Tsinghua alumni develop the "eyes" for robots: Annual revenue reaches 433 million yuan, ranking first globally and listed on the Hong Kong Stock Exchange
Does the floor-cleaning robot in your home have a small, constantly rotating "head" that protrudes like a tower? That's its "eye" for observing the world - LiDAR (Light Detection and Ranging).
This device enables the robot to accurately identify the position of the sofa and the width of the table legs, and even avoid a kitten dozing in the corner.
Today, the company that supplies this "eye" to more than half of the world's floor-cleaning robots - Shenzhen HC Technology Co., Ltd. (hereinafter referred to as "HC Technology") - is gearing up for a Hong Kong IPO.
According to data from CIC, in terms of revenue, HC Technology ranked first in the global space perception solution market for floor-cleaning robots in 2024. In the LiDAR product field, its shipments reached approximately 8 million units, accounting for over 50% of the global market share, maintaining its leading position.
Since its establishment, HC Technology has received investments from several well-known institutions, including Oriental Fortune Capital, Siyang Capital, Yicun Capital, Chenhui Venture Capital, AVIC International, Qianhai Mother Fund, and Cornerstone Capital. After the latest round of financing before the IPO, its valuation reached 3.04 billion yuan.
01
Zhou Kun, the helmsman of HC Technology, was born in 1979.
In 2003, he graduated from the Department of Automation at Tsinghua University with a postgraduate degree, specializing in computer vision and image analysis under the guidance of Academician Dai Qionghai of the Chinese Academy of Engineering. At that time, he was already fascinated by visual space positioning technology. After graduation, he went to the Bell Labs in Chicago, where he gained practical experience in human-computer interaction.
After returning to China, he worked at companies such as China Mobile Zhuowang Technology and Taishan Online, delving into visual algorithms for more than a decade and accumulating more than twenty national invention patents.
During that time, he was like a quiet explorer, always pondering how to enable machines to "understand" the world. However, he found that technology implementation in large companies was often slow - this restlessness became the starting point for his future entrepreneurship.
In 2013, Zhou Kun and Wang Jian, an alumnus from Xi'an Jiaotong University, founded HC Technology, initially targeting the somatosensory interaction market in the TV industry.
Qvod was impressed by their technology at that time. It not only invested but also provided office space, hoping to combine HC's cameras for remote control and somatosensory games.
However, three months later, Qvod was investigated, and HC Technology suddenly lost its customers and funds. Paying the salaries of more than a dozen employees became a problem.
In a critical moment, the team managed to secure an order from Skyworth, developed cameras for its TVs, and achieved mass production within three months, thus stabilizing the situation.
After that, the company briefly shifted to the VR field, developing a positioning system with 10% higher accuracy than Oculus, which attracted potential partners such as Tencent and Huawei.
However, at the 2017 CES exhibition, the absence of VR giants made Zhou Kun realize that without an ecological leader, it would be difficult for this market to grow. HC Technology had to find another track.
02
The turning point came in April of that year.
The investment department of Ecovacs asked him a question: "Can your technology be used to make LiDAR for floor-cleaning robots?" This question caught Zhou Kun's attention. After more than five months of research, he found that although floor-cleaning robots had been developed for a decade, they were not well-received due to inaccurate indoor positioning and poor user experience.
There are mainly two approaches in the market: LiDAR offers high accuracy but is expensive, while visible light navigation is cheap but inaccurate.
Zhou Kun judged that the key lies in whether the cost of LiDAR can be reduced in the next three to four years. In terms of mass production consistency and reliability, HC's long-accumulated monocular space positioning technology happens to meet the requirements of "short-range, high-precision".
"This is neither a blue ocean that no one notices nor a red ocean with fierce price competition. It's a 'pink market' suitable for entrepreneurs to break through," he later summarized.
In September 2017, HC Technology officially launched the LiDAR project for floor-cleaning robots. After that, the team immersed themselves in research and development, creating cost-effective products with self-developed algorithms and chips.
Facts have proved that this move was right - HC's LiDAR quickly became the industry benchmark, enabling robots to "shuttle freely" at home.
The following year, the company launched its first self-developed ASIC chip, C1, replacing the industrial linear array solution with a consumer-grade area array sensor, reducing the cost by 40% - 50%. Its first product, Camsense® X1, was mass-produced that year, solving the industry's pain points of "high cost and low accuracy" at once.
Relying on the combination of "self-developed chips + core algorithms", HC Technology quickly became the core supplier for leading brands such as Xiaomi, Roborock, and Ecovacs. By 2024, its LiDAR shipments reached 8 million units, with a market share of over 50%, firmly ranking first globally.
While deeply involved in the robot track, HC Technology has not given up its initial VR/AR dream.
The team continues to invest in the research and development of space positioning technology for VR controllers, collaborating with international giants such as Google, Meta, and HTC. In his view, although the VR industry has temporarily entered a trough, this long-distance race is not only about innovation but also about "foresight and will".
In addition to floor-cleaning robots, HC Technology has extended its technology to new fields such as lawn mowing robots and industrial inspection. Its partners include industry giants such as COMAC and Qualcomm.
