Sensor is the source of data acquisition, it is everywhere. The situation awareness required at the front end of intelligence basically starts with sensors. Whether it is intelligent manufacturing, smart city, smart medicine, intelligent equipment and big data analysis, and then a huge intelligent system, we should start from the tip of the sensor.

Leonardo da Vinci's surgical robot, a marvel in the field of medical devices, has more than 400 sensors; The famous Boston robot dog, which can jump freely, needs 1300 sensors.

Lizhu lixil, a famous toilet brand in Japan, is launching an intelligent toilet. An image sensor is installed on the back of the toilet cover, which can automatically identify the shape of feces. The whole toilet is automatically detected and connected to the cloud through more than 70 sensors, which can realize chronic disease health management. The factory cooperative robot assistant APAs launched by Bosch has built-in hundreds of sensors to quickly perceive people's state.

These amazing intelligent products actually have something in common.

The pace of digitization in this world can't leave a small sensor.

However, in China's strategic and supportive industrial territory, the location of sensors can hardly be found. When the new infrastructure is in full swing, the sensor, which is a vital support, is almost forgotten. The picture is like this. When all the bright guests want to enter the hall, their heels are clamped outside the door. Sooner or later, this embarrassing situation will be shouted out in pain.

Six monsters of amphibious species sensor

Widely used sensors, which belong to the industrial category of small to large, are the basis for other industries. But sensors are also a very unique industry. Many sensors have amphibious properties: on the one hand, the core of sensors is chips, which will follow Moore's law and have a rapidly evolving brain; On the other hand, it is also together with sensitive materials and mechanical devices, which is restricted by many mechanical theorems. This is a unique industry, so the sensor must be taken care of carefully in order to develop well. However, in China, sensors have become a surprising "six strange" industry.

The first step of the sensor:

The capacity is not small, but the domestic head players are very small. In 2019, the scale of China's sensor market will reach 170 billion yuan, and it is estimated that there are more than 1700 enterprises. In addition to Goethe and Ruisheng, relying on the strong shipments of Apple mobile phones, the volume has reached 10 billion, and they have occupied the territory in the field of acoustic sensors. Other fields, such as mobile phone, automobile, industrial control, wearable and Internet of things, are basically the markets of foreign brands. In areas other than consumer electronics and security, there are not many enterprises with an output value of more than 100 million. Only Zhengzhou Hanwei, Baoji Mike and Nanjing Gaohua are ahead. Other domestic sensor enterprises basically belong to potato club.

The second part of the sensor:

There are many kinds, but this market is very hidden. There are more than 30000 foreign molding products and types under research, and more than 20000 in China. These quantities may not be accurate, but sensors are undoubtedly a large category of products. This product is rarely known by people outside the industry. In fact, mobile phones, automobiles, industrial measurement and intelligent equipment are all large users of sensors. In recent years, smart manufacturing and industrial Internet are inseparable from small sensors. Of course, artificial intelligence is no exception. It can be said that no matter how fast AI runs, it still wears sensor shoes on its feet.

The third part of the sensor:

Civil products are most afraid of supply interruption, and military industries are not afraid of high prices. Military sensors have been highly autonomous, mainly due to the negligible cost of military procurement. If we want to compete in the civil market, we should not only fight for scale, but also have high cost performance. If the power consumption is lower and the cost is lower, the winner takes all. Therefore, the breakthrough in the civilian market is still very difficult, and it is impossible to obtain support from the military market. The two tunnels, each on one side, did not get through. The sensors of civil instruments are highly dependent on foreign countries. Yokogawa, Japan, has a joint venture with Chongqing Chuanyi to produce Yokogawa instruments. The resonant pressure sensor provided by Yokogawa, Japan, is the highest precision pressure sensor. Domestic key problems have not been solved. The joint venture can only rely on Japanese sensors.

The fourth part of the sensor:

Although made in China is famous for its cost, the cost advantage of sensors is not as obvious as that of foreign countries. At present, most of the sensors produced in China are low-end sensors. The import of medium and high-end sensors accounts for 80%, and the import of sensor chips accounts for more than 90%. China's production cost is also very high, and the income is only tens of millions. How can we be willing to invest tens of millions to build a production line? Now many sensor manufacturers still work alone and assemble a lot by hand. Because the output can't go up, some can only produce 5000 in one month, which can't talk about economies of scale at all. Bosch and OMRON have long set up factories in China, and the cost advantage is also huge.

