Yu Yang, from Afei Temple
QbitAI | Official Account
Ren Shaoqing, the author of ResNet, has personally entered the fray of embodied intelligence entrepreneurship.
QbitAI has learned and confirmed that Ren Shaoqing, a well-known AI scientist, the author of ResNet, the director of the General AI Research Institute at USTC, as well as NIO’s senior vice president and head of intelligent driving, and also the chief director of the 2023 Future Science Award, Dr. Ren Shaoqing, has officially established a company and embarked on a venture in embodied-intelligence robots.
According to a report from LatePost, at an all-staff meeting for NIO’s intelligent driving business, NIO CEO Li Bin informed employees internally that Ren Shaoqing, the head of NIO Intelligent Driving, has founded an independent company focused on physical AI foundational models and embodied intelligence.
NIO has clearly made a strategic investment and carried out business cooperation. The news also says that Ren Shaoqing’s new company has already completed registration, with a valuation at the unicorn level (more than $1 billion), but the company name and the specific details of the financing are still undisclosed.
Notably, Ren Shaoqing’s current startup model for embodied intelligence robots is not yet fully clear—
Although he started a robotics venture, he is still employed at NIO.
How should I put it? “Divorce but don’t move out” (manual doge head).
Ren Shaoqing dives into embodied intelligence
Regarding Ren Shaoqing’s robotics startup this time, people close to NIO’s management said:
For years, NIO has kept a close eye on cutting-edge AI technologies and has laid out its industrial strategy in a way that balances independent R&D with open collaboration.
In recent years, NIO has continued to invest in key technologies such as chips and operating systems, achieving major breakthroughs. At the same time, through strategic partnerships and industry collaboration, it has continued to support innovative enterprises around directions including embodied intelligence, algorithm models, intelligent hardware, new energy, and compute infrastructure.
Further information revealed that internally, NIO believes that:
Intelligent driving is one of the most mature large-scale deployment scenarios for AI in the physical world. It has continuity with embodied intelligence in foundational capabilities such as perception, prediction, planning, control, world models, and reinforcement learning. Over the past two years, more and more talents in autonomous and intelligent driving have flowed into the robotics industry, which is also related to this kind of technology transfer.
But it is worth noting that NIO has blurred the “change” related to Ren Shaoqing this time.
Considering NIO’s business and development situation, perhaps one can glimpse the strategy behind this subtle and ambiguous backdrop.
On the one hand, as a world-class AI scientist and the absolute driving force behind NIO’s intelligent driving transforming for the better, if Ren Shaoqing were to leave and start a business, it would not be good news for NIO—which is already in a strong momentum phase. On the other hand, if this is part of NIO’s incubation or internal innovation business, the current embodied intelligence robotics sector’s characteristics—high investment and long cycle—would also add pressure on NIO, which has just completed a financial turnaround.
What connects the dots is that Xiaopeng, the earliest company betting on embodied intelligence robots, just before the release of its robot quarterly financial results, announced its first round of independent external financing—an unmistakable statement that robot finances would be independent.
As for NIO, although its trial run of embodied intelligence robots isn’t early, its internal thinking and external statements have long been clear.
As NIO faced pressure on financial performance in 2025, both Li Bin and Ren Shaoqing clarified separately on Quantum Black that the next chapter of AI + the automotive business is for NIO to lean toward robots, not Robotaxi.
Now that Ren Shaoqing has entered the robotics space, it may be seen as NIO’s first foray into embodied intelligence robots.
Previously, in the AI frontier area, NIO had already launched internal R&D of world models in the second half of 2023.
In July 2024, it was Ren Shaoqing—representing NIO—who released NIO World Model (NWM) at the Tech Day event. Starting in 2025, the model began mass production and deployment, and this year it was upgraded to “World Model + supervised fine-tuning + closed-loop reinforcement learning.”
And Ren Shaoqing himself has repeatedly emphasized that world models are the foundational paradigm shared by autonomous driving and robots.
The spatiotemporal understanding built on the vehicle platform, long-term sequential reasoning, and closed-loop optimization capabilities can be directly transferred to physical AI and embodied intelligence scenarios. This is also one of the technical foundations behind the embodied intelligence companies he recently founded.
