for open‑pit mines, and a team from Xi’an University of Science and Technology that provided an intelligent coal‑rock microscopic composition recognition solution.

Overall, 389 teams and 1,460 participants joined the competition. Huawei’s mining AI model platform enabled hundreds of AI‑driven scenarios, many of which have already been applied in production.

Huawei’s mining AI model is presented as an industry‑pretrained model that provides automated tools, reducing the need for manual design and enabling rapid scenario modeling.

The competition identified four major gaps to be addressed for smart‑mine AI adoption: the core‑technology gap (AI penetration in energy is below 3%), the scenario gap (diverse mining processes require collective effort), the talent gap (scarcity of professionals skilled in both mining and AI), and the ecosystem gap (need for open collaboration among regulators, enterprises, equipment makers, and developers).

Huawei proposes a "crowdfunding model" to build an AI scenario map and applications through collective contributions, turning AI into a continuous source of innovation for the mining sector.