Peng Zhihui discusses the commercialization of Zhiyuan Robots: "Laying eggs along the way" has covered eight types of scenarios

robot
Abstract generation in progress

On February 28th, the Annual Conference on Standardization of Humanoid Robots and Embodied Intelligence (HEIS) was held in Beijing. Peng Zhihui, Vice Chairman of the Standardization Technical Committee for Humanoid Robots and Embodied Intelligence under the Ministry of Industry and Information Technology, Co-founder, President, and CTO of Zhiyuan Robotics, stated in his speech that, so far, the flexibility of humanoid robots has reached a practical stage. The next competition is about “who has stronger work capability,” not only domestically but also compared to leading overseas embodied intelligence companies, focusing on who can truly implement and deploy.

Regarding industry concerns about commercialization, Peng Zhihui said, “Different stages of technological maturity match different application scenarios, known as the ‘laying eggs along the way’ approach.” Currently, Zhiyuan Robotics has applied its technology in eight typical scenarios, including entertainment performances, guided tours and shopping, industrial manufacturing, logistics sorting, security patrols, education and scientific research, data collection training, and commercial cleaning.

“The ultimate goal of embodied intelligence is infrastructure.”

In Peng Zhihui’s view, “the end goal of embodied intelligence is infrastructure, not just single-point products.” He pointed out that in our physical world, buildings, tools, equipment, transportation systems, etc., are all designed and built for the human body—such as the height of door handles, the size of stairs, the shape of tools, and the organization of tool flows—essentially assuming a human-shaped operator. Based on this, “there should be a clear evolutionary path from digital AI to physical AI. To maximize versatility and compatibility, the final interaction terminal is likely to look like a human.”

Peng Zhihui further explained, “Humanoid robots are expected to become future intelligent infrastructure, not because they look like humans, but because they are the first to unify ‘emotional value productivity’ and ‘labor productivity’ into scalable and evolvable physical terminals.”

“The challenge for the next decade is to scale physical AI, moving from digital AI to physical AI,” Peng Zhihui introduced. Based on three years of exploration and practice, Zhiyuan Robotics has summarized a practical engineering paradigm called “One Body, Three Intelligences.”

Specifically, “One Body” refers to the hardware core of the humanoid robot. Its design is not just a simple hardware stack but a comprehensive safety, reliability, and supply chain engineering system. The two most critical components are the joints and dexterous hands, which determine the upper limits of movement and manipulation capabilities, respectively, accounting for the largest portion of the overall cost. The “Three Intelligences” refer to the robot’s AI “soul,” including motion intelligence, interaction intelligence, and operational intelligence. Motion intelligence is fundamental, providing the basic ability for free movement and full-body actions. Based on this, interaction intelligence and operational intelligence offer emotional value and productivity value, respectively. It is important to note that these three intelligent systems are not isolated; they share a common underlying platform, each providing base models and standardized interfaces, allowing flexible combination and deployment on different robot forms.

“We must get these robots ‘running’ in real scenarios.”

According to China Business News, unlike other humanoid robot companies, Zhiyuan Robotics’ core product lines cover humanoid robots, quadruped robots, commercial robots, and dexterous hand systems. The humanoid robots include the Expedition A2 series, Lingxi X2 series, and Spirit G1, each used in different scenarios with distinct appearances and main functions.

At the conference, Peng Zhihui also revealed Zhiyuan Robotics’ joint series planning, expecting to produce 2 million joints by 2026, with the lowest cost per joint dropping to 500 yuan.

Peng Zhihui summarized Zhiyuan Robotics’ commercialization as “laying eggs along the way,” with applications already in eight typical scenarios. He admitted, “We don’t believe that entering a certain scenario is the absolute correct choice, but we are confident that real-world scenarios are the most effective markets for training systems, exposing boundaries, and verifying robot reliability.” Therefore, robots must be tested in real environments, collecting data and iterating in cycles to achieve breakthroughs in technology and application.

The national strategic window for humanoid robots and embodied intelligence has fully opened, and signals from the industry are very strong. According to data from the Ministry of Industry and Information Technology, there are over 140 domestic humanoid robot companies with more than 330 products released, indicating that the industry has moved from laboratory experiments to the second half of engineering and scenario-based competition.

Peng Zhihui pointed out, “From late 2024 to early 2025, everyone was competing on who could make robots walk straighter and more naturally. Now, robots can perform actions that humans can do, and even those humans cannot do. The next competition is about who has stronger work capability—not only domestically but also internationally—to see who can truly implement and land these technologies.”

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
0/400
No comments
  • Pin

Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate App
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)