36Kr Exclusive | Building "eyes" for pool robots, the chip startup founded by former Ecovacs employee has secured tens of millions of yuan in angel financing.
By｜ZHANG Ziyi
Edited by｜YUAN Silai
36Kr has learned that underwater sensing chip enterprise "Xinyang Micro" has recently completed a multi-million-yuan angel round of financing. The investors of this round are Keystone Ventures and Hidden Peak Capital, and the funds will be mainly used for team building and talent system development.
As underwater robots evolve from remote-controlled operation to autonomous operation, perception has become an unavoidable technical threshold.
According to data from Fortune Business Insights, the global underwater/marine robot market size reached approximately 5.82 billion U.S. dollars in 2025, and is expected to grow to 19.66 billion U.S. dollars by 2034. However, compared with land robots, the underwater environment is not friendly to common sensors: particles in water and turbidity will accelerate the attenuation of light and affect visual imaging; the propagation distance of laser in water is also limited. The NOAA's monitoring of seawater light attenuation also shows that the higher the turbidity of water, the lower the effective propagation depth of visible light.
Therefore, ultrasonic wave has become an important technical route for underwater detection.
Traditional underwater ultrasound mainly uses piezoelectric ceramic (PZT) transducers to generate and receive ultrasonic waves through the vibration of piezoelectric materials. Such devices can provide high acoustic power, so they are still suitable for applications such as long-distance sonar, but they have limitations in miniaturization, high-density array and imaging scenarios required by robots.
The technical stack of Xinyang Micro adopts the CMUT (Capacitive Micromachined Ultrasonic Transducer) route. It uses MEMS semiconductor technology to manufacture micro diaphragms on wafers, which can be arranged into high-density two-dimensional phased arrays while having a wide frequency band.
Existing studies have shown that CMUT arrays can realize underwater ultrasonic imaging and have broadband and array characteristics, which makes them more suitable for being installed on the robot body to complete near-field environment perception and 3D acoustic imaging.
CAO Jinlong, founder of Xinyang Micro, told 36Kr, "Traditional sonar is more suitable for long-distance detection at the kilometer level, and the company is currently targeting near-field perception within the range of about 0-20 meters around the robot."
What Xinyang Micro currently provides is not a single CMUT chip, but an acoustic perception module composed of CMUT array, ASIC integrated chip and related software and hardware. The first commercial application scenario selected by the company is pool cleaning robots.
According to data from CIC Consulting, there are about 34.1 million swimming pools worldwide in 2025, with an annual shipment of about 4.9 million pool cleaning robots, which is expected to increase to 7 million units by 2030; the global retail sales of pool robots in 2025 is about 2.8 billion U.S. dollars.
The pool robot market is developing rapidly, and many participants are Chinese enterprises.
According to CAO Jinlong, many current pool robots still mainly rely on gyroscopes, single-point ultrasound or vision-assisted positioning. The robot can move, but it is difficult to accurately know which areas have been cleaned and which have not.
Xinyang Micro hopes to change this situation with array acoustic imaging. According to the company's test data, after being equipped with its module, the cleaning coverage rate of the pool robot can reach more than 95%, and it can further realize area identification and path planning, making the pool cleaning robot more intelligent and controllable.
At present, Xinyang Micro has completed multiple rounds of tapeout and entered the small-batch production stage, and has carried out cooperation with many pool cleaning robot enterprises in product requirement research, testing and supply chain.
CAO Jinlong said that the swimming pool is the first stop the company actively chooses. Compared with medical ultrasound, it has lower requirements for imaging accuracy, certification and product introduction, which can verify chips and modules faster and generate early revenue.
After completing the layout in the swimming pool scenario, Xinyang Micro plans to continue expanding along underwater scenarios. The next step covers consumer-grade underwater navigation mapping and scanning robots, which can be applied to scenarios such as mobile mapping and scanning of fish schools, underwater equipment and underwater rescue, then expand to aquaculture, where it can obtain data such as the number of fish schools and their activity depth through acoustic perception, so as to provide information for feeding and aquaculture management. In the future, Xinyang Micro's CMUT technology will also be applied to the interaction between robots and the human brain, forming a wearable smart device that enables the human brain to directly control the robot.
In terms of the team, CAO Jinlong, founder and CEO, previously worked successively at ECOVACS and Honeywell, with work experience spanning consumer-grade robots and upstream sensors; the R&D team has more than 10 years of CMUT R&D experience, and has accumulated a variety of chip design and process technology expertise.
In the long run, Xinyang Micro hopes to extend the capability of CMUT from "perception" to "communication". Since ultrasonic devices can both transmit and receive sound waves, the company will eventually make it an information port for underwater robots - which is responsible for not only "seeing" the surrounding environment, but also exchanging information with the outside world.
CEO Dialogue
36Kr: Chips adopting the CMUT technology route were mostly used in the medical field before. Why did you choose pool robots first instead of starting with medical products with higher technical thresholds?
CAO Jinlong: In the early stage, we actually looked at vascular ultrasound, cardiac chamber ultrasound, craniocerebral ultrasound, and non-invasive brain-computer interfaces. However, these products require huge investment, and the R&D and commercial return cycles are very long. For an early-stage startup, if it enters a market with high investment and no short-term revenue at the very beginning, the risk will be extremely high.
Pool robots do not have as high requirements for imaging accuracy as medical products, and the cost of product introduction and trial and error is much lower. For us, it can form both a technical closed loop and a commercial closed loop. We choose the path of verifying products through a real market first to keep the company alive, accumulate customers and capital, and then expand to higher-end scenarios.
36Kr: CMUT technology has been developed overseas for many years. Where are the opportunities for Chinese companies to enter the underwater robot track now?
CAO Jinlong: China's biggest advantage lies in the development speed of the downstream robot industry. Many pool robots and underwater smart hardware are developed by Chinese teams and sold to the global market. We are very close to customers, so we can get real demands faster and then define chips according to the demands.
In the past, foreign CMUT companies were mainly targeted at the medical market, and they have formed relatively mature products and business models. However, medical chips cannot be directly used underwater, and the devices and ASICs need to be redesigned. On the contrary, we can define products around underwater scenarios from the very beginning.
Of course, China still has shortcomings at present, mainly in the chip manufacturing process. Our current production volume is not large enough, and the available foundry resources are limited, so we are also solving the mass production problem through pilot test platforms and new manufacturing partners.