Gespräch mit Zhu Qiuguo, CEO von Dialogue Cloud Deep: Wir sind die "Sechs kleinen Drachen" aus Hangzhou, und wir sind immer noch ein kleines Unternehmen.
Business customers will not pay for the "Six Dragons from Hangzhou". After all, it depends on the product.
After Deep Robotics was included in the "Six Dragons from Hangzhou" together with Unitree Technology, it gained great fame. However, the founder and CEO Zhu Qiuguo has rarely made himself available for external interviews so far.
In November 2017, when autumn was in the air, the company was founded. Zhu Qiuguo loves traditional Chinese culture, and the name "Deep Robotics" refers to the poem "On the Way to the Mountain" by the Tang Dynasty poet Du Mu: "The path winds steeply up the cold cliff, and hidden in the deep clouds lie the villages."
For a long time, Deep Robotics has behaved as its name suggests: discreet and reserved. Zhu Qiuguo told "China Entrepreneur": "The poem says that one should explore the unknown and serve humanity."
In December 2024, Deep Robotics' Chinese robot dog "Jueying X30" carried out an autonomous inspection tour in the power tunnels of Singapore Power, an energy company. In case of abnormalities such as water accumulations or cracks that could cause disruptions in the power grid, it could immediately record and report them.
Jueying X30
This was the first successful application of a four-legged robot from China in a foreign power grid. For this, Deep Robotics received praise from Lin Jian, the spokesman of the Chinese Ministry of Foreign Affairs, on social media: "This is so cool!"
Deep Robotics' robot dogs can operate in any environment and are therefore referred to by users as "special units for dangerous situations". In addition, Deep Robotics is also a master in cost management. The costs of robot dogs with the same performance are only one-third to one-half of the costs of products from Boston Dynamics.
In August this year, "China Entrepreneur" met Zhu Qiuguo at the World Robot Conference. He said that he had long had the dream that robots would move from the laboratory into practice.
"They should be able to be deployed outside and help people in dangerous, difficult and complex situations. And when people really rely on them, they should be there."
This is also the philosophy that Deep Robotics has followed since its founding.
Zhu Qiuguo announced that Deep Robotics' business volume in 2024 increased by more than 100% compared to 2023. According to him, Deep Robotics will be able to deliver 10,000 robots in 2025. Currently, the company has already developed markets in the Asia-Pacific region, the Middle East as well as in Europe and America.
Currently, Deep Robotics' team has nearly 400 employees, who are mainly engaged in the fields of research and development, production and sales. In July 2025, Deep Robotics completed a financing of nearly 500 million yuan. The participation was jointly led by Fortune Capital and Guoxin Fund, followed by investors such as Beijing Robot Industry Development Investment Fund, Qianhai FOF, CCTV Media Convergence Fund and Huaying Capital.
To date, Deep Robotics has completed a total of seven rounds of financing. The investors include, among others, Inno Angel Fund, Yuanhe Origin Point Capital, Bangsheng Capital, Shenzhen Smart City Investment, Huajian Function Investment, Fangguang Capital and Saizhi Bole Investment. Many of them have state backgrounds.
Zhu Qiuguo said that the latest financing funds will mainly flow into the expansion of the production line for four-legged robots, the research and development of humanoid robots and the recruitment of high-quality professionals.
From laboratory work to practical application
Deep Robotics was founded in November 2017 by Zhu Qiuguo and Dr. Li Chao, who also comes from the laboratory of Zhejiang University.
Deep Robotics' self-developed robots of the "Jueying" and "Bobcat" series are already being used in various environments, including power grid inspections, security guards, exploration missions, public rescue operations, tunnel inspections, metal smelters, construction surveys as well as in research and teaching.
An industry study by IDC shows that the global market for four-legged robots reached a turnover of more than 180 million US dollars in 2024 and about 20,000 devices were sold. The demand in the fields of energy, oil and public safety increased. Deep Robotics is the market leader in several industries for business customers.
In 2018, Deep Robotics presented the first four-legged robot in China that could climb stairs, navigate autonomously and perform intelligent interactions. In 2019, it released the first four-legged robot in China that could recharge itself. At the World Robot Conference 2024, Deep Robotics presented its first humanoid robot DR01 and thus officially announced its entry into the field of humanoid robots. In April 2025, it released the wheel-legged robot Bobcat M20, which can also operate in difficult environments such as mountain roads and rubble.
Zhu Qiuguo spent nearly three years on solitary research until he moved from technological innovation to practical application.
When Zhu Qiuguo and Li Chao started to found the company, capital did not look favorably on university professors who wanted to become founders. At that time, they had very modest goals: "We thought at that time that if we could survive for three to six months, that would be good. We dared to hope that we could finish the first prototype before the money ran out." Zhu Qiuguo recalled to "China Entrepreneur".
In fact, Deep Robotics needed less than a year to present its self-developed first four-legged robot. This was also the first four-legged robot in China that could climb stairs and navigate autonomously. Later, the Swiss Federal Institute of Technology Zurich bought a robot from Deep Robotics for research work and published an article in the renowned professional journal "Science Robotics", which was also selected as the monthly cover.
At that time, Zhu Qiuguo and his colleagues named the four-legged robot dog "Jueying". The inspiration came from the horse "Jueying" of Cao Cao from the novel "Romance of the Three Kingdoms". The horse was known for its unparalleled speed and elegance on the battlefield. The name is supposed to symbolize that the robot is as fast as a shadow.
