李嘉：让数据驱动研发成为“创新引擎”

Core Viewpoint - The integration of data elements with industries is crucial for companies to achieve the core goal of "supply, flow, usage, and security" in their operations [3]. Group 1: Data Utilization - Companies need to establish three empowering systems to ensure effective data usage: making data a "personal expert" for frontline operations, creating "scenario solutions" for key applications, and driving innovation through research and development [3]. - The focus on "supply" involves addressing the internal data asset management, which includes activating and clarifying data resources accumulated over the years [3]. - The company has initiated an AI product expert project to organize 2 million sets of raw material formulas and experimental logs into a knowledge graph, transforming implicit knowledge into explicit and systematic data for production [3]. Group 2: Scenario-Based Applications - In the development of cooling liquids for supercharging piles, the company utilizes "federated learning + knowledge graph" technology to integrate multi-party data while ensuring data sovereignty, achieving cross-domain collaborative modeling [3]. - The company has identified over a hundred practical scenarios in sectors such as transportation, industry, and new energy, which are essential for driving innovation [4]. - In a data center cooling scenario, AI has successfully reduced the Power Usage Effectiveness (PUE) from above 1.5 to 1.05, resulting in annual electricity savings exceeding 30% [4]. Group 3: Innovation and Research - The "smart fluid brain" platform at the company's low-carbon lubrication and intelligent cooling research institute can filter millions of formulas within a minute, predicting performance and carbon footprint, thus reducing the research cycle from 90 days to 45 days and cutting experimental waste by 42% [4]. - The company advocates for the establishment of industry standards and collaborative platforms to gather data effectively, emphasizing the need for leading enterprises and associations to co-create vertical industry data platforms [4]. - The company believes that data transformation should be a strategic priority, requiring commitment from top management to ensure successful AI digitalization [4].