Embodied AI FRONTIER
- Your Embodied AI Intelligence Hub -
/ Daily / Zheng Nanning: Embodied AI Is Not Humanoid Robots, Evaluation Isn't About Resembling Humans
Tech

Zheng Nanning: Embodied AI Is Not Humanoid Robots, Evaluation Isn't About Resembling Humans

Zheng Nanning, an academician of the Chinese Academy of Engineering and director of the Institute of AI and Robotics at Xi'an Jiaotong University, published an article on embodied intelligence growing in the physical world.

He points out that a major direction for artificial intelligence is to enable machines not only to process information and reason about complex problems, but also to flexibly complete various tasks in real, changing environments.

From fine operations on industrial production lines to autonomous walking in complex terrain to dynamic movements in performance settings, robots are appearing frequently across scenarios.

Behind these capabilities, AI is undergoing a significant shift: from being able to compute to being able to act, and further evolving into an intelligent form that continuously interacts, learns and evolves in the physical world.

This shift signals a broader transformation in how intelligence is understood and deployed.

It must be clarified that embodied intelligence is not equal to humanoid robots; humanoid form is only one carrier. The key lies in whether the intelligent system can truly enter a real environment, form a closed loop of perception, cognition, decision, action and feedback, and continuously adjust its behavior within this loop, improving internal models and strategies through learning.

Therefore, the criterion for evaluating embodied intelligence is not whether its form resembles a human, but whether it can truly solve real-world problems.

An intuitive way to understand embodied intelligence is to return to how an infant comes to know the world. Infants do not first learn concepts or language; they build understanding through constant touching, grasping, falling and retrying. Infant cognition is not detached from the body but gradually formed through body-environment interaction: vision provides spatial structure, touch provides physical feedback, and motor experience continuously corrects predictions about the world.

Intelligence is not merely a computational process existing in the brain; it is a product of body-environment interaction. Thus, the body is not just a tool for executing commands but a component of the cognitive system, participating in perception, understanding and action. A typical embodied intelligence system consists of three parts: a perception system for acquiring environmental information, a decision system for forming understanding and planning, and an execution system for acting on the physical world.

For machines to truly enter the world and act in it, embodied intelligence requires multiple capabilities. First, world model formation: the system needs to build an internal simulatable environment so it can predict and reason before acting.

Second, multimodal representation: real-world information is highly heterogeneous, including vision, language, touch and sound; embodied intelligence must integrate these into a unified semantic space.

Third, causal reasoning: current large models mainly learn based on statistical correlations and have limited grasp of causal structures; their judgments may sometimes seem linguistically reasonable but do not hold in the physical world.

Fourth, integrated generation and action: an end-to-end system that jointly models vision, language and action, enabling robots to generate action strategies from natural language instructions and to execute them through control systems and actuators.

The development of embodied intelligence relies on two fundamental conditions: high-quality multimodal data as fuel, and high-fidelity simulation environments as training grounds. Training in simulation and then transferring to the real world has become an important technical path, but how to narrow the sim-to-real gap remains an urgent problem.

Current systems still have limitations: the black-box problem, where model decision-making is insufficiently interpretable; the contradiction between computing power and deployment, as large models are difficult to run in real time on edge devices; and an immature industrial ecosystem, with standards and interface specifications still evolving.

In industrial manufacturing, intelligent systems are moving from execution tools to intent-driven systems. In healthcare and rehabilitation, they are becoming important assistive tools for doctors. In urban governance and public safety, drones and ground robots collaborate on infrastructure inspection tasks. In transportation, autonomous driving systems are among the embodied intelligence forms closest to large-scale application.

The essence of embodied intelligence is not to make machines resemble humans, but to let intelligence truly enter the real world, forming continuous learning and adaptation capabilities in complex, uncertain environments.

✓ Verified 2026-09-24
Flash Briefs Live
Today · Sep 24 Thu 13
Yesterday · Sep 23 Wed 31
22:45
Industry 98
World's First Large-Scale Embodied AI Theme Park Opens in Hengqin
22:12
Industry 98
Tesla Supply Chain Audit Rumors Lift Chinese Humanoid Robot Stocks
19:11
Industry 98
Light Rare Earth Prices Rise Sharply as Supply-Demand Dynamics Shift
18:27
Industry 98
PCB price hikes transmit smoothly downstream, more industry applications begin to scale
16:00
Tech 98
KAIST Four-Legged Robot Completes Marathon on Single Battery Charge
13:30
Tech 98
ETH Zurich's New Robotic Hand Can Walk, Press Keys, and Move Objects
13:10
Industry Skild AI 99
Skild AI brings its trained robot to life
13:10
Tech Skild AI 98
Skild AI brings AlphaGo-style self-play to physical world
13:10
Tech Skild AI 99
Skild AI trains robot to play football with 140 years of self-play
12:28
Industry Agility 98
Agility CTO: data security can't be an afterthought for humanoids
10:55
Global Astribot 99
Astribot to debut T1 robot overseas at IROS 2026
10:31
Tech Boston Dynamics 99
Boston Dynamics engineer on Stretch endurance testing
10:21
Capital 99
D-Robotics Completes $400M Series C, Its Third Round in a Year
10:20
Industry 99
Sanrui Intelligent's CubeMars Robot Joints Target Humanoid Robots, Exoskeletons and Wearables
09:05
Industry Noetix Robotics 99
Noetix Robotics unveils Boomy humanoid, hits Kickstarter soon
08:00
Industry lumos robotics 99
Lumos Robotics CEO Yu Chao Named to Fortune 2026 40 Under 40
07:29
Capital 99
Qualcomm to Acquire U.S. Embodied AI Firm Picknik
07:29
Industry 99
Qualcomm to acquire PickNik Robotics, pledges to keep MoveIt open-source
07:00
Industry lumos robotics 99
Lumos Launches Station Lite Real-Robot Validation Platform
06:58
Industry LimX Dynamics 99
LimX Dynamics Teases Luna Self-Boxing Feature via OTA
06:20
Tech 99
Humanoid Robots Use Whole-Body AI to Navigate Narrow Gaps and Obstacles
05:11
Capital 99
Yale PhD Team's Embodied Startup Bets on Psychological Coach Robots, Avoiding General Humanoids
05:08
Industry 99
Kaizhi Data Launches Lighthouse Plan to Gather 1 Million Hours of Embodied AI Data
04:16
Industry 100
FA Robotics and Boyin Hechuang Sign 2,000 Humanoid Force-Control Arm Deal
04:04
Industry 99
Lens Technology Achieves Small-Batch HDD Glass Disk Shipments, Expands in Embodied AI
00:43
Industry 94
Massage Robot Debuts in Huizhou, Combining AI Face and Tongue Analysis with Robotic Therapy
00:38
Tech 94
UCLA's 11-Centimeter Frog Robot Jumps Using Twisted Elastic Rods
00:30
Industry 94
Newport Hospital Deploys Rhode Island's First da Vinci 5 Surgical Robot, Fully Donor-Funded
00:22
Industry 94
Goldman Sachs Raises Humanoid Robot Forecast to 6.5M, Sees 49% Price Drop
00:18
Industry 96
Musk Predicts 100 Billion Humanoid Robots Within 20 Years
00:01
Capital 94
Zhongyi Embodied Raises Nearly ¥100M Angel Round to Build Robot Operating System