In 2018, Deep Robotics received an external investment of several million yuan thanks to its first robot dog. However, Zhu Qiuguo's worries and pressure did not decrease. "After the investment, we thought that we could survive for another year. As Jack Ma once said, you think every day that the company could go bankrupt soon." Zhu Qiuguo said.
At that time, the market for four-legged robots was still far from practical application, as the costs were high and the technology was not yet mature.
Zhu Qiuguo also did not see a clear direction for the applications of "Jueying". "At that time, the robot couldn't even climb curbs and slipped on curved surfaces. It was already a success if it just stood still."
Good fortune returned in 2020. Employees of the Southern Power Grid contacted Zhu Qiuguo and asked if a robot dog could be used for the inspection of substations. Zhu Qiuguo saw this as a "stroke of luck", but behind this trust was the years of work of Deep Robotics' team on the basic development of operating systems, communication architectures and navigation algorithms.
"The power grid is a very serious area. Even today, they won't buy another device just because one belongs to the 'Six Dragons from Hangzhou'. In the end, it depends on the performance of the product." Zhu Qiuguo told "China Entrepreneur".
Since then, Deep Robotics has defined the field of power grid inspection as its core market. The strict requirements in this area have also improved the product quality of Deep Robotics to ensure more robust motion control, better environmental perception and interference resistance as well as a more stable long-term function. Finally, Deep Robotics has achieved success not only in China but also in foreign markets such as Singapore.
After the strategy in the power grid area was successful, Zhu Qiuguo quickly expanded into the field of fire fighting and rescue. In this area, the robot dog can penetrate into dangerous environments such as toxic atmospheres, oxygen-poor rooms or unstable buildings instead of rescue workers to carry out exploration and rescue work. It can also penetrate into complex structures to explore objects, terrain and toxic gases and thus provide "vital information" for the fire department. In addition, it can transport rescue materials and equipment to support the rescue.
By 2025, the four-legged robot market has moved from technological testing and scenario research to mass commercial application. The market is growing rapidly and the application areas are constantly expanding. Deep Robotics' four-legged robots are now also being used in police stations, tunnel construction works, steel mills and even in the "last mile" of logistics.
The necessity of self-development
"It is very difficult to develop a stable robot conversation in the whole world." Zhu Qiuguo emphasized his cautious optimism in several interviews.
In the past eight years, the self-development of core technologies and the hard work on the supply chain have helped Deep Robotics to close the gap between the prototype and a reliable industrial product.
At the beginning of the founding, Zhu Qiuguo and his team tested countless versions of the robot operating system. The supply chain was the biggest challenge for Zhu Qiuguo and his team. "We had to wind the motors by hand. To increase the power density, we had to develop them ourselves." Zhu Qiuguo recalled to "China Entrepreneur".
At that time, there were no established suppliers and no mass production systems. For the key components of "Jueying", such as high-performance motors, precision gears and special controllers, the team and the suppliers had to learn step by step from the design, material selection to the manufacturing technology.
Jueying Lite3
This "forced" self-development has finally become a core competence of Deep Robotics. To date, Deep Robotics has more than 100 authorized patents, including more than 20 invention patents that cover the entire chain of Embodied-AI technologies such as hardware design, intelligent motion control, multi-modal environmental perception, autonomous navigation and decision-making.
An important milestone in the technology development of four-legged robots was the intensive application of the reinforcement learning algorithm in motion control. "This has significantly improved the stability, flexibility and adaptability of the robot in complex, unstructured environments. This is the key for robots to be actually used." Zhu Qiuguo said.
Based on the new technology, Deep Robotics released the new robot "Bobcat M20" in April 2025, a wheel-legged robot for industrial applications in difficult terrains and dangerous environments. It combines the speed of wheeled robots with the flexibility of legged robots. It can work at temperatures between -20 °C and 55 °C, climb 80-cm-high steps and has autonomous perception, navigation and all-directional collision avoidance.
Bobcat M20
In addition, the Bobcat M20 can adapt its movement in extreme environments such as steep mountains, muddy ground or rubble fields and thus work in all areas.
In addition to the technological challenge, cost control is another key for mass production.
With the increasing attention of the state and political leadership to the humanoid and four-legged robot industry and the increasing market attractiveness, the supply chain for core components such as motors, drivers and sensors has developed rapidly. There are more options, the prices of robots are falling and the manufacturing technology is becoming more stable. Thanks to its leading position and mass production in several application areas, Deep Robotics has also reduced its costs.
This strong technological basis comes from the academic background and engineering competence of the founding team. Zhu Qiuguo studied mechanical and electronic engineering at Zhejiang University and thus laid the foundation for the hardware. In his master's and doctoral theses, he switched to the field of control technology and worked on the control of two-legged humanoid robots. He was also involved in the development of the two-legged robot series "Wukong" of Zhejiang University and experienced the entire process from the theoretical drawing to the functioning prototype.
In addition, the co-founder Dr. Li Chao and the core team of Deep Robotics come from top universities such as Zhejiang University, Peking University, Shanghai Jiao Tong University, Harbin Institute of Technology, University of Chinese Academy of Sciences as well as the Swiss Federal Institute of Technology Zurich and the Georgia Institute of Technology. The close cooperation between research, teaching and industry has continuously supplied Deep Robotics with technology and innovation power.