Site Change story Top five use cases Impact Beijing Shougang Cold Rolling Beijing, People’s Republic of ChinaTo meet the challenge of stricter quality standards for high-end automotive manufacturing and increasingly diverse stock-keeping units (SKUs), the site deployed 67 4IR use cases, 61% of which utilized AI. This enhanced end-to-end process accuracy, resolved customer quality challenges and eliminated key quality and throughput constraints. As a result, high-end sales increased by 36%, customer complaints decreased by 55%, product defects dropped by 35% and production line efficiency improved by 21.2%.Knowledge graph-enabled AI expert system for customer stamping quality improvement 248%Number of customers’ stamping parts ML-enabled process settings optimization 35.1% Product defect rate AI neural networks-enabled galvanizing process real- time close loop control 46.4% Thickness defect rate Advanced analytics-based production line efficiency improvement 18.7% OEE Intelligent production planning & scheduling based on advanced algorithms 24.4%Production line utilization CEAT Limited Sriperumbudur, IndiaTo support global expansion, CEAT needed to manage three times more SKUs, faster order fulfilment, coupled with new product launches at twice the speed with productivity improvement in assembly process. To achieve this, CEAT deployed over 30 digital solutions, including operational research models for reducing turnaround time, advanced analytics for predictive control and ML-based design. These solutions improved labour productivity by 25%, reduced dispatch turnaround time by 54%, accelerated product ramp-up by 30% and cut scope 1 and 2 emissions by 47%.Advance analytics- controlled cycle time (CT) optimizer for batch mixing 18% Cycle time ML-based die design for improved ramp-up time 78%New product ramp-up time Operations research (OR) model for export container plan optimization and TAT2 reduction 288% Dispatch throughput Analytics-driven operator performance monitoring & skill enhancement linked to VR training system 29% Labour productivity Vision system-driven zero- touch automation system for dispatch 57% Truck turnaround time CITIC Dicastal Ameur Seflia, Morocco“Lightweighting” trends are reshaping the automotive sector, by driving intense competition for high-quality auto parts with smaller environmental footprints. In response, CITIC Dicastal Morocco deployed over 40 digital use cases for high-precision, flexible production and efficient use of materials. The site deployed advanced algorithms for casting and machining, an AI-generated content- enhanced vision inspection system, and a process control to manage natural gas quality volatility in furnaces – a local challenge. This led to a 17% improvement in overall equipment effectiveness, a 27% increase in labour productivity, a 31.1% reduction in defects and a 53% reduction in scope 1 and 2 emissions.Multi-task ML algorithms for intelligent casting 19.6% Defect rate Advanced sensors and deep learning algorithms for precise control of product weight 40.1%Weight accuracy ± % tolerance AIGC-enhanced industrial inspection robots for complex parts 98.1%Escape rate of critical defects Predictive maintenance of mold enabled by knowledge graphs elevates life-cycle performance 5.7% Mold OEE Advanced process control in furnaces to manage variable energy sources for carbon reduction 37.1%Natural gas consumption Global Lighthouse Network: The Mindset Shifts Driving Impact and Scale in Digital Transformation 40

Global Lighthouse Network 2025