Summary of Key Points
The humanoid robot industry is on the verge of an explosion (2024-2025 will be the period of significant growth), and dexterous hands are the key to determining whether robots can truly take up work roles (a factor often underestimated in competitive scenarios). Hangzhou Xino Future Technology has overcome the so-called “impossible triangle” of cost, performance, and reliability for dexterous hands through hybrid drive technology. With its full-stack self-development capabilities (from motors to algorithms), the company has become a focal point in the industry and has received investment of 1.5 billion yuan from investors such as Meituan and NIO Capital. Its Flex2 product has achieved a balance of lightweight design, high precision, and low heat generation. The commercialization path is clear: mass production will begin in the third quarter, with a capacity of 10,000 units by the end of the year. Investors generally regard dexterous hands as a crucial component in the field of embodied intelligence.
I. Dexterous Hands: The Final Barrier for Robots to Go from “Observers” to “Workers”
Have you ever wondered why robots can walk and talk but have not yet entered factories and homes? The answer lies in their hands—human tools (cups, screws, keyboards), production lines, and home environments are all designed for human hands. If a robot’s hands are insufficient, unable to handle fragile items or tighten parts, it can only serve as a display in an exhibition hall.
Dexterous hands are not just “executive organs” capable of grasping; they are also “sensory organs” that can detect the hardness and weight of objects. Without them, embodied intelligence (the physical embodiment of AGI) would remain at the stage of “understanding” (visual recognition) and unable to perform practical tasks. In the words of Xino’s founder, this breakthrough is as significant as the last industrial revolution—just as the steam engine replaced human labor, dexterous hands will enable robots to truly integrate into our world.
II. The Technical Dilemma: The “Impossible Triangle” of Dexterous Hands
What makes a useful dexterous hand so challenging? There is what’s known as the “impossible triangle” in the industry: it’s impossible to simultaneously achieve low cost, high performance, and high reliability.
The current mainstream approaches for electric dexterous hands fall into two categories:
- Pure rope-driven: Uses ropes to control the fingers, similar to operating a puppet. The advantages include being lightweight (the palm doesn’t weigh much), being resistant to drops, and having a long lifespan. However, the downside is lag in response time (commands take longer to be executed by the fingers), resulting in poor precision and an inability to grasp delicate objects.
- Pure direct-drive: Motors are installed directly in the palm for fast response and high precision. The problem is that more than 20 motors add up to a heavy palm (equivalent to carrying a brick), and the motors generate significant heat, causing them to shut down after 3-5 minutes, making them unsuitable for continuous 7×24-hour operation (e.g., on factory assembly lines).
Both approaches have their drawbacks, and the industry has been struggling to find a balance between performance and reliability.
III. Xino’s Solution: Hybrid Drive + Full Stack Self-Development
Xino Future Technology has taken a middle approach with hybrid drive technology, mimicking the human body’s structure:
- A “large motor” is installed in the forearm (similar to human forearm muscles) to exert force during gripping tasks (e.g., lifting heavy objects).
- A “small motor” in the palm (similar to small muscles in the human hand) controls fine movements of the fingers for precise operations (e.g., cracking an egg).
This approach solves the heat generation issue associated with direct drive motors (since the large motor is not in the palm) and overcomes the lag from rope-driven systems (the small motor in the palm provides direct control). Their Flex2 product demonstrates these advantages: it weighs only 400 grams (lighter than a smartphone), has a repeat positioning accuracy of 0.2 millimeters (twice the thickness of a human hair), and precise force control (able to gently pick up a piece of paper) without generating heat during continuous use.
Moreover, Xino is a “full-stack player” that develops and manufactures everything from motors, electronic controls, reducers, to algorithms. The founder explains this as a necessity due to the lack of ready-made components for embodied intelligence in the supply chain, forcing them to rely on their own capabilities. This allows them to integrate hardware, software, and data effectively (customers use the data collected by the dexterous hands to train AI models, which in turn optimize the products), creating a closed loop.
IV. Commercialization: Who Is Buying? How Much Can Be Produced?
Xino’s customers fall into three categories:
1. Robot body manufacturers: They need dexterous hands to enable their robots to perform tasks.
2. Data factories: Use the dexterous hands to collect data for training AI models.
3. Research institutions: Universities and laboratories for research purposes.
In terms of production capacity, mass production will start in the third quarter of this year, with plans to produce 10,000 dexterous hands and 200,000 micro-electric cylinders by the end of the year. The cost of dexterous hands accounts for 20%-25% of the total cost of humanoid robots (comparable to the cost of an automobile engine), making this a critical component with significant profit potential in the industry.
V. Why Are Investors So Optimistic? What Makes Dexterous Hands Such a Solid Investment?
The main reason investors are investing heavily in dexterous hands is the certainty they offer:
- Clear market trajectory: Embodied intelligence represents a major trend for the next few decades (similar to the rise of the internet), and dexterous hands are essential for turning robot concepts into reality.
- High value potential: Dexterous hands account for 20% of the cost of a robot, indicating significant revenue potential.
- High barriers to entry: The technology (hybrid drive, full-stack capabilities) is complex, and only a few companies can compete. The industry is already in a process of consolidation, with only 4-5 leading manufacturers expected to remain in China.
Short-term challenges include integrating large AI models with hardware (to enable the hands to make autonomous decisions, such as knowing to handle glass gently) and reducing costs (full-stack self-development can help achieve this). However, in the long run, dexterous hands are the strongest asset for China’s embodied intelligence industry. Once robots overcome this final hurdle, they will truly become part of our daily lives.
In summary: Dexterous hands are the “soul” of robots, and Xino’s technological breakthrough has opened the door to commercialization. Investors and the entire industry are betting that this development will enable robots to move from laboratories to production lines.