SUPCON Technology (SH:688777) Xin Lang Cai Jing·2025-09-28 06:08Core Insights - The conference highlighted the importance of industrial AI in addressing complex industrial scenarios, emphasizing the need for safe, reliable, and efficient applications of AI technologies like ChatGPT in the process industry [2][3] - The company introduced its "Industrial Embodied Intelligence" model, which integrates AI with industrial knowledge systems, enabling traditional factories to evolve into autonomous industrial AI factories [3][4] Industry Applications - The TPT model demonstrated significant practical value, achieving a prediction accuracy of 99.79% and increasing ethylene yield by 0.373% at PetroChina's Yulin Chemical, resulting in an estimated annual benefit of 15 million yuan [4] - In Inner Mongolia, the TPT model optimized the closed-loop control of an electrolytic cell cluster, reducing daily electricity consumption by over 6,700 kWh, which could save over 120 million kWh over five years, equivalent to the annual electricity needs of 704,000 rural households [4] Market Potential - The industrial added value in China is projected to reach approximately 40.5 trillion yuan in 2024, with the process industry accounting for 45% of this value [4] - The TPT model can reduce labor input by 30-50% and improve production efficiency by 1-3%, potentially decreasing carbon emissions in the manufacturing sector by 1-3% if widely adopted, contributing nearly 1% to national GDP growth [4] Recent Developments - The TPT 2 model has been successfully deployed in Hubei Xingrui within a month of its release, with over 7,000 online users from more than 4,000 enterprises, generating an estimated annual benefit of over 100 million yuan [5] - The company aims to replicate successful models like "San Ning" and "Xing Fa" in high-energy, high-value scenarios to accelerate the application of industrial AI and build a robust industrial AI technology ecosystem [5]