AI software wrings more capacity from a Taiwan slag mill Rockwell Automation's FactoryTalk Analytics PavilionX Model Predictive Control software raised capacity 4% to 5% and cut energy use per metric ton 3% to 10% at CHC Resources' slag grinding line at Taichung Port, Taiwan, with no new equipment added, Rockwell said. The results come from about nine months of rollout and testing comparing manual and AI control, and Rockwell estimates savings of about US$340,000 per year at 2025 energy prices. CHC, Taiwan's largest maker of blast-furnace cement and slag powder and a company controlled by China Steel, plans to extend the software to its southern grinding plant, with no date given. Manufacturing https://heavytechasia.com/tag/topic-manufacturing-3/ AI software wrings more capacity from a Taiwan slag mill Rockwell says its control software lifted capacity 4% to 5% at CHC Resources, with operators now supervising instead of adjusting settings by hand. CHC Resources 中聯資源 , a company in Kaohsiung, Taiwan, controlled by steelmaker China Steel 中鋼 https://www.csc.com.tw/?ref=heavytechasia.com , has added AI control software to a grinding line at its plant in Taichung Port, central Taiwan, with no new equipment. CHC says it is Taiwan's largest maker of blast-furnace cement and of slag powder, a concrete ingredient. Supplier Rockwell Automation, a US industrial automation group with fiscal 2026 sales https://www.sec.gov/Archives/edgar/data/0001024478/000102447826000045/fy2026ex99.htm?ref=heavytechasia.com of US$8.99 billion, says capacity rose 4% to 5% while energy use per metric ton fell 3% to 10%. It is industrial AI reaching a heavy process plant as a software add-on. The software, FactoryTalk Analytics PavilionX Model Predictive Control, predicts how the process will change and adjusts settings such as fuel input, feed rate and mill outlet temperature. It puts the line under semi-automatic control, Rockwell's case study says https://www.rockwellautomation.com/zh-tw/company/news/case-studies/chc-ai-model-predictive-control.html?ref=heavytechasia.com . Operators monitor and manage the process instead of adjusting it by hand. Tang Chih-peng 湯志鵬 , manager of CHC's Taichung plant, says in the case study that the system merges "the operating experience of different shifts into a consistent control mode." Rockwell's https://www.rockwellautomation.com/?ref=heavytechasia.com figures come from about nine months of rollout and testing, in which CHC compared manual and AI control. Rockwell estimates savings of about US$340,000 per year at 2025 energy prices. Rockwell calls the project the first use of the software in Taiwan's slag-grinding industry, and says CHC plans to extend it to its southern grinding plant, without giving a date. Rockwell reports output gains of a similar size from cement mills abroad. Its product page https://www.rockwellautomation.com/en-us/products/software/factorytalk/operationsuite/pavilionX.html?ref=heavytechasia.com cites throughput gains of 4.8% and 5.7% on two mills at an unnamed North American company. A case study https://www.rockwellautomation.com/en-us/company/news/case-studies/cement-optimize-operation.html?ref=heavytechasia.com says Brazil's Cimento Itambé https://www.cimentoitambe.com.br/?ref=heavytechasia.com got 5% to 10% more productivity from its mills with Rockwell's Pavilion8 model-predictive control software, in a phase completed in 2023. Demand for slag powder was growing, CHC https://www.chc.com.tw/en/index.html?ref=heavytechasia.com said in November 2025, udn 聯合新聞網 reported https://udn.com/news/story/7252/9162621?ref=heavytechasia.com . It said the shift from burning coal to gas was cutting supplies of fly ash, forcing concrete plants to use slag powder instead. Sources Editor: Ian Lamont