—  Company news, media report, industry trend  —

Story behind the "Invisible Champion" in Robots Vacuum Industry

Introduction: If Christopher had not left the school, then Tsinghua University might have one more teacher, but for the global lidar industry chain, there would be one less leading company from China.

"We're going to be number one!".

"No. 1 in China or No. 1 in the world?" I asked.

"No. 1 in China is No. 1 in the world." Christopher replied with a smile.

When answering, Christopher's eyes were only firm and steady.

"The Restless" Christopher

When it comes to Shenzhen Camsense Technology Company (hereinafter referred to as "Camsense"), many people may not know it. But in the field of sweeping robots, Camsense is like HiSilicon in the chip industry. The lidar provided by Camsense effectively solves the high cost and low accuracy of the current lidar.

Walking into Camsense, I found that behind this lidar unicorn company, there is a "restless" leader - he could live a stable and comfortable life, but he chose to start his own business. He is Christopher, an entrepreneur who likes to challenge himself.

(CamsenseCEO Christopher | Image source OFweek)

"I've never been a security-seeker," Christopher said with a smile.

After completing undergraduate and master studies at Tsinghua University, Christopher worked for several years at Bell Labs in Chicago. Several years after returning to China, Christopher founded Camsense and was hired by Tsinghua University as a master tutor of Shenzhen Graduate School.

"Actually, I have always hoped that I could do something by starting a business. Staying in school as a teacher is really too 'quiet'." Christopher said, "Schools are also very competitive as teachers, but when it comes to enterprises, especially leading a company Development, the things we face are much more complex, and we need to mobilize all the capabilities of the individual.”

"As a teacher in a school, anyone will respect you. But in a business, you have to have good products and technologies to be respected," Christopher said.

During the interview, it was obvious that Christopher had the dual qualities of an ivory teacher and a corporate CEO. The concentration and love for scientific research, and the ambition and ambition for the enterprise are perfectly blended and folded in him, corresponding to each other.

We visited that, unlike other companies today, Camsense provides a full-chain model from product design, R&D to production and sales. This model requires great effort and cost for enterprise managers; but the benefits It is also very obvious - the quality of all products is completely controllable. As a component supplier, this model can effectively control production costs and provide downstream enterprises with technically reliable and low-cost products.

Creating a new operating model is also an unprecedented test and challenge for Christopher.

"But I like the challenge of life," Christopher said with a smile.

Today's Christopher has also increased many titles: founder and CEO of Camsense, member of the Master's Steering Committee of Tsinghua University Graduate School, master's tutor, high-level talent in Shenzhen, and leading talent in Nanshan District.

But for Christopher, the title is not important, the important thing is how to provide users with low-cost, technically reliable and stable lidar.

Three years to sharpen a "sword"

"In fact, before April 2017, we have been focusing on monocular spatial positioning technology, and have not landed on the sweeping robot." Christopher said.

For Christopher, to establish a company, it must consider the implementation of its basic technology. In the early stage, Camsense chose the TV industry as the technology landing field, and used monocular sensing cameras to do somatosensory games.

"In the first three years, we have actually been looking for the direction of the company's technology." Christopher said frankly.

This was a lucky time for Camsense, as giants cultivated the market and followers also got a piece of the pie. Christopher is well aware that the market bottleneck and characteristics of the TV market itself will not be the final market for monocular sensor cameras, but this market allows Camsense to not worry about funding issues. Sufficient funds to operate, and insist on focusing on the research and development of monocular spatial positioning technology. In the process of research and development, Camsense is also looking for new tracks.

Later, it was proved that Camsense's prediction was correct. In the field of TV somatosensory games, in addition to Microsoft, domestic companies such as Obi Zhongguang and many other companies have not been able to fully develop this market. And Microsoft canceled this division in the following years.

But for Camsense at this time, it has found a new market.

"It was a very accidental opportunity to enter the field of lidar." Christopher said with a smile.

April 2017 is a lucky date for Camsense - Camsense after that, will be rewritten at this moment.