Moreover, virtue day brand enterprises are eyeing China's sensor market and pay close attention to market share. As soon as China makes progress, it will be squeezed by the price reduction of foreign brands. In 2010, one switch of OMRON in Japan was close to 400 yuan, but now with the gradual rise of Chinese brands, it only needs 60 yuan. Flexible price reduction and firm defense of market share are common marketing means for foreign manufacturers. In this way, domestic brands have been suppressed near the sallow line, which is difficult to turn over.

The fifth part of the sensor:

The market is huge and financing is the most difficult. Originally, intelligent manufacturing and artificial intelligence were very hot, and the sensor should finally welcome the salted fish. But, No. This is a market that investors don't want to see. Due to the lack of domestic understanding of the importance of this industry, the investment community has been in a cold period. This has something to do with the concealment of products and the difficulty of becoming bigger and stronger. The state's "cold treatment" attitude towards this industry naturally also affects the judgment of investment funds.

The sixth part of the sensor:

This is an important weapon of the country, but it fell to the ground and became a little carrot head. As the first line of defense of perception, sensor is the key source of human society towards intelligence. However, this industry has not been paid attention to. In the early 1980s, in the subject research presided over by the State Science and Technology Commission, in the process of discussing which technologies are included in information technology, "sensor technology" caused great differences. But because the volume was too small, it was eventually cut off. In a flash, forty years have passed, and the situation has hardly changed. Although some policies to encourage the development of sensors have been introduced in recent two or three years, there is no strength and no funds, which is basically a cursory step.

Sensors are actually the five senses of interconnected things, eyes, ears and all kinds of touch. Although so important, no one pays attention to it. The six strange sensors are a strange thing in themselves. This is really a heart piercing thorn.

Amazing profits

In China, sensors are not easy to make money. Because the chip can not be independent, the investment in process R & D is huge, and the Red Sea competition is fierce, the profit of Chinese sensors has been pressed very low. According to the financial reports of 40 listed companies of domestic sensor enterprises, nearly 40% of enterprises have a profit margin of less than 5%; There are six companies with negative profits.

It is said that manufacturing profits are low, and sensors seem to be one of them. However, not making money is not the real situation of this industry.

Japan's Kearns sensor company can be said to be the most profitable company in Japan. In 2019, the turnover was close to 36 billion yuan, while the profit reached an amazing 18 billion yuan. The profit margin is more than 50%, and it is the same all the year round. Sensor, an industry that can hardly make achievements in China, has been made a real cash cow by Japan.

This sensor company, which started with fabless, mainly designs and sells sensors, measurement systems, laser engraving machines, etc. From the perspective of product development strategy, it never customizes products and adheres to the completely "I-oriented" standardized product R & D. This strategy maintains the regularity of product R & D, while customized products will have great cycle uncertainty, which often leads to the loss of flexibility. In order to continuously develop new products, Kearns has adopted a wide range of R & D information sources to promote the diversification of products. In terms of product series, a deeply nested product portfolio is adopted. Both sensor products and measurement systems based on sensors have become the leader in the field of measurement.

Domestic leaders such as Hikvision and Dahua all take large-scale projects. Although it also makes a lot of money, it doesn't have much to do with sensors. Even Zhengzhou Hanwei, which started with gas sensors, has focused on EPC projects such as water and environmental protection in recent years. The sensor business sector is just the name of the high-tech of this listed company, which is basically insignificant in terms of volume.

Sensors are mainly used in electronic products, industrial control and measurement, equipment and other plates. The development of sensors was first promoted by industrial automation. But the darkest thing in China is the branch of industrial control and measurement. The most typical can be regarded as Shanghai Weitai instrument company. Taking nuclear power as the starting point, this enterprise has entered the field of sensing and instrumentation, which is a pure industrial automation product. From the financial statements of the listed company, the company has been listed for 14 years, but its income in the latest year is about 60 million yuan. I have to say, the operation is bleak. You know, Honeywell, another giant company, has a global revenue of nearly 6 billion yuan in its sensing and IOT department.

Nobody cares about design software

Industrial software is a weakness made in China, especially for sensors. The sensor design software is also a very hidden dagger. In recent years, MEMS sensors are very popular. There are several in each mobile phone, such as sensing acceleration. There are at least a dozen cars in general. Bosch of Germany, Broadcom of the United States and NXP of the Netherlands are all giants in the industry. China only pulls back one corner of the MEMS sensor of the microphone, which is done very well.

However, the design of MEMS sensor needs two professional CAD software. One is Intellisuite, which was founded in 1991 in the United States. It is also the earliest special CAD design and drawing software for MEMS.

