Hangjiashuo Robotics
Not long ago, Pacini announced the completion of a B++ funding round of hundreds of millions of CNY.
This tactile perception company, which has raised a cumulative total of over CNY 4 billion in half a year, is now retailing a dexterous hand online:
In JD.com's self-operated store, Pacini's DexH13 dexterous hand is priced at less than CNY 98,000. It features four fingers, 16 degrees of freedom, 1,140 tactile sensing units, 3,420 channels of tactile signals, and an 8-megapixel hand-eye camera.
Looking at this product link alongside the funding news reveals a subtle dislocation.
It Originally Sold Components
Founded in 2021, Pacini's founder, Jincheng Xu, comes from Sugano Shigeki's laboratory at Waseda University in Japan. The company took a detour initially—originally aiming to develop a complete teleoperation system solution, collecting tactile motion data through teleoperation to train models. After traveling the globe, the team scrapped the complete system and consolidated all resources into fundamental tactile perception.
In retrospect, this contraction was the right move. Over the past year, Pacini's self-developed tactile perception chips have achieved a cumulative shipment of over 1 million units. It has served more than 1,000 customers, with a repurchase rate of over 50%. A widely circulated evaluation in the industry states: the vast majority of domestic dexterous hand and end-effector manufacturers use its tactile hardware.
From the perspective of most peers, Pacini is a supplier, not a competitor. It does not stand in the foreground, but it appears in the bill of materials of almost all peers.
At the Same Time, It Is Also Selling Hands
Looking at Pacini's product matrix, this impression begins to waver: tactile sensors, dexterous hands DexH5 and DexH13, humanoid robots TORA-ONE and TORA-DOUBLE ONE, data acquisition devices PXCap III and PXDex III, and a 6-axis force sensor named PX6D, officially described as "the world's first Hall-effect-based 6-axis force sensor for embodied intelligence."
The DexH13 is clearly priced in JD.com's self-operated store. The slogan for the TORA series is "the industry's first humanoid robot equipped with a multi-dimensional tactile sensor." On the 6-axis force sensor side, the competitors are companies like Link Touch, Kunwei Sensing Technology, and Sunrise Instruments (SRI). The 6-axis force sensor is the most expensive single item in the complete system's perception system. With one sensor at each of the four joints, priced at approximately CNY 15,000 each, a single robot costs CNY 60,000 just for this item.
Thus, the question arises: a company that sells tactile hardware to almost all dexterous hand manufacturers is simultaneously selling dexterous hands; a company that supplies sensors to system integrators is simultaneously selling complete systems. The positions of supplier and customer are beginning to overlap.
Imperative, Not Just Expansion
It is hard to simply interpret this as "stealing business." A closer explanation is: the path of selling hardware alone is inherently a dead end.
Ma Yang, founder of Tashan Technology, once made a very straightforward remark. In the second quarter of 2025, his company sold nearly 10,000 tactile fingers, but he found that, "The hardware was bought, but I don't see anyone actually using it. I sold shovels, but the downstream can't use them. These 10,000 units are one-time purchases, which means my shovels are wasted."
Tactile sensors are not plug-and-play devices. Their value depends on whether downstream users know how to use them and whether there are supporting algorithms and scenarios. If customers just install the sensors into prototypes to shoot a video, this market will never grow.
Therefore, tactile sensor manufacturers have to take an extra step downstream: building a dexterous hand and a complete system themselves, using them to prove what tactile perception can truly deliver, and to collect data under real working conditions. This is not an impulse for expansion, but the cost of self-proving value.
The industry has already started viewing this as a phenomenon. In public reports, "component manufacturers extending downstream" is listed separately, with the list including ZHAOWEI Machinery, Choho Industrial, Jiecang Group, Pacini, Leadshine, and RoboSense.
The Cost Is Blurring Boundaries
For dexterous hand manufacturers, a core component supplier that simultaneously sells dexterous hands is something that needs to be re-evaluated: Should they develop tactile technology in-house? In-house development means giving up a supplier that has already reached a shipment volume of 1 million units and has the most mature batch consistency; not developing it in-house means that part of the product definition is held in someone else's hands.
Meanwhile, the industry's buffering mechanism has not yet been established. There is still no unified testing specification and data interface standard for domestic tactile sensors, and the relevant national standards have just initiated the project proposal. When customers receive sensors from different manufacturers, they often have to do a lot of secondary adaptation. Who will guarantee compatibility?
There is another layer of more intuitive confusion: Pacini claims to "account for half of the tactile perception market," while Tashan Technology claims a "market share of over 80% in the field of humanoid robot tactile sensors." Both claim over 80%, so the denominators are obviously not the same. It is clear that this industry does not yet have a universally recognized yardstick.
Returning to the DexH13 on sale, it may not exist solely to be sold.
What this round of financing truly changes may not be Pacini's valuation, but the originally clear boundary between components and complete systems. After raising CNY 4 billion, Pacini has the ability to stand on both sides simultaneously. And its customers must start thinking about which side they stand on.
This article is compiled based on publicly available network information, for reference only, and does not constitute investment advice.