Exclusive Financing Announcement | Backed by two listed companies, Leading Optoelectronics and Woan Robotics, Zhenjue Wanxiang aims to build the "data foundation" for embodied intelligence.
【Shenzhen, August 2026】 The embodied intelligence track has welcomed a key infrastructure player. Physical interaction data infrastructure firm Physen AI announced today that it has closed a multi-million-yuan angel round of financing. This round is led by the Worth Star Torch Fund under Yingke Zhide, with follow-on investment from institutions including Shanjing No.1 Fund and Mingyao Huichuang.
What is particularly striking is that two listed companies, leading A-share optical enterprise Leading Optoelectronic (605118.SH) and Hong Kong-listed "No.1 AI embodied home robot stock" SwitchBot (06600.HK), have rarely joined hands to support this round through direct industrial investment. The raised funds will be mainly used for the mass production and delivery of three-ring data acquisition equipment, the construction of data algorithm platform and team expansion, to solve the core industry pain point that robots "can understand tasks but cannot perform them correctly".
"The 'Oil Field in the Robot Era': The Rise of the Independent Data Layer"
The data layer for autonomous driving is naturally built into automakers, but embodied intelligence does not have such inherent advantages. "The Internet era has spawned the massive text data required by large language models, while the Physical AI era requires full-modal interaction data between humans and the real world," said Huang Zhishen, founder of Physen AI.
Videos can only record "what has been done", but physical interaction signals such as contact, force application and slip can tell robots "how to do it". Physen AI judges that the embodied intelligence industrial chain will present a pattern where data, brain and ontology develop in parallel. For long-tail scenarios such as home services and medical health care, independent data infrastructure will become the core of the industrial chain.
Extreme Cost-Effective "Three-Ring Solution": Redefining the End State of Data Acquisition
Centering on this judgment, Physen AI has launched the "Three-Ring Solution", which is known as the end-state form of EGO data acquisition.
Different from the common motion capture gloves or teleoperation devices in the industry, this solution forms a complete perception closed loop through the collaboration of headband (global visual positioning), dual armbands (six-dimensional force matrix) and IMU (state monitoring).
What is more disruptive is its cost control and technical route:
Extreme cost-effectiveness: The total cost of the whole set of equipment is controlled at around 150 US dollars. With a "near-sensory-free" wearing mode that does not change people's natural operations, large-scale deployment becomes possible.
Technical generation gap: Adopting FMG (force myography signal) instead of the interference-prone EMG (electromyography signal), it covers traditional blind areas such as static holding and slow force application, with stronger robustness.
Efficiency leap: Relying on the exclusive data alignment model, Physen AI can efficiently transfer the prior knowledge of human operations to robots. Tests show that the physical machine data demand is reduced by more than 70%, and the generalization speed for new objects is increased by more than 3 times, greatly lowering the threshold for robot training.
Industrial Collaboration: Full Verification from Supply Chain to Scenarios
This round of financing is not only financial support, but also in-depth integration of industrial resources.
Liding Optoelectronics has more than 20 years of experience in the field of fisheye and panoramic optical lenses, and will form supply chain and joint R&D collaboration with Physen AI on optical modules to consolidate the hardware moat.
As a world's leading provider of AI embodied home robot systems, SwitchBot's products are sold in more than 90 countries and regions, and it is an important industrial investor and scenario verification entrance for Physen AI. Relying on the strategic linkage with the Baoan District Government, SwitchBot has deployed a large-scale physical robot data acquisition base in the Jubao Pen AI Ecological Innovation Community; the physical AI engine interaction data system of Physen AI will be deployed as the core acquisition solution of this center, and the two sides will jointly open up the link from home scenario interaction to physical signal annotation, providing a large-scale data source close to real applications for the algorithm iteration of Physen AI.
At present, Physen AI's solution has been recognized by leading companies in the industrial chain, and has achieved delivery of tens of millions of scale. Its partners include leading harmonic reducer firm Green Harmony (model verification), auto parts giant Ningbo Huaxiang (data backflow), and DER Co., Ltd. (task verification), covering key scenarios such as grasping, handling and assembly.
Full-Stack Team: From Tesla, Amazon to the Chinese Academy of Sciences
The core team of Physen AI spans the fields of embodied models, hardware mass production, optical engineering and robotics:
Founder Huang Zhishen, PhD in AI from the Chinese University of Hong Kong, has more than 10 years of experience in multi-modal motion perception;
Liang Yu, head of hardware products, PhD from the Hong Kong Polytechnic University, once led the R&D of consumer-grade quadruped and humanoid robot products, and set a record of completing the whole process from design to shipment in 4 months;
Wang Chen, optical and robotics engineering expert, PhD from Northwestern University, once served as senior optical engineer at Tesla, and led the sensor development of Astro, Amazon's first home robot;
Zhao Wentao, head of robot engineering, PhD from the Institute of Automation, Chinese Academy of Sciences;
Chen Lei, head of VLM and semantic understanding, PhD from Fudan University, member of Shanghai AI Laboratory.
In addition, Xu Songlin, assistant professor of the Hong Kong Polytechnic University, and Sheng Lü, professor of Beihang University (top 2% of global scientists), serve as chief scientist and technical consultant respectively.
Defining Physical AI Infrastructure
According to the plan, Physen AI will follow the three-step path of "Build (2026) → Scale (2027) → Ecosystem (2028)", and finally form a Physical Token data market and a cross-ontology data supply network.
"Data acquisition is undergoing a paradigm shift from motion capture gloves to wearable, near-sensory-free solutions," Huang Zhishen said. "Extreme flexible operation requires making up for the mechanical signals beyond vision without changing human's natural operations. Physen AI is committed to defining the next-generation Physical AI infrastructure, so that robots can truly have the 'sense of touch' to change the physical world."
About Physen AI:
Shenzhen Physen AI Technology Co., Ltd. is the world's leading physical interaction data infrastructure company. With the mission of "Let machines perceive all phenomena in the world", it provides underlying data support for embodied intelligence and world models through self-developed wearable data acquisition equipment, mechanical decoding algorithms and data alignment models.