NISCO and Huawei deploy YuanYe steel model powered by swarm intelligence

Nanjing Iron & Steel Co., Ltd. (NISCO) has worked with Huawei and ecosystem partners to commercially deploy the YuanYe Steel Large Model, described as the world’s first steel model powered by swarm intelligence, across multiple scenarios at scalesource.

The model embeds AI into ironmaking, steelmaking, rolling and quality inspection. In ironmaking, furnace temperature and health are predicted two hours ahead with over 90 % accuracy, raising prime-grade molten-iron yield by over 10 % while cutting unit energy consumption.

In rolling, a millisecond-level force-prediction model narrows thickness tolerance to ±0.2 mm, reduces steel-thickness deviation to 1.9 %, and lifts the rolling-force hit rate from 88 % to 98 %, supporting an annual output of over 2.5 million tons of steel plates and a 0.5 % increase in overall yield.

Quality inspection now runs fully automated metallographic checks, boosting accuracy to over 95 % and shortening per-sample time from 10 minutes to 3–5 minutes, an 80 % gain in efficiency.

Li Jinyan, General Manager of the Digital and Intelligent Integrated Development Department and Director of the AI Research Institute at NISCO, said the transformation is “human-centric” and aims to free employees for higher-value work. To date NISCO has deployed over 300 AI models, more than 1,000 low-code agents and lightweight applications built by frontline staff.

Cheng Yongfu, Senior Industrial Solution Expert at Huawei’s Steel & Non-ferrous BU, noted that industrial AI’s real value emerges when it expands from single-point applications to comprehensive integration across production, operations and management.

Our read: The YuanYe deployment shows how a tightly integrated, multi-model AI architecture can turn a legacy heavy-industry operation into a data-driven enterprise, delivering measurable cost cuts and productivity gains while keeping human workers in supervisory roles.