How crazy is the embodied intelligence track in 2026? Some companies raised ¥2 billion or ¥3 billion in a single round, with funding figures becoming increasingly exaggerated. A few days ago, IT Juzi released the latest 'China Embodied Intelligence Entrepreneur Ecosystem Portrait Report', which garnered significant attention from many readers.
In this wave of enthusiasm, we have noticed another prominent force: the academic school. How do we define the 'academic school'?
It's not the typical 'universities supplying graduate talent' scenario, but rather universities personally founding/incubating embodied intelligence companies.
Can you imagine it? Professors are no longer just burying themselves in research in labs and publishing papers in top-tier journals. They are directly registering companies, becoming founders/co-founders, and even going out to secure funding and manage production lines. Moreover, this is no longer an isolated phenomenon.
According to IT Juzi statistics, within the embodied intelligence circle, there are over 70 companies where the founding team currently holds or previously held professor positions in universities. However, the 'academic school' filter criteria I want to discuss here are stricter:
The first category is: Professors serving as founders are still employed by universities while simultaneously starting businesses, and they can mobilize resources and support from their alma mater. It is even more typical if the university holds direct equity. Companies that merely invite university professors as advisors or chief scientists are not included.
The second category is: Embodied intelligence companies that are themselves the product of the platformization of universities/research institutes and laboratories. This is another representative of the academic school.
Therefore, based on this standard, I have compiled a list of 18 'academic school' embodied intelligence companies, presented in the table below.
How capable are these 'university professor-led + research institute incubated' companies?
The financing data we have compiled shows that: In the first 8 months of 2026, the total disclosed financing of these 18 'academic school' embodied intelligence startups amounts to approximately ¥15 billion (estimated value), with over ¥13 billion explicitly disclosed.
It can be said that they are holding high the banner of the 'academic school' in embodied intelligence and have become a core force within this track.
Now, we can simply divide these companies into two major categories. The first category consists of companies whose core founders/co-founders are affiliated with prestigious universities, covering top-ranked 985/211 comprehensive universities like Tsinghua University, Peking University, Zhejiang University, Shanghai Jiao Tong University, Fudan University, etc.
The second category comprises companies directly incubated by universities/research institutes, such as Westlake University, Tsinghua AIR, Zhiyuan Research Institute, Chinese Academy of Sciences, etc. It's clearer to look at them separately.
When 985 University Professors Personally Take Command and Dive into the Fray
The number of companies where 985 professors personally venture into entrepreneurship for embodied intelligence is not vast, but they naturally attract the most attention. Founders often hold professor titles and are driven by the motivation to commercialize academic research results.
Tsinghua University-invested Xingdong Jiyuan (Star Motion Era) is a typical example, focusing on humanoid robot hardware. Founder Chen Jianyu is an assistant professor at Tsinghua University's Institute for Interdisciplinary Information Sciences (IIIS) and holds a Ph.D. from UC Berkeley. The company was founded in August 2023, starting as an incubation project from IIIS. In July 2026, it completed a ¥1 billion Series B financing, growing into a unicorn in just over two years.
Another in the Tsinghua University camp is Accelerated Evolution, also working on humanoid robot hardware.
Founded in June 2023, the founding team hails from Professor Zhao Mingguo's humanoid robot team in Tsinghua University's Department of Automation, with CEO Cheng Hao responsible for commercialization. In April 2026, Accelerated Evolution completed a ¥1 billion Series B financing, also entering the unicorn ranks.
The representative embodied intelligence company from Peking University is Galaxy General, pursuing an integrated route of 'embodied multimodal large model + general humanoid robot hardware'. Founder Wang He is concurrently an assistant professor at the Frontiers of Computing Research Center, School of Computer Science, Peking University.
The company was founded in May 2023, received ¥1.1 billion financing led by CATL in June 2025, and completed a ¥2.5 billion B+ round in March 2026. Last year, Galaxy General launched a smart retail store in Haidian, Beijing, where humanoid robots 'work'.
