The year 2017 just passed was a year when the concept of AI was extremely popular, and artificial intelligence became the most popular technology word in 2017. At present, the application hotspots of artificial intelligence in China are mainly concentrated in the fields of computer vision, speech recognition and natural language understanding. With the help of the huge video surveillance resources accumulated through the construction of "safe city" in the past ten years, computer vision technology naturally has a clear and fast implementation of application scenarios. Therefore, security will be the fastest industrialization, marketization and most competitive industry application of artificial intelligence.
This article attempts to analyze the change trend of the third ten-year security cycle in the era of artificial intelligence from the past ten-year cycle law of the security industry, as well as the impact of artificial intelligence on the market competition pattern of the security industry, for reference by peers.
1. The ten-year cycle law reappears, and the security industry will usher in great changes in the era of artificial intelligence
Looking back at the development of the domestic security monitoring industry in the past 20 years, we found that the development of the security industry is very similar to the development of the social economy, basically a cycle of ten years. In the mid-1990s, the domestic security market started, and 2006 was the first cycle. In this cycle, the technology is represented by analog technology, and the products are mainly dominated by analog monitoring products such as black and white/color cameras, tape recorders, multi-screen splitters, video matrices, and monitors. Mintong, Panasonic, Sony, Samsung, Axis, PELCO, AD and other Taiwan and overseas security brands are the main ones;
Since 2006, it has entered the era represented by digital video surveillance technology. The products are dominated by digital surveillance products such as hard disk recorders, video encoders, IP cameras, high-definition cameras, and network video storage. The protagonist of the market has been replaced by Domestic brands such as Hikvision, Dahua, Uniview, etc., with the help of the transformation opportunities in the digital age and the wave of "safe city" construction, the domestic security market and security industry have developed rapidly, resulting in many heavyweight listed companies, some The company has also become a world-class security giant.
Also in this cycle of change, the influence of the original foreign big names in the domestic security market gradually declined, or even disappeared. As of the end of 2016, the traditional "safe city" construction is coming to an end, the homogeneity of technology and products has been very serious, and price competition has become the most effective competitive strategy among domestic security giants, indicating the second ten. the end of the annual cycle.
The arrival of deep learning and artificial intelligence has opened the third ten-year cycle of the security industry, and 2017 has naturally become the first year of the third change cycle. The first ten-year cycle is the era of analog surveillance, the second ten-year cycle is the era of digital and high-definition monitoring, and the third ten-year cycle is undoubtedly the era of artificial intelligence.
2. Analysis of the market change trend in the third ten-year cycle
From the changes in the first two ten-year cycles, we found that changes in technology and products will simultaneously bring about huge changes in market demand and competitive landscape.
The first change in the security industry has realized the change from analog to digital technology and products. Just the technological upgrade in transmission and storage has brought about the huge market demand for the construction of "safe city" and the re-washing of the market competition pattern. Brands, those world-class security industry giants who were once invincible have almost completely retreated in this round of non-disruptive technological upgrades, and reluctantly handed over the industry status and market to the latecomers they once dismissed. Then the technological change led by artificial intelligence will inevitably bring greater or even subversive changes to the market demand and competitive landscape.
First, market demand will be redefined
In the last 10-year cycle, the application requirements of the security monitoring market mainly revolved around the three levels of "seeing clearly, encircling and transmitting", defining requirements with "information", focusing on information (digital video) collection, transmission and transmission. Storage, upgrade and optimize from three aspects: image quality, monitoring distribution and transmission network.
With the application of artificial intelligence technology, the demand will be based on "inquiry, analysis, and prediction", and the demand will be defined by "data", focusing on data (face, vehicle, human body and other video structured data) collection and large-scale data. Data analysis, judgment and risk warning. Security monitoring will move from the age of information and digitalization to the age of data, only then can we truly be able to meet and embrace the era of security big data.
The most obvious one is the demand for public security big data application. In the past few years, under the influence of Internet companies, public security organs in various places have built so-called public security big data platforms, which integrate the personal identity information, hotel registration data, vehicle checkpoint data controlled by the public security organs, and data from Internet companies, mobile operators, The data obtained by banks, etc. are summarized and analyzed to provide more extensive data support for the public security organs to accurately crack down on illegal and criminal acts.
However, due to the uncertainty of data sources, the validity of data, the authenticity of the data itself, and the non-real-time nature of data acquisition, it is difficult for the so-called big data platform to achieve the expected results. The large-scale application of artificial intelligence technologies such as face recognition, vehicle face recognition, and video structuring has largely filled the gap of effective data sources, making the application of public security big data no longer "water without source" and "cook without rice". ".
