The digital transformation of traditional industries, particularly the iron and steel industry—a pillar of the national economy—has become a crucial driver for high-quality industrial development and the transition to new growth engines. CERI Digital Technology (Beijing) Co., Ltd. (hereinafter referred to as “CERI Digital Technology”), as a wholly-owned subsidiary of CERI, keeps committed to driving industrial upgrading through digital intelligence. The Company is deeply engaged in the digital transformation of industrial sectors, focusing on three main areas: intelligent manufacturing, green and low-carbon development, and smart cities. Leveraging advanced digital technologies, it is dedicated to upgrading traditional industries and fostering a new, intelligent, efficient, green, and low-carbon industrial ecosystem. The Company has established an integrated implementation framework encompassing strategy, data, scenarios, technology, and organization, delivering full-chain solutions from industrial internet platforms to AI models and large model applications. This approach empowers the iron and steel industry to transition from "experience-driven" to "data- and model-driven" intelligent manufacturing, providing a replicable path for the digital transformation of long-process industries. 

Identifying Pain Points in Industrial Upgrading 

As a core sector within heavy industry, the iron and steel industry possesses distinct characteristics. First, the production process is highly complex, comprising over 10 main stages, each further divided into multiple sub-processes. Second, a single plant operates thousands of key pieces of equipment, requiring coordinated operation. The equipment varies widely in type, runs in complex environments, and poses significant maintenance challenges. Additionally, the production process involves intricate physical and chemical transformations—including high temperatures, high pressures, and multiphase reactions—making it susceptible to chain reactions triggered by fluctuations in operating parameters. Finally, the reaction sequence is lengthy: From raw material intake to finished product dispatch, the full-chain process generally spans several days to weeks. The industry currently faces six major challenges: i) Collaboration complexity—production scheduling relies heavily on manual experience; ii) Optimization bottlenecks—capacity mismatch between steelmaking and rolling often leads to resource waste; iii) Traceability issues—quality problem tracing is time-consuming, frequently taking over 8 hours or even up to a week; iv) Prediction limitations—equipment failure prediction accuracy is below 60%, and sudden critical equipment failures can cause line shutdowns; v) Slow decision-making—production data is scattered and delayed, resulting in sluggish management response; vi) Low efficiency—some processes still depend on manual labor, leading to inefficiencies and higher error rates. These challenges have severely restricted the high-quality development of steel enterprises. 

In response to industry-specific characteristics and challenges, CERI Digital Technology has developed a comprehensive set of scientific digital transformation implementation strategies: i) Strategy-driven leadership: Foster consensus among senior management, with top leadership spearheading a dedicated digital transformation task force to define clear objectives and pathways; ii) Data foundation establishment: Build a unified data platform, ensuring data quality through cleansing and standardized processing, while establishing robust data security and governance protocols; iii) Scenario-driven implementation: Prioritize high-value, readily deployable use cases—such as quality traceability and equipment failure prediction—to achieve rapid deployment and iterative refinement; iv) Technology integration: Leverage real-time data processing, AI algorithms, and large model technologies to build a technical framework aligned with the real-time operational demands of the steel industry. v) Organizational adaptation: Break down departmental silos by forming cross-functional expert teams integrating IT, OT, process engineering, and management, ensuring technical solutions are deeply rooted in production realities. 

Building Specialized Digital Foundation 

As an industrial digital integrator, CERI Digital Technology adheres to the core philosophy of “focusing on specialized fields and enhancing external cooperation.” The Company concentrates on edge-side data acquisition, intelligent terminal deployment, and enterprise-level application development, leveraging its strengths across areas including quality and energy management. This has enabled the conventional big data platform to upgrade to an advanced industrial internet platform. In the foundational development of its data middle platform and intelligent platform, the Company has established an integrated lakehouse data platform to combine the strengths of data lakes and data warehouses, meeting the iron and steel industry's requirements for massive data storage and efficient analysis. 