"In April 2017, a partner asked us if we could make lidars for sweeping robots. After more than 5 months of research, we officially launched our efforts in this field!" Christopher said.

Due to its development in the TV field, Camsense has become well-known in the industry, and a word from a partner inspired Christopher to conduct a research on the cleaning robot industry, which lasted for nearly half a year.

Through investigation, Camsense found that the robot vacuum cleaner market has been developing for more than ten years, but the overall progress is very slow.

"We have contacted a lot of investment institutions, and most of them are not optimistic about the sweeping robot." Christopher said, "Most investors do not think that the sweeping robot will be a product that will be used in thousands of households in the future, because the user experience is too bad."

But it is undeniable that the rise of Xiaomi's sweeping robots has allowed Christopher to see the business opportunities - the problem that hinders the development of sweeping robots is not the essence of sweeping robots, but the inability to achieve optimal intelligent functions such as indoor positioning.

But after observing the global sweeping robots, Christopher found that the technological evolution direction in the market has shown a bipolar division - the lidar solution represented by Xiaomi, and the visible light visual navigation solution represented by iRobot.

These two types of sensors face different markets and different treatments.

In 2017, lidar sensors are relatively cutting-edge technologies with high precision, but the cost is relatively expensive; while visible light visual navigation benefits from AI algorithms, which are not only sought after by capital, but also relatively cheap.

How to choose, it was difficult for Christopher to make up his mind for a while.

"We can't just look at the present, but in the next 3-4 years, whether we can reduce the cost through our technology." A passage from Christopher made everyone focus on lidar.

For Christopher, it's not a fantasy.

In half a year's research, Christopher found that the cost of lidar is too high, in addition to technical reasons, more importantly, the consistency of mass production product quality, long-term reliability and other issues have not been well resolved. If a technical solution can be found that not only enables the full production of lidars and effectively reduces their costs, but also ensures the reliability and consistency of mass production of products, then lidars will inevitably become fully popularized.

The visible light visual navigation equipment is sought after by many companies, and the market price cannot be greatly reduced in the future - more importantly, the visible light visual navigation (VSLAM) technology, due to the limitation of the principle, makes it difficult to achieve the high precision of lidar.

"We wouldn't be where we are today if we didn't get the technology right three years ago," Christopher said.

After determining the technical direction, Camsense finally set up a project in September 2017 and began to develop lidar for sweepers.

Directly hit the pain point, based on the market

But in the process of research and development, Christopher found that the existing market lidar is subdivided into two technical directions.

The first type is TOF lidar, which can measure the distance of surrounding objects through TOF technology during operation, so as to control the travel route of the sweeping robot. However, when the radar is moving and ranging, the accuracy will deteriorate as the distance from the obstacle decreases. Moreover, in the home environment, since there are often highly reflective materials such as skirting lines around the walls, the multi-path interference problem is serious, which makes the short-range ranging may be seriously inaccurate. This is not the best technical path for household cleaning robots. .

Another type of monocular vision-based geometric ranging lidar technology is also applied to low-speed mobile robots. This radar has a unique advantage - the closer the distance, the higher the accuracy, and it is suitable for spatial positioning of 6-8 meters. It is suitable for use in sweeping robots.

"This lidar is very suitable for the monocular spatial positioning technology we have been developing," Christopher said.

Under the combination of these two factors, Camsense has carried out corresponding research and development for this lidar.

At this time, the robot vacuum cleaner market has also developed with the maturity of spatial positioning technology, and Camsense has also caught up with the "development peak" of this industry.

"For start-ups, the 'blue ocean' and 'red ocean' markets are not recommended for direct entry. The 'blue ocean' market is unpopular, and the development of enterprises is limited; while the 'red ocean' market is too competitive and not suitable for companies without capital participation. And the 'pink' market is the best place for startups to grow," Christopher said. "When we entered, it was the pink window."

But Camsense, which entered this market, is also facing a new round of challenges - price war.

