Currently, large models and intelligent agents are accelerating their transition from the digital world to physical scenes, but embodied intelligence and robot landing still face challenges such as motion lag. In the future, with the commercialization of 6G technology, what changes will it bring to embodied intelligence?

During the 2026 Zhongguancun Forum, Zhang Ping, an academician of the CAE Member and professor of Beijing University of Posts and Telecommunications, said in an interview with Xinhuanet that 6G will break the current bottleneck and promote AI to take root in the real economy through synaesthesia integration, intention transmission and real-time interaction.

Zhang Ping pointed out that the current human-machine object communication is difficult to meet the needs of deep and wide range information transmission and collaborative control in the calculation and control loop. The core reason why existing embodied intelligence and robots are not smooth is that they do not fully follow the operating logic of human information systems and still remain in a purely digital world solution, relying on pre trained models to match decisions, resulting in significant delays.

The future development direction is to enable intelligent agents to act as network nodes with autonomous and learning capabilities, independently completing scene perception, goal orientation, and decision control. In Zhang Ping’s view, 6G is the key support for achieving this goal, and its core breakthrough lies in the integration of sensory perception. By directly driving instantaneous decision-making through real-time perception of the physical world, robots can autonomously adapt to dynamic scenes such as potholes and obstacles without the need to call models, achieving a seamless process of “perception decision execution”.

Unlike 5G and Wi Fi, which are only responsible for data transmission, 6G will revolutionize the communication paradigm. Zhang Ping believes that 6G has shifted from “transmitting data” to “transmitting intent”. Through intent understanding and efficient compression, only necessary instructions are transmitted, and non full data is filtered after transmission. For example, when the left turn light of a vehicle lights up, it knows its intention to go left. As long as this fact is conveyed, it knows what corresponding action to take, achieving real-time, symmetrical coordination and solving the problem of response lag.

Information has an ‘age’, taking yesterday’s data and calculating it today is like a game. Zhang Ping emphasized that artificial intelligence will develop towards reality in the future, deeply dealing with the real economy and physical world, and 6G is the key support for promoting intelligent agents to truly enter the physical world, achieve smooth operation and efficient collaboration.