Second, the market size will be greatly increased
In the second half of the last ten-year cycle, the growth rate of the security monitoring market has become a slowing trend year by year. Lack of technological change and the need for new applications to drive, the decline is very obvious.
The arrival of the era of artificial intelligence will not only release a large number of unmet market needs, but also bring more and richer new application demand scenarios, and the market scale will be greatly increased. It is expected that by 2022, the compound growth rate of the artificial intelligence security market will reach about 40%-50%.
In addition to the algorithm hardware, software and big data platform required by artificial intelligence itself, the use, scene, installation location, installation height, installation angle and equipment indicators of surveillance images under the application of artificial intelligence technology will be redefined. The system will be re-engineered to meet the application needs of artificial intelligence and big data.
In addition, some places and areas that were not valued originally are the best collection and control points under the artificial intelligence and big data system, such as the entrances and exits of communities, banks, bars, hotels, bath centers, supermarkets, parking garages, etc. These places All have to be redone. Some places where monitoring equipment is already installed need to add supplementary light facilities, and some ordinary cameras need to be replaced with ultra-low illumination cameras or cameras with some front-end intelligence, which will bring great market opportunities to the entire industry.
The application of face recognition technology will bring a revolutionary change to the one-card and access control industry. The original access control and one-card mainly rely on RFID technology, which is easy to lose, easy to forge, and needs to be produced separately. These are precisely the face recognition. technical advantages. Therefore, face recognition will replace the original RFID technology on a large scale, and the original system needs to be upgraded on a large scale, thus bringing a large market space.
Third, the market competition pattern faces an opportunity to reshuffle
Artificial intelligence and traditional security have essential differences in technology categories, and artificial intelligence has a high threshold for technology accumulation, enabling artificial intelligence start-ups to reinterpret the security industry from a new starting point and height, and obtain the same An opportunity for security giants to compete equally. In view of the temptation of the huge market volume of the security industry, it will inevitably attract more non-traditional security industry companies to use the sword of artificial intelligence to enter the security circle that has become a three-nation confrontation, and reconstruct the competition pattern of the security industry.
Fourth, a good security ecosystem will be rebuilt
The past ten years have not only been a decade of rapid development of the domestic security industry, but also a decade of rapid development from a hundred flowers to the monopoly of a few industry giants. A good industry and industry ecology is an essential foundation for industrial innovation and sustainable development. However, a few industry giants have used their advantages in market channels and costs to form an oligopolistic position, which has greatly damaged the industrial ecology, stifled the creativity of industrial technological innovation, and is extremely detrimental to the sustainable development of the security industry.
The arrival of the era of artificial intelligence has not only shaken the monopoly position of these giants, but also made it possible to reconstruct a healthy industrial ecology. At the same time, due to the relatively long industrial chain of artificial intelligence, it involves basic algorithm models, basic hardware, basic products, and industries. Different links from application, scenario application, solution to application delivery inherently require the division of labor and cooperation between different links to create a complete industrial chain.
3. Analysis of the competitive situation of the security industry in the era of artificial intelligence
As mentioned earlier, in the era of artificial intelligence, the security market competition pattern will face an opportunity to reshuffle. At present, in addition to traditional security companies, there are Internet companies such as Baidu, network communication equipment manufacturers such as Huawei, and entrepreneurial teams specializing in artificial intelligence technology research such as Yitu, Megvii, and SenseTime. security industry. What are the competitive advantages and disadvantages of these competitors, and what position will they be in in the future market structure?
The industry generally believes that the competition of artificial intelligence at the technical level is determined by three elements: algorithm, computing power and data resources, of which the algorithm is the core, especially the completely autonomous deep learning original algorithm, which determines how far the team can go and whether it has Sustained leadership.
The second is computing power, which includes two levels: one is whether it has a powerful deep learning computing platform, which determines the speed of algorithm model iteration using deep learning; the other is the adaptation ability of algorithm model and hardware, under the same hardware conditions The computing power determines the stability of the product and the availability in terms of large-scale deployment and operating costs; and the data source used for algorithm model training also largely determines the speed of algorithm model iteration.
At present, only a small number of domestic enterprises or teams have mastered the original deep learning technology. Most of them use open source algorithms or some technologies brought back from abroad for targeted model training, while more do not make any algorithm models themselves. Take the SDK provided by the third party for product packaging. Therefore, at the technical level, the professional artificial intelligence algorithm team will have more advantages. Although the routine of the security industry is still unclear in a short period of time, it is impossible to immediately adapt to the play of the security market. The requirements for computing power are becoming more and more demanding, and the advantages of the latter will become more and more obvious.