Another conventorware is also an American company. Many sensor enterprises in China are almost using it, which can occupy 80% of China's market. All the 30 research groups that undertook the MEMS Research Project of 863 Program in China used this software. Its position in MEMS sensor is basically the same as that of MATLAB software cut off from Harbin Institute of technology in June. In China, there is almost no such software. Unfortunately, this software was acquired by Pan forest lam in 2017; Lam is the second largest semiconductor equipment manufacturer in the United States. This is the easiest place for the US government to cut off its supply.

Industrial software is very subdivided. If you don't go deep into the industry, many software are hidden and invisible. This situation is as like as two peas. Sensors and industrial software seem to be wearing invisibility vests. It is these invisible industrial software that secretly block many lifelines made in China. Sensor design software is one of the most important hidden holes. Without software, these sensors are difficult to design.

It's almost all stuck

In China, consumer electronic sensors have made great progress in recent ten years due to the pull of the market. However, in industrial grade sensors, the jamming is worse than that of chips. Around control and measurement, especially instruments and sensors, almost 100% are imported.

The two giants of China's instrument transmitters are "foreign cores". The annual output of Chongqing Yokogawa instrument is 300000 sets of resonant transmitters, and the sensors are made of Yokogawa, Japan; Beijing Far East Rosemount has 300000 metal capacitance transmitters every year, using Rosemount sensors from the United States. It can be said that these two leading enterprises, which occupy more than 70% of China's market, basically work for Japan and the United States. The situation of other enterprises is the same. Suzhou enderhaus e + H uses about 50000 units a year, using German e + H; The leading enterprises of domestic brands basically use German firstsensor. What's worse, this company was acquired by American sensor giant Tektronix in March this year. This is actually very dangerous for China's appearance. There is great uncertainty about whether German sensor chips can be bought in the future.

This means that for the transmitters widely used in petrochemical, pharmaceutical and other process industries, in addition to the chips of Yokogawa, Japan and Rosemount, the sensors originally used by German companies must now rely on American companies.

Other industries are basically in a similar situation. According to the statistics of the blue book on sensor development in China by the National Engineering Research Center of sensors, the dependence of automotive sensors, high-end chemical gas sensors, optical fiber sensors and environmental detection sensors on foreign imports is more than 95%. As for ocean sensors, automatic buoys, underwater gliders and sea buoys for mobile observation platforms are 100% imported.

The PM2.5 value that Chinese people are very concerned about is basically the product of American thermoelectric company, an instrument giant. The micro oscillation balance used in it measures the concentration of air particles by measuring the mass of small particles on the filter membrane and causing the frequency change of the oscillation tube. Based on the precision measurement sensor, a PM2.5 measuring instrument of thermoelectric company often costs hundreds of thousands of yuan or even millions of yuan. Only national measuring stations can really afford this instrument. Until recently, this technology was basically conquered by Tianjin Tongyang company founded by a doctor graduated from Jingyi College of Tianjin University. This is a very lucky development.

The neck clamping method of the sensor is different from most other industrial products. It is like a minefield all over the mountains, scattered and hidden. It will be a long process to break through this kind of neck short board item by item. Moreover, it is basically unrealistic to break through one by one.

Vacillation in history

Sensors, communications and computers are known as the three pillars of modern information technology. However, the sensor, which was originally at the strategic point, has been basically marginalized in China's industrial position.

There have been historical vacillations in China. According to the domestic informatization elders, in the early 1980s, some experts participated in the research and drafting of relevant policies on the subject of "information technology development policy" presided over by the national science and Technology Commission. At that time, the first question to be solved was: what technologies does information technology include? The four technologies of computer, integrated circuit, communication technology and software have been unanimously agreed by experts. The problem lies in "sensor technology", and everyone disagrees.

Wechat pictures_ twenty trillion and two hundred and one billion sixteen million one hundred and one thousand nine hundred and fifty-eight

Figure 2 composition of information technology in China

In theory, we all agree that sensor technology is an important part of information technology. Without sensors, information technology is incomplete and the system cannot be self consistent. However, from the perspective of industry turnover, the sensor industry was too small at that time, not to mention the large industry such as communication industry, but also the "small industry" such as software at that time. If they are listed in the document, it is very difficult to write. After a long discussion, I finally "reluctantly give up my love".

It can be said that information technology has just started. As one of the fulcrums, sensors have been marginalized from the beginning. This deviation means that China's informatization has always been lame informatization. In the digital age, industrial Internet has become a national strategy, and this lameness is more obvious. However, the invisible deviation formed by this historical vacillation has become more and more deformed after 40 years, and has not been corrected until now.

Now, it's time to go back to the origin and rebuild the foundation.

Notes

The chip is stuck in the neck, and the whole country is excited