Inside the roaring workshops of a special steel plant, the traditional production model that relied on “experience and intuition” is quietly giving way. Real-time data, intelligent algorithms and digital systems have become the “new stewards” of production management.

At Hebei Yongyang Special Steel Group Co., Ltd., artificial intelligence operates around the clock across multiple areas, including raw material inspection, intelligent steelmaking, equipment maintenance, energy management and financial operations. AI has been deeply integrated into the entire production and business chain, driving a profound transformation powered by intelligent technologies.

Step by step, this digital transformation has delivered tangible results. Yongyang Special Steel has been selected as a national digital supply chain exemplary case by the Ministry of Industry and Information Technology of China, and has received AAA certification under the Integrated Management System for Informatization and Industrialization Integration, as well as recognition as a model enterprise for quality improvement and brand development.

Bringing Digital Technology to the Production Frontline

Founded in 1993, Yongyang Special Steel has grown into a leading enterprise in China's light rail and heavy rail sectors. It is a member of China's Top 500 Private Manufacturing Enterprises, a National High-Tech Enterprise, and a nationally recognized Green Factory.

The company's achievements are impressive: its light rail products hold approximately 70% of the domestic market share, industrial heavy rails account for around 65% of the core domestic supplier market, and fork steel flat bars represent about 70% of the market for logistics equipment steel materials.

Today, Yongyang Special Steel has become one of China's largest production bases for light rails, heavy rails and specialized steel profiles for mining applications, with the widest product range in the industry.

Facing the wave of manufacturing transformation and upgrading, the company has not simply pursued expansion in scale. Instead, it has made intelligent and digital transformation a core development strategy. Through years of continuous improvement, Yongyang Special Steel has established a transformation path from digital foundation building to intelligent manufacturing advancement.

Digital transformation is not achieved overnight

Before 2017, Yongyang Special Steel completed the electronic transformation of its basic business data.

From 2017 to 2021, the company established a private cloud platform and ERP system, integrating business operations, financial management and information flows.

In 2022, the company accelerated its intelligent transformation by building a 5G private network, breaking down barriers between IT and OT systems, and applying AI technologies across raw materials, steelmaking, equipment management, energy control and business operations.

This created a “1234+2” intelligent architecture consisting of:

• One cloud platform
• Two networks
• Three major platforms
• Four categories of application scenarios
• Dual security protection systems

“Digital transformation is not simply about purchasing equipment or installing software systems. The key is to address real production challenges and solve practical problems,” said Dai Ziwei, Director of the Enterprise Management Department at Yongyang Special Steel.

In the past, many production processes relied heavily on manual experience. This not only increased labor intensity but also created challenges in quality consistency, production efficiency and workplace safety. By introducing AI, Yongyang Special Steel aims to precisely solve these operational bottlenecks and make digital technology truly serve frontline production.

From Experience-Based Steelmaking to Data-Driven Steelmaking

Raw material inspection is the first and one of the most critical steps in steel production. Traditionally, scrap steel grading relied entirely on manual visual inspection. The complex working environment created safety risks, while subjective judgment and inconsistent standards could lead to quality disputes and management issues. Paper-based records also made full-process traceability difficult.

Today, the AI-powered scrap steel intelligent grading system has been fully deployed.

High-definition cameras across the plant operate 24/7 to capture the entire unloading process. Using edge computing technology, the system performs millisecond-level intelligent analysis. Drivers can initiate the grading process through self-service terminals, while operators in the control center can monitor images and preliminary grading results in real time, conducting remote verification and adjustments when necessary.

This creates a new human-machine collaboration model combining “machine intelligence + human verification.”

“AI has transformed inspection standards into measurable and standardized rules, achieving grading accuracy of 98%,” said Xi Caihong, Information Department Manager at Yongyang Special Steel.

After implementation, vehicle waiting time has been significantly reduced, inspection efficiency has improved by more than 40%, and raw material quality disputes have decreased by 80%.

The system not only removes workers from hazardous inspection environments but also creates complete digital records, providing valuable data support for production planning, material allocation and procurement management.

Steelmaking has long been regarded as a “black box” within the steel industry. Inside furnaces operating above 1,600°C, chemical reactions cannot be directly observed. For decades, production quality depended heavily on the “hidden experience” of senior steelmaking experts, making product consistency difficult to guarantee.

To overcome this industry challenge, Yongyang Special Steel developed an intelligent steelmaking system.

The system combines metallurgical mechanism models with machine learning models, integrating real-time data from key equipment such as converters and refining furnaces. It collects critical parameters including temperature, chemical composition and pressure changes with millisecond-level response.

Based on big data analysis, the AI system automatically calculates optimal process routes and intelligently provides operational instructions, including material charging ratios, oxygen blowing intensity and power supply curves.

This enables closed-loop control from decision-making to execution, achieving a fundamental transition from “experience-based steelmaking” to “data-driven steelmaking.”

Meanwhile, through the converter large language model intelligent Q&A module, real-time operating data, process optimization results and expert knowledge are transformed into an accessible “steelmaking knowledge brain.”

The system preserves generations of steelmakers' practical expertise as digital assets, enabling standardized and sustainable knowledge transfer.

Turning Computing Power and Algorithms into Green Benefits

Continuous steel production requires extremely high equipment stability. Traditional maintenance approaches based on “repair after failure” or scheduled overhauls often result in unexpected downtime or inefficient maintenance resource allocation.

To address this challenge, Yongyang Special Steel launched an equipment predictive maintenance platform.

The system continuously collects operating parameters such as vibration, temperature and current signals. Through AI algorithms, it identifies early-stage equipment risks, provides graded warnings, analyzes potential failure causes and recommends maintenance solutions.

This shifts equipment management from “reactive repair” to “predictive maintenance.”

So far, the system has detected 23 potential equipment issues within six months. Overall maintenance costs have been reduced by 10%–20%, effectively minimizing the risk of unplanned production interruptions.

AI is not limited to production workshops. Its capabilities have also expanded into energy management and financial operations.

The intelligent energy management system predicts and optimizes gas generation and consumption, increasing converter gas recovery efficiency by 17% and saving approximately RMB 11 million in annual external energy purchasing costs.

The intelligent payment and collection system enables direct bank-enterprise connectivity. AI robots can complete intelligent matching and automatic reconciliation between bank transaction records and ERP accounts receivable and payable documents within milliseconds.

Even complex cases such as abbreviated customer names or irregular transaction descriptions can be accurately processed through semantic recognition technology.

As a result, financial employees are freed from repetitive reconciliation tasks and can focus more on risk management and business analysis.

From a production model dominated by human experience to one driven by intelligent data, and from isolated equipment upgrades to full-chain intelligent empowerment, Yongyang Special Steel's transformation represents a vivid example of the digital and intelligent upgrading of Hebei's steel industry.

AI-based digital technologies are not merely additional tools. They directly address challenges across every stage of production, transforming computing power and algorithms into measurable improvements in quality, cost efficiency, safety and environmental sustainability.