Environmental (E)

Social (S)

Economic (E)

The system saves process costs, accurately controls steel tapping time and reduces energy consumption by 1%

The operators may pass on their experience and reduce the occurrence of industrial accidents

It reduces unnecessary energy loss, accurately controls steel production, and improves steelmaking quality

Solutions

ITRI collaborates with China Steel Corporation to establish a digital twin of the tapping process, integrating AI technology and conducting molten steel fluid dynamics simulation using six million particles. This is combined with prediction and recognition technologies and a steel flow pattern estimation model to obtain simulation information as a reference for process control. The simulation results correspond to actual data from China Steel Corporation with an accuracy rate exceeding 90%. This system has won the recognition of the 2024 R&D 100 Awards.

Technology Application

²  Simulation of Steel Flow particles: Conducting dynamic simulation of steel flow using a large number of particles to replicate the physical characteristics of molten steel flow.

Five AI models for real-time analysis and feedback on process patterns:

ž    Fluid Physical Simulation Model: Simulating the physical patterns and behaviors of molten steel flow.

ž    Molten Steel Inventory Estimation Model: Accurately measuring and estimating the remaining amount of molten steel in the converter.

ž  Steel Flow Information Recognition Model: Analyzing key information such as the width of the steel flow and the amount of slag falling.

ž  Steel Flow Deposition Prediction Model: Providing real-time prediction of the location of steel flow deposition to enhance accuracy.

ž  Empirical Converter Wear Model: Dynamically adjusting the parameter of converter wall thickness based on the information of furnace wear.

Achievements

²  The digital replica, which is the result of the collaboration between ITRI and China Steel Corporation, has won the recognition of the 2024 R&D 100 Awards.

²  This project represents the first application of a digital twin system in the steel industry. By integrating visualized virtual AI model calculations, physical engine simulations, and million-particle technology, it accurately replicates the actual steel tapping process. Furthermore, it estimates critical information that cannot be directly measured in reality, thereby establishing a reliable experience curve. This lays a solid foundation for the realization of smart and automated steel tapping.

Environmental (E)

Social (S)

Economic (E)

The Technology Baseball System has reduced data analysis time by 40%, minimizing resource waste caused by inefficient repetitive training.

Assisted the Taitung Red Leaves team in winning the championship, reduced the digital divide between urban and rural areas, and promoted talent development.

Doubled the data transparency and analysis efficiency, reduced injury probability by 30%, and established a new model of data-driven sports training governance.

Solutions

Utilizing the "Technological and Automated Ball-Strike System," "Batter Training Analysis," and "Training Load Algorithm," among other sports sensing technologies combined with AI algorithms, personalized data monitoring and training planning are conducted for players. This technology was also introduced to the Hung Yeh Elementary School baseball team to verify and enhance their training effectiveness. Gradually, it is being extended to professional baseball fields and sports venues, forming an industrialized application model that effectively meets baseball training needs at various levels.

Technology Application

²  Image Data Conversion Technology: Utilizing deep learning to identify the trajectory of the ball and the positions of players' joints, capturing the movement data of the ball and players on the field.

²  3D Ball-Strike System: Employing computer visualization technology to determine the pitching strike zone location with a coordinate error of less than 2 centimeters and a ball speed error of only ±2.9 kilometers per hour.

²  Application of Data Visualization Analysis: Through data integration and visualization, the efficiency of team data analysis has been improved by more than twofold. It enabled the prediction of pitching behaviors under different scenarios, assisting coaches in enhancing training effectiveness.

Achievements

²  Training System Enhancement: A new foundation for baseball data analysis has been established to assist coaches in improving instructional effectiveness and achieving scientific training for athletes.

²  Enhancement of Data Analysis Quality: Ensuring data consistency and accuracy to guarantee the quality of analysis, which has been practically applied in the 2024 Chinese Professional Baseball League Championship Series and the exhibition game between Chinese Taipei and the Czech Republic before the 2024 WBSC Premier12 tournament.

²  Talent Development and Community Impact: Through the introduction of the Technology Baseball System, the Hung Yeh Elementary School baseball team won the championship in the 2024 Taitung County Elementary School Baseball League, enhancing the local baseball strength.