In fact, Ren Shaoqing is among the earliest in the industry to bet on world models—a hands-on practitioner—and this is also a key reason behind NIO’s overtake in intelligent driving on the straightaway. And world models are becoming a new converging direction for robots’ “brains.”
NIO’s own business scenarios and supply-chain manufacturing are also naturally suitable for deploying robots in closed-loop systems.
And those who know Ren Shaoqing may also know that although, after earning his PhD, he became a direct entrepreneur as a co-founder of Momenta, his personality and strengths lean more toward the role of a chief scientist or CTO…
World-class AI scientist Ren Shaoqing
Ren Shaoqing was born in 1988 in Bengbu, Anhui. In 2007, he graduated from Bengbu No. 2 High School and then entered the University of Science and Technology of China, majoring in information security.
In 2016, under the guidance of Sun Jian, one of the authors of ResNet, Ren Shaoqing obtained a joint-cultivation PhD degree from the University of Science and Technology of China and Microsoft Research Asia.
And he himself is also one of the four authors of ResNet.
ResNet is a milestone in the field of deep learning. Its achievement lies in fundamentally solving the problem of vanishing gradients in training deep neural networks. It makes building high-performance, deep neural networks practically feasible, and has made a tremendous contribution—especially in fields such as computer vision and autonomous driving, where perception capabilities are crucial.
This research won the CVPR 2016 Best Paper Award, and the number of citations has already exceeded 340,000.
Core authors: Ke Ming Why, Zhang Xiangyu, Ren Shaoqing, and Sun Jian.
He was also the first author of Faster R-CNN. This is an efficient two-stage target detection algorithm that successfully addresses the previously time-consuming problem of generating candidate regions in target detection algorithms, striking a good balance between speed and accuracy. It has been widely used in autonomous driving, security, and medical image analysis.
After graduating with his PhD, Ren Shaoqing joined the autonomous driving startup Momenta, serving as R&D Director and co-founder.
In 2020, he officially joined NIO. Initially, he served as Chief Expert for autonomous driving R&D and Assistant Vice President, mainly responsible for the autonomous driving algorithms team; later, he was reassigned as Vice President of Intelligent Driving R&D, reporting directly to CEO Li Bin. By the end of 2024, NIO’s intelligent driving R&D department underwent a structural reorganization: a new Technical Committee was established. Ren Shaoqing was put in direct charge of this team and led the Large Model Department, with his scope of authority further expanded.
After Ren Shaoqing took over NIO’s intelligent driving R&D system, he was the first globally to put “world models” into vehicles, and combined them with NIO’s self-developed chips, enabling NIO’s intelligent driving to overtake on the straightaway. To a certain extent, he became one of the key contributors to NIO’s first profitable turnaround in 2025–2026.
After the success of the world model, Ren Shaoqing was promoted to Senior Vice President at NIO.
Meanwhile, Ren Shaoqing also began giving back to his alma mater, USTC. In September 2025, Ren Shaoqing returned to USTC to host enrollment classes, serving as a distinguished professor, doctoral supervisor, and director of the Institute of General Artificial Intelligence Research. Even earlier, Ren Shaoqing donated all the prize money from the Future Science Award to USTC.
On the academic official website page of Professor Ren Shaoqing at USTC, the list of his main research fields on his homepage already includes “embodied intelligence,” specifically: artificial intelligence, world models, embodied intelligence, autonomous driving, computer vision, deep learning, and AI for Science.
Of course, even more striking is the proof of Ren Shaoqing’s achievements:
As of March 2026, his academic papers have been cited more than 490,000 times, ranking first among scholars in China and in the top three among Chinese people worldwide.
Awards include: the 2023 Future Science Award—Mathematics and Computer Science Award; the 2025 Wu Wenjun AI Science and Technology Progress First Prize; the CVPR Best Paper Award; the 2025 NeurIPS Test of Time Award; the 2025 Helmholtz Prize (ICCV 10-Year Impact Award); the Software-Defined Vehicle Innovator Award; global champion in the ImageNet and MSCOCO competitions; and third overall on the AI 2000 global AI scholars ranking list.
Only worship~