The representative company founded by a Zhejiang University professor is CloudWalk Deep, focusing on quadruped and humanoid robot hardware, with strengths in power inspection and emergency rescue. Founder Zhu Qiuguo is an associate professor and doctoral supervisor at Zhejiang University. This company was established earlier, starting in Hangzhou in 2017, and is one of the 'Hangzhou Six Dragons'.
In July 2025, CloudWalk Deep completed nearly ¥500 million in financing, co-led by Fortune Capital and Guoxin Fund, followed by a Pre-IPO round in December. Among these professor-founded companies, CloudWalk Deep is the earliest to achieve commercialization.
The representative company from Shanghai Jiao Tong University is Qiongche Intelligence, which does not manufacture robot hardware but specializes in the algorithm layer of the embodied brain.
Founder Lu Cewu is the vice dean of the School of Artificial Intelligence, Shanghai Jiao Tong University, and a Changjiang Scholar. Founded in November 2023, the company developed the self-developed Noematrix Brain. In June 2026, Qiongche Intelligence completed several hundred million yuan in financing, and notably, this round's investors included the 'Shanghai Jiao Tong University AI Future Fund' and a wholly-owned subsidiary of Shanghai Chuangzhi College.
The representative from Southern University of Science and Technology (SUSTech) is Zhuiji Dynamics, developing general-purpose humanoid robot hardware.
Founder Zhang Wei is a tenured professor at SUSTech, previously a tenured professor at The Ohio State University in the US, with over ten years of research in humanoid robots. Founded in 2022, the company completed a $200 million Series B in February 2026 and a nearly $200 million Pre-IPO round in July, reaching a valuation of ¥15 billion. Its shareholder list includes Alibaba, SAIC Motor, Nio, JD.com, etc.
More notably, nearly 70% of this round's funding came from overseas institutions, with Middle Eastern and European capital betting on China's humanoid robot sector.
The representative embodied intelligence company incubated by Fudan University is Moushen Intelligence, working on generative and lightweight general embodied brains, belonging to the algorithm layer. Co-founder Chen Tao is a professor at Fudan University and director of the Deep Learning Laboratory. Founded in January 2025, the company just completed a ¥500 million Pre-A+ round in August 2026.
From Moushen Intelligence and Qiongche Intelligence to Deta Intelligence and Xingyuan Intelligence, they are essentially competing for the ecological niche of the 'embodied brain', indicating that the competitive focus in embodied intelligence is shifting from 'hardware' to 'brain'.
The representative company from Xi'an Jiaotong University is Youai Zhihe, manufacturing industrial mobile robot hardware (wheeled, non-humanoid), focusing on factory logistics and intelligent inspection. Founding members like Zhang Chaohui are from Xi'an Jiaotong University, with their supervisor Mei Xuesong being a senior scholar in mechanical engineering.
Founded in 2017, its products are applied in logistics handling for semiconductor, new energy, and 3C factories, as well as inspection scenarios in pipelines, power distribution rooms, and substations. Youai Zhihe raised ¥150 million in Series C and is recognized as a 'Little Giant' specialized and sophisticated enterprise by MIIT. This company's approach is more pragmatic—make money first, then discuss form factors.
The representative company from Harbin Institute of Technology (HIT) is Leju, developing large humanoid robot hardware. Founder Leng Xiaokun has a background at HIT. Founded in 2016, it is one of China's earliest companies in large humanoid robots. The core team is distributed in Shenzhen and Harbin, with members all being HIT masters and Ph.D.s in computer science, mechanics, and mathematics.
In October 2025, Leju Robotics completed a ¥1.5 billion Pre-IPO round, jointly invested by over ten institutions including Shenzhen Investment Holdings Capital, Longhua Capital, and CITIC Capital, with cumulative financing exceeding ¥1.7 billion. Leju's 'Kuafu' series has established ecosystems with over 40 enterprises like Huawei, Tencent, China Mobile, and FAW Group. Leju's characteristic is prioritizing mass production—while peers were still unveiling prototypes, Leju had already begun small-batch deliveries.
