November 12 news: From September 20 to October 26, 2026, the Embodied Intelligence Testing Laboratory of the Institute of Artificial Intelligence at the China Academy of Information and Communications Technology conducted the EAIBench special evaluation for industrial scenarios. The evaluation aimed to advance real-world training and iterative optimization of embodied intelligence in industrial settings, and to leverage testing to drive research, evaluation, and application. It focused on the cross-embodiment, cross-model, and cross-scenario operational capabilities of embodied intelligence in industrial scenarios. Targeting 7 representative tasks, 7 open-source models, and 3 embodiments, the evaluation employed unified test questions, unified criteria, and a standardized environment to examine real operational capabilities, completing a cumulative total of 147 large-scale real-machine tests and 3,234 rounds.