Baosteel's AI-powered blast furnace cuts carbon emissions and operating costs in Shanghai steel plant

Baosteel is retrofitting its No 1 Blast Furnace in Shanghai with full AI control, making it the last of four furnaces at the Baoshan base to go fully AI-powered. The system predicts furnace temperature and sulfur content in real time, solving a long-standing problem in steelmaking where extreme heat prevents direct measurement. Each furnace is expected to save over 10 million yuan annually and cut carbon emissions by 5 kilograms per metric ton of hot metal. The upgrade targets the 'black box' nature of blast furnace ironmaking, which has historically relied on human experience despite highly complex conditions. By stabilizing temperature around 1,510 degrees Celsius, the AI reduces downstream disruptions and production costs. This sets a benchmark for intelligent manufacturing in steel, one of the hardest industries to decarbonize. For carbon markets and climate policy watchers, this is a concrete example of how AI can drive efficiency gains and emissions reductions in heavy industry. The 5 kg CO2 reduction per ton of hot metal may seem small, but scaled across China's massive steel output, the cumulative impact is significant. It also shows that industrial AI is moving beyond hype into measurable operational improvements.