Additionally, several other companies are often categorized or self-described as 'university-incubated' but don't strictly qualify. For example, Qianxing Technology was once affiliated with Zhejiang University but later exited; Mocet Intelligence is a project incubated by Tsinghua University's x-lab, but Tsinghua x-lab is an 'on-campus entrepreneurship incubator' under the School of Economics and Management, more oriented towards students and alumni. After incubation, projects have no direct relationship with Tsinghua University.
New Research Institutes Are Batch-Incubating Startups
If the first category is professors starting businesses independently with university support, the second category is research institutions/institutes directly spinning off teams and platform achievements to establish registered companies. The financing enthusiasm for this type of company is equally high.
Tsinghua University's Institute for Artificial Intelligence (AIR) incubated Qiuzhi Technology. AIR is headed by Dean Zhang Yaqin.
Qiuzhi Technology develops mobile lifting dual-arm robot hardware (non-humanoid), focusing on scientific research training and commercial service scenarios. In 2026, Qiuzhi Technology completed an over $100 million angel round, jointly funded by eight institutions.
Founder Zhou Guyue holds a Ph.D. from HKUST, was a student of Li Zexiang, and a core early member of DJI. He joined Tsinghua AIR as an associate professor in 2020 and industrialized his laboratory achievements in 2023.
Then, Beijing Institute for General Artificial Intelligence (BIGAI) incubated Deta Intelligence.
Deta Intelligence's positioning is specializing in embodied intelligence foundational models and data infrastructure. The company released a head-mounted full-body data collection device, Delta D1. Users can normally perform tasks wearing the device to collect motion data.
Founded in January 2026, Deta Intelligence secured nearly ¥500 million in an Angel++ round by July, completing six rounds of financing within half a year. Founder Ma Xiaojian holds a B.S. from Tsinghua University in Computer Science and a Ph.D. from UCLA. He previously worked at Google Robotics and NVIDIA Research and later served as a researcher at BIGAI.
Beijing Academy of Artificial Intelligence (referred to as: Zhiyuan Research Institute, a national-level platform in China's AI large model field) incubated Xingyuan Intelligence.
Xingyuan Intelligence focuses on cross-platform deployable embodied brain models and integrated software-hardware architecture (brain + high-computing power edge hardware). Founded in August 2025, it raised ¥1 billion within 10 months: including a ¥200 million angel round, an over ¥100 million angel+ round, and a Pre-A round. Currently, Xingyuan Intelligence's integrated robot solution covers over 70% of leading embodied intelligence companies and is also the global largest supplier of NVIDIA's Jetson Thor.
Tsinghua University, in conjunction with the Jianghuai Frontier Technology Collaborative Innovation Center, also co-incubated Ling Cifang.
Ling Cifang focuses on generalized grasping algorithms and also introduces wheeled humanoid robot hardware, following a combined 'algorithm + hardware' route. Registered in Hefei, it completed over ¥100 million in financing in April 2026, led by Runze Group. Its generalized grasping technology doesn't require custom programs for specific items; robots can autonomously determine grasping methods for unfamiliar objects. Its ZERITH-H1 wheeled humanoid robot was exported to Malaysia as complete units in 2026.
The representative from Beijing Institute of Technology is Ligong Huahui, developing humanoid robot hardware and core components (joints, actuators, etc.). Founded in 2015, it completed a nearly ¥100 million Pre-A round in May 2026. Beijing Institute of Technology has deep expertise in biomimetic and special-purpose robots. Ligong Huahui is a typical 'professor + university-run industry' path.
A few others belong to institute incubation but are still in early stages: Zhangyu Technology incubated by Shenzhen Institute of Artificial Intelligence and Robotics (AIRS), and Deep Wit under the Zhongguancun Academy system.
It's worth noting that Westlake University has incubated 18 startups, including two major embodied intelligence companies—Westlake Interaction, incubated from Professor Jiang Hanqing's Cross Mechanics Laboratory at Westlake University (School of Engineering), mainly providing embodied intelligence solutions and having released a flexible variable-stiffness robotic arm; Westlake Robotics, focusing on autonomous general-purpose humanoid robot hardware and embodied brain R&D. Founder Wang Donglin is a tenured professor and associate dean of the Department of Artificial Intelligence at Westlake University.