To achieve a closed-loop process for the acquisition, transmission, storage, and utilization of data throughout steel production, CERI Digital Technology has developed a hierarchical, categorized foundational data flow architecture for plants. For terminal device data access, second-level data is channeled through the gateway into the Resource IoT Platform and ingested by the Data Platform via Kafka. Millisecond-level data is processed by a high-speed acquisition program and sent directly to the platform through Kafka; high-frequency data is accessed in batches. For legacy IT systems, such as plant Class II systems, inspection and testing, ERP, and metering, data is integrated through a self-developed unified technology platform, which leverages CDC or OGG to extract data from underlying databases. The AI middle platform is directly interfaced with the data middle platform, enabling the development of quality, energy, and process-related applications based on complete data sets. 

Establishing Full-Process Quality Management 

Among the wide range of digital applications, quality management solutions in the iron and steel industry are particularly representative. The design of quality management applications generally centers on four key aspects: First, establishing full-process monitoring that spans the entire value chain—from raw materials intake, pre-ironmaking, hot metal, steelmaking, rolling, to finished products dispatch, and enabling a shift in quality management from post-incident control to pre-incident and in-process control. Second, implementing data stratification, whereby quality data is divided into strata of raw materials, process, and finished products to maintain clear data logic and facilitate systematic management. Third, integrating all data through a unified entry point and main process flow, enabling full-process tracking of quality data without the need to switch between multiple systems. Fourth, adopting closed-loop quality management, whereby data acquisition, aggregation, analysis, and feedback form a closed loop for continuous improvement. 

The core functionalities of this quality application are structured into five major modules: Tracking Platform, Quality Diagnosis Platform, Quality Traceability Platform, Quality Analysis Platform, and Feedback Platform. The Tracking Platform implements real-time monitoring and risk alerts throughout the production process, reinforcing in-process controls. The Quality Diagnosis Platform identifies the grade of finished product quality, enabling refined quality management. The Quality Traceability Platform provides a comprehensive reconstruction of the production process, allowing the traceability of quality risk root causes based on solid evidence. The Quality Analysis Platform leverages advanced analytical tools and scenario-based analysis to uncover intrinsic features within the full-process indicator framework, building a robust quality analytic knowledge base. The Feedback Platform channels insights from tracking, diagnosis, and analysis back to the front line, guiding production and ultimately achieving the PDCA continuous improvement cycle. Additionally, this quality application features a variety of typical interfaces, including real-time monitoring, quality assessment and analysis for the steel rolling process, quality dashboards generated from aggregated data, and a quality cockpit. 

Enabling Full-Process Intelligence in Steel Production 

Currently, CERI Digital Technology has developed extensive AI models supporting over 60 scenario-based applications, covering the entire iron and steel rolling process from raw material intake and steel rolling to finished product dispatch. Production logistics tracking and quality inspection rely primarily on CV models, while production quality prediction is mainly driven by data models, which significantly enhance production stability and quality management efficiency. The Company is actively advancing the application of large-scale models, with a focus on two key areas: intelligent collaborative scheduling and cross-process quality control. Leveraging technologies such as multi-modality data fusion, knowledge graph augmentation, and federated learning, the Company continuously strengthens model performance while ensuring data accuracy. Meanwhile, the Company is driving initiatives in data governance and large model applications within the group, and, as an integrator, promotes related platforms to users. 

Through these initiatives, CERI Digital Technology aims to break down process barriers and enhance production efficiency via full-process collaborative optimization. By employing cross-process quality prediction and closed-loop control, the Company ensures proactive quality management. This approach shifts the Company from experience-based decision-making to a data- and model-driven paradigm, enabling precise decisions. Additionally, the Company leverages large models to support R&D of new materials and processes, significantly shortening the R&D cycle. Currently, CERI Digital Technology's digital transformation practices for the iron and steel industry have been successfully implemented in multiple enterprises, delivering tangible results. Looking ahead, the Company will continue to deepen its focus on intelligent manufacturing applications, driving the digital transformation of the conventional steel industry through innovative technologies.