In the field of consumer electronics, once the market has entered a mature stage, "price wars" cannot be avoided. How does Camsense reduce product prices while ensuring quality?

In order to reduce the price, Camsense is adjusted on two levels:

1. Use the consumer-grade area array Sensor to replace the industrial-grade linear array Sensor, and change the mobile phone and PC-side chip algorithm to make it suitable for lidar;

2. Independent research and development of ASIC chips suitable for lidar, which can effectively replace expensive traditional DSP chips.

Compared with the linear array Sensor, the area array Sensor solution has a large algorithm processing capacity and high requirements on the processor algorithm capability, but its fault tolerance is good, and it can adapt to harsh environmental tests such as high temperature, low temperature, drop, vibration, etc., and ensure the measurement accuracy. It can significantly reduce the repair rate of lidar, which is the most important reason why Camsense's products can catch up with the latecomers. In addition, the use of large-scale consumer-grade sensors and self-developed ASIC chips can also significantly reduce product costs.

Through a series of methods, Camsense's lidar has successfully reduced the price to the level of 100 yuan on the premise of ensuring performance and quality.

The "price war" in the market cannot shake Camsense's position in the field of sweeping robots.

In the development in recent years, the lidar sensor with high cost performance and stability created by Camsense has been recognized by Xiaomi ecological chain, Haier, 360 and other customers, and its market share has risen repeatedly.

Under the good market effect, Camsense received nearly 50 million yuan in A+ round of financing in November 2019.

The brand strength behind "10W+"

"The lidar market also has low and peak seasons, and the procurement cycle is very long." Christopher said calmly, "From the meeting and communication between the companies to the final batch procurement, the interval may be 8-12 months."

In such a long procurement cycle, users of Party A have plenty of choices: whether the product can run stably, the user experience in the past, the evaluation of network users, the word of mouth among friends and businessmen... This is Under the series of operations, the comprehensive strength of Party B has been effectively demonstrated. For start-ups located in Shenzhen, it is not easy to be able to engage in a certain segment for a long time and have a good reputation among many users.

And Camsense even more so.

But unlike other companies, Christopher values user word-of-mouth rather than media publicity. With reliable technical strength, Camsense achieved a monthly shipment of over 10W+ in September 2020, which is close to Camsense’s 2019 full-year shipments. Annual shipments!

Behind this is Camsense's commitment to customers: Camsense LiDAR has a lifespan of 1500 hours in harsh environments.

"Actually we can reach 2000 hours." Christopher said calmly.

Unlike other consumer products, LiDAR is an integrated product that integrates optics and electromechanics, and there is a risk of damage if it is bumped. Especially in the home environment, problems such as the flatness of the ground, the touch of pets/children, and low-slope obstacles will cause lidar to face various accidents.

To this end, Camsense continuously adjusts its internal structure, so that the lidar can still maintain normal operation in the harsh home environment.

It is precisely because of its reliable quality that Camsense's lidar has become a "sweet pastry" in the sweeping robot industry, and many companies have chosen Camsense. This also prompted the partners to place orders for more than 100,000 units directly with Camsense after the epidemic.

"Worry" by Christopher

The sudden increase in orders is something to celebrate for start-ups, but for Christopher, it will put great production pressure on the production line - although Christopher has already predicted that there will be a larger increase in 2020, But the huge shipments in an instant made Christopher feel the pressure.

The reason why Christopher felt stressed was the chain reaction caused by the new coronary pneumonia.

In 2019, the market has been improving all the way. Christopher predicts that there will be a new outbreak in 2020, and plans to develop another production line.

However, with the emergence of the new crown pneumonia, Camsense was hit the pause button.

Christopher is worried, but unlike other startup CEOs,


Story behind the "Invisible Champion" in Robots Vacuum Industry


Camsense:Being Committed in High-Precision Visual Sensor R&D


Related News

Reaching VR/AR again, "Camsense" completed nearly 100 million yuan of C1 round financing


Camsense's Short-range Positioning LiDAR Driving the Popularity of Service Robots


Camsense is Booming with the Industry