There is also 'Zhongke Xinsong / Duko Robotics', incubated by the listed company 'Siasun Robot & Automation', not listed in the table.
In 2000, Shenyang Institute of Automation, Chinese Academy of Sciences founded China's first robot enterprise, Siasun, named after Academician Jiang Xinsong, the 'father of Chinese robotics', which later listed on the A-share market, known as the 'first robot stock'. Siasun produces a full range of industrial robot hardware. In 2025, this veteran company officially entered the embodied intelligence track, launching 'Songyi' wheeled and 'Songxing' bipedal humanoid robots, and established an Embodied Intelligence Research Institute.
Zhongke Xinsong completed several hundred million yuan in an A+ round in July 2026. In April 2026, nearly a hundred Siasun heavy-load industrial robots entered Geely Automobile's welding main line. Reportedly, this marks the first batch application of domestic spot welding robots in automotive welding lines.
High Density of Tsinghua Affiliates, University Industry Funds Simultaneously Investing
Finally, within this 'academic school' system, we summarize a few observations:
First, the density of Tsinghua University's presence in the embodied intelligence field is extremely high.
Xingdong Jiyuan, Accelerated Evolution, Qiuzhi Technology, and Ling Cifang are all 'direct affiliates' of the Tsinghua system. Combined with the Tsinghua/Beida forces behind many research institutes, Tsinghua University alone almost supports half of China's embodied intelligence industrial chain.
The reason is not hard to understand: Tsinghua's internal 'university — research institute — company' transformation chain is very complete, with multiple mature frontier industry experimental platforms like IIIS, AIR, Zhiyuan Research Institute, Communication Research Institute laboratories, etc. Core technology, talent (professors and students), capital, and application scenarios are basically within the same network.
Second, university funds are beginning to directly enter the market.
Xingdong Jiyuan's shareholder list includes Hua Kong Technology Transfer Co., Ltd. (Tsinghua Holdings' technology transfer platform). Qiongche Intelligence's financing round featured the Shanghai Jiao Tong University AI Future Fund. Zhiyuan Research Institute participated in Xingyuan Intelligence's financing.
In the past, the relationship between universities and companies was more like 'licensing/transferring paper and patent technology', with technology element flow as the core. Now it has transformed into a deeply equity-bound, incubating relationship. From joint laboratory technical support to capital injection, elements like technology, capital, and talent are being comprehensively connected. Alma maters giving 'entrepreneurship red packets' to companies founded by professors was unimaginable many years ago.
Third, the 'professor concentration' in this entrepreneurship wave is unprecedented.
While the protagonists of the previous AI entrepreneurship wave were tech-savvy prodigies, this time the protagonists are professors still within the establishment system.
The technological barriers in the large model era lie in data and computing power, which can be tackled with engineering prowess; but the barriers in embodied intelligence lie in the physical world, in fields like mechanics, control, and materials that require over a decade of accumulation. The advantages of professors are thus magnified.
When the 'academic school' becomes a label, it also has apparent shortcomings.
Academic thinking and business pace are completely different. Many professor-led teams possess first-class technology, but mass production, supply chains, overseas channels—these tangible aspects are the real lifelines. Fortunately, leading companies have recognized this and are addressing these gaps through complementary talent team building.
We believe embodied intelligence will be an upgrade of China's industry-university-research transformation model. The old model was 'publish papers, sell patents, professor continues teaching'; now it's 'professor steps into the fray, research institute incubates, independent company founded, alma mater invests'.
As professors well-versed in the physical world are now researching to solve physical AI, the development path of embodied intelligence will become more scientific. This list will grow longer in the future. And the next professor to venture into entrepreneurship might be in the university lab next door to you.
This article is from WeChat public account 'IT Juzi' (ID: itjuzi521), author: Wu Meimei
