Group 1 - The Beijing Shijingshan humanoid robot data training center, inaugurated in March 2025, is the first comprehensive facility in China for data collection, scene training, and industrial collaboration [1] - The robot data set is identified as a core driver for the development of robot technology, with the Vision-Language-Action (VLA) model requiring data from real-world interactions, which is costly and inefficient to collect at scale [1] - The robot training industry chain is structured like a pyramid, with core technology at the top, training services in the middle, and application solutions at the bottom [1] Group 2 - TrendForce predicts that the global large language model (LLM) market for robots, including AI training and AIGC solutions, will exceed $100 billion by 2028, with a compound annual growth rate of 48.2% from 2025 to 2028 [2] - Companies like Suochen Technology focus on physical AI platforms, while Liard has successfully completed the "data collection-simulation training-real machine training" process in just two weeks, based on the precision and quality of virtual motion point data [3]