This equipment is developed in accordance with the "Technical Requirements for Intelligence Grading of Humanoid Robots" (T/CIE 298-2025) as the core standard, covering the classification of humanoid robot intelligence levels into five grades from L1 to L5, with different grades corresponding to different capability boundaries and technical requirements: L1 Basic Function Level, L2 Environment Response Level, L3 Task Collaboration Level, L4 Scene Self-Adaptation Level, and L5 Humanoid Intelligence Level. Through the deployment of advanced intelligence capability evaluation equipment (voice command and question-output systems, dynamic environment systems, teaching systems, and visual verification systems), it systematically conducts the "four-dimensional, five-level" evaluation of humanoid robots. The evaluation system is built upon four key capability dimensions — Perception and Cognition (P), Decision-Making and Learning (D), Execution and Performance (E), and Collaboration and Interaction (C) — combined with the five-level intelligence grading standards, establishing a scientific and comprehensive capability assessment framework. It ensures high alignment between evaluation results and the robot's actual intelligence level, effectively overcoming potential interference from different testing environments and significantly reducing subjective evaluation bias caused by human factors. By providing a unified and quantifiable benchmark, it offers an important basis for the R&D and iteration of humanoid robot technologies, guaranteeing the objectivity, consistency, and comparability of the entire intelligence evaluation process.