With technological breakthroughs, capital has also followed. Since its establishment, HC Technology has completed multiple rounds of financing. The investors include Oriental Fortune Capital, Siyang Capital, Yicun Capital, Chenhui Venture Capital, AVIC International, and Roborock. After the latest round of financing before the IPO, the company's valuation reached 3.04 billion yuan.
03
The emergence of HC Technology has solved a core problem that has long restricted the popularization of robots - the contradiction between performance and cost.
In the past, the industry generally adopted a combined solution of "general-purpose chips + linear image sensors + EEL lasers".
Although this solution offers stable performance, it has many problems: linear sensors are expensive, general-purpose CPU or FPGA chips have insufficient computing power and limited reliability when processing high-speed point cloud data. Coupled with the bulky EEL lasers, the cost of the entire module remains high.
As a result, this type of perception system cannot be used for large-scale consumer-grade robots, and its commercialization space is extremely limited.
In 2018, HC Technology rewritten the industry rules with a "radical innovation".
They proposed a new technological approach: "self-developed ASIC AI chips + CMOS area array image sensors + VCSEL lasers".
They boldly replaced the expensive industrial linear array solution with the CMOS area array sensor, which is common in mobile phone cameras and has extremely low cost. Then they replaced the EEL lasers with the more compact and mature VCSEL lasers.
The key to enabling these "affordable components" to achieve high-precision perception is the self-developed ASIC AI chip. HC has directly embedded the core space perception algorithm inside the chip, fundamentally solving the shortcomings of insufficient computing power and low reliability of general-purpose chips.
This innovation bore fruit in 2020: the X2 LiDAR was launched. It is smaller in size and lower in cost, quickly becoming the mainstream solution in the floor-cleaning robot industry. It was this breakthrough that triggered the growth wave of the entire market in the following years.
Under this technological system, the most core and profitable product of HC Technology is the "triangulation LiDAR".
The financial report shows that this product almost accounts for all of the company's revenue:
In 2022, the revenue was 144 million yuan, accounting for 99%.
In 2023, it increased to 327 million yuan, accounting for 98%.
In 2024, it further rose to 408 million yuan, still accounting for 94%.
It was not until the first half of 2025, when the dTOF LiDAR and line laser sensors began to gain momentum, that the revenue share of the triangulation LiDAR dropped to 72.7%, but it still remains HC Technology's most stable "cash cow".
04
Today, four out of the world's top five floor-cleaning robot brands have chosen HC Technology as their core supplier - these customers together account for more than 60% of the global market share. On average, HC has been collaborating with them for more than five years.
Among the few global space perception solution suppliers with "self-developed chip capabilities", HC Technology ranks first in terms of scale.
This technological and customer barrier is directly reflected in the company's financial data.
Its total revenue increased from 146 million yuan in 2022 to 432 million yuan in 2024, with a compound annual growth rate of 72.5% over three years. This momentum continued in 2025 - in the first half of the year alone, the company achieved a revenue of 292 million yuan, a year-on-year increase of 63.1%.
However, behind this rapid growth is equally aggressive R & D investment.
Over the past three and a half years, the company has continuously maintained a "strategic loss": the net losses were 28.695 million yuan, 0.883 million yuan, 31.375 million yuan, and 4.157 million yuan respectively.
According to the prospectus, the main reason for the losses is the continuous high-intensity R & D. From 2022 to the first half of 2025, the company's cumulative R & D investment exceeded 198 million yuan.
05
The entire robot industry is moving from "task execution" to "embodied intelligence" - machines are no longer just "seeing" the world but "understanding" and "responding" to it.
Today's floor-cleaning robots can only avoid obstacles, with only geometric recognition ability. In the future, intelligent robots will be able to recognize that it is a slipper and put it back in the shoe cabinet - this means the integration of perception, cognition, and action abilities.
This intelligent transformation is giving rise to a new market.
CIC predicts that the global market for space perception solutions for intelligent robots will grow from 26 billion yuan in 2024 to 78.6 billion yuan in 2030, with a compound annual growth rate of 20.2%. The Chinese market will grow even faster, from 10.8 billion yuan to 35.7 billion yuan, with a compound annual growth rate of 22%.
In other words, the market growth of the "brain" and "eyes" is outpacing that of the "body" itself.
Future competition will no longer belong to single sensor manufacturers but to players who can provide complete system solutions. The three generally recognized trends in the industry are:
1. Sensor fusion will become the mainstream.
Future devices need to integrate multi-source data from vision, LiDAR, millimeter-wave radar, and inertial measurement units (IMUs) simultaneously to achieve a more stable understanding of the environment.
2. AI empowerment will drive "understanding the world".
The industry is moving from "geometric recognition" (identifying obstacles) to "semantic reasoning" (understanding the meaning of objects), which relies on stronger AI algorithms and large model capabilities.
3. The rise of hardware-software integration and edge AI.
As the complexity of algorithms increases, computing needs to be completed in real-time on the device side to reduce latency and improve reliability. This requires a deep integration of chip design and algorithm optimization to achieve low power consumption and miniaturization.
This article does not constitute any investment advice.
This article is from the WeChat official account "Pencil News" (ID: pencilnews), author: Huataishi. It is published by 36Kr with authorization.