Some people will say that the essence of deep learning is data. Whoever masters the data resources will lead in the future, but the data ownership in the security industry belongs to the owner, not the operator, nor the integrator, nor the product manufacturer. of. Moreover, data is only material, without the ability to process data, no amount of data will produce a good algorithm model. Now, the neural networks mastered by some algorithm teams have reached thousands of layers and have huge supercomputing platforms. These technical advantages cannot be surpassed in a short time.
In terms of market competition, in addition to technical competition, there are also factors such as brand awareness, market sales channels, on-site technical services and project delivery. At this level of competition, traditional security companies have obvious advantages, which are unmatched by Internet companies and professional artificial intelligence companies. They have a good relationship with users, it is easier to understand the ideas of users, it is easier to instill their own products and ideas into users, and they have a strong ability to implement projects. For users, it is often difficult to distinguish the difference between artificial intelligence and monitoring. They simply think that it is just an upgraded version of traditional security monitoring.
Another more important competitive advantage of traditional security companies is the ability of product hardware and platformization, especially the advantages of front-end products such as cameras - the rapid advancement of artificial intelligence algorithms to the camera side. However, since the existing algorithms are still far from meeting the needs of users, far from being able to support more and more demanding application requirements and complex application scenarios, the face recognition algorithm model still needs continuous iteration and optimization, and even in the future there will be significant It is a breakthrough possibility. The algorithm currently solidified in the front end cannot be iteratively upgraded, and it will inevitably be eliminated in the future, which is also the price of "eating crabs".
Although edge computing + cloud computing is an inevitable trend in the development of artificial intelligence in the future, when the algorithm is far from fully mature, reluctance to solidify and marginalize is only an initial attempt. At present, the professional artificial intelligence algorithm team has also deeply realized the importance of edge computing in the future, and is also developing and producing front-end artificial intelligence cameras or providing front-end algorithms, and even actively deploying AI-specific chips, in order to occupy the future competition. favorable position.
In the future market competition, traditional security companies will rely on their "domestic power" to take the lead and implement some early-stage pilot projects, thus achieving the effect of the education and training market, and laying a good market foundation for the full blooming and landing of artificial intelligence in the security industry. , which is also a due contribution to the next ten-year cycle of the security industry as a beneficiary of the previous ten-year cycle.
With the continuous enrichment of application scenarios and the continuous changes in user needs, the market's expectations for artificial intelligence continue to increase, and the advantages of algorithm companies will be more fully exerted. By building a win-win artificial intelligence + security ecological chain or industrial investment mergers and acquisitions, etc. Capital operation can effectively make up for its shortcomings in the security industry and occupy a very important position in future competition.
In addition, Internet companies will have a strong competitive advantage in artificial intelligence public safety cloud services by virtue of their advantages in Internet access and big data resources. Network communication equipment companies will bring more comprehensive solutions to high-end security users such as public security and finance by virtue of their advantages in data communication, data storage and processing, and will surely occupy a place in the future competition.
The future has come, but the road ahead is even farther. May China's security industry continue to lead the world in the industrial wave of artificial intelligence + security and benefit all mankind.
The Anti-explosion Screen Protector is a Screen Protector that can effectively buffer impact, prevent the screen from bursting, or prevent the glass panel from breaking and falling apart due to accidental impact of the mobile phone. It has gloss, texture and surface hardness. The Explosion-proof Screen Protector strengthens the screen and reduces the chance of cracking, broken tempered glass is no longer frequently replaced.
Compared with other Screen Protectors, Explosion-proof Screen Protectors have the following advantages:
1. Anti-shattering: The PET material of the Explosion-proof Screen Protective Film has high strength, flexibility and elasticity, can fully decompose the impact force, and can effectively prevent the screen from being broken or damaged when the screen is impacted or dropped by external forces.
2. HD Clear: The Explosion-proof Screen Protector is made of high-quality PET material, which can maintain a high-definition screen display effect and truly display the original screen color.
3. Anti-fingerprint: The Anti-explosion Screen Protector has a built-in oleophobic and waterproof coating to prevent fingerprints and stains on the phone screen, providing your phone with original texture and perfect touch screen response speed.
The use of an Explosion-proof Anti-explosion Screen Protector can effectively enhance the durability and protection of the mobile phone screen, reduce the risk of screen cracking and damage, and at the same time protect the clarity and quality of the screen display. Ideal for protecting your screen.
PET Screen Protector, Explosion-proof Screen Protector, Explosion-proof Screen Protective Film, Explosion-proof Protective Film,Anti-explosion Screen Protector
Shenzhen Jianjiantong Technology Co., Ltd. , https://www.tpuprotector.com