Exciting News! Zhongkong Technology TPT Selected as an Outstanding Case in the "2025 China AI+ Application Top 50"

(Source: Zhongkong Technology)

News

Recently, the Shanghai Publishing Group’s Financial Associated Press and Science and Technology Innovation Board Daily officially announced the list of “2025 China AI+ Application Top 50” outstanding cases. Zhongkong Technology’s time-series large model TPT (Time-series Pre-trained Transformer) stood out from over 300 applicants and was successfully selected, demonstrating Zhongkong Technology’s deep expertise and innovative strength in the industrial AI field. It also serves as a benchmark model for empowering the real economy through industrial AI.

The “2025 China AI+ Application Top 50” selection focuses on the large-scale implementation and practical value of AI technology. It aims to discover high-quality AI applications rooted in various industries, showcasing China’s achievements in moving from technological exploration to industrial empowerment. The selection process strictly adhered to the core standards of “high scene fit and effective empowerment,” with multiple rounds of professional review. Ultimately, 50 outstanding cases were chosen that combine technological innovation, industrial practicality, and industry demonstration, covering key sectors such as manufacturing, finance, education, and healthcare. Industry leaders like Huawei, Ant Group, and Midea are among the list, which is highly recognized within the industry.

As Zhongkong Technology’s core engine for industrial AI designed for process industries, TPT’s successful inclusion is mainly due to its original algorithm architecture, significant energy efficiency improvements, and extensive industry deployment practices. It also affirms Zhongkong Technology’s innovative approach to integrating “AI + Industry.” TPT is a generative AI algorithm framework independently developed by Zhongkong Technology based on automation scenarios in production processes. It pre-trains by integrating massive data from various industries, factories, and equipment, learning general rules of industrial device operation. This breaks the traditional “one scene, one model” development mode, allowing for reuse across different devices and conditions with minimal or zero fine-tuning. It has strong generalization capabilities, adaptable to various production processes, and supports online real-time optimization and closed-loop control. Currently, TPT has been successfully applied in over 100 industrial scenarios, covering petrochemical, chemical, new energy, and other fields.

Typical Value Cases

▶ At Wanhua Chemical’s Ningbo Chlor-Alkali Plant, TPT reduced electrolyzer caustic soda power consumption by about 5%, achieved 95% accuracy in ion membrane lifespan prediction, and compressed waste liquid pH neutralization time from 5 hours to 1 hour, saving over 10 million yuan annually.

▶ At Sinopec Lanzhou Petrochemical Yulin Chemical Ethylene Plant, the ethylene yield per furnace increased by 0.373%, coking early warning was advanced by 3-4 days, and key indicator prediction accuracy reached 99.79%, generating an annual economic benefit of over 20 million yuan.

▶ At Guangxi Huayi Energy’s billion-level base, TPT reshaped factory operations through the core scenarios of “safety, optimization, and talent empowerment,” reducing the standard deviation of hydrogen sulfide fluctuations by 30%. It can automatically calculate optimal settings for main oxygen, combustion air, and natural gas flow within 30-40 seconds, stabilizing the temperature of the main combustion furnace and achieving safety, environmental protection, and efficiency improvements.

▶ In the chlor-alkali plant of Hubei Xingrui Silicon Materials under Xingfa Group, TPT deeply integrates with UCS, helping to create the world’s first fully autonomous AOP plant. The concentration of caustic soda remains stable within a narrow range of 32%-32.1%, with overall production efficiency increasing by 1%-3%.

Zhongkong Technology will continue to deepen its focus on industrial AI, iteratively optimize the core capabilities of the TPT large model, and promote the integration and innovation of “AI + Industry.” It aims to enable more industrial scenarios to upgrade intelligently, truly root AI technology in industries, serve industrial development, and inject stronger intelligent momentum into activating new productivity and high-quality industrial growth in China!