The first industry standard in the field of embodied intelligence, jointly drafted by the China Academy of Information and Communications Technology and more than 40 units, was officially released on March 26th. This standard establishes a unified benchmark testing framework for the embodied intelligence field, marking a new stage of embodied intelligence evaluation with standards to follow.
It is understood that this standard focuses on key foundational technologies of artificial intelligence and benchmark testing methods for embodied intelligence, while clarifying the framework and capability requirements for embodied intelligence systems. It will be officially implemented on June 1, 2026. The standard specifies the benchmark testing framework, methods, and indicators for embodied intelligent systems in both simulated and real environments. The evaluation system proposed by the standard supports the testing of basic abilities, cognitive reasoning abilities, and full loop closed-loop abilities, covering four testing methods: static simulation testing, dynamic simulation testing, real environment testing, and combined testing.
Experts pointed out that in order to ensure the operability of the standards, a supporting testing task library, testing tools, and standard testing environment were synchronously constructed during the standard preparation process. At present, more than 10000 testing task libraries have been built, covering 300 types of tasks such as industrial, household, retail, and logistics. We have developed testing tools for data collection and management, simulation task generation, and automated calculation of indicators.