Author | Li Xi
Edit | Yu Kuai
This year, AIoT seems to be on fire again.
AI was born in 1956 and was widely recognized by the 2016 man-machine Go battle.
Compared with AI, IoT is still a “little fresh meat”, which was proposed by Bill Gates in “The Road to the Future” in 1995.
Ten years later, the “ITU Internet Report 2005: Internet of Things” issued by the International Telecommunication Union formally proposed the “IoT Internet of Things”, which laid the conceptual foundation for the “Internet of Everything”.
In other words, the birth of AIoT, without even a clear time point, was formed by the gradual integration of different technologies in development.
In 2017, companies led by Xiaomi began to aggressively attack AIoT, and AIoT ushered in its first highlight.
After several years of silence, in the past two years, from security companies to AI companies, AIoT has been used as a company positioning or future strategy, and AIoT has become “scented” again.
A concept that has been born for many years, is coming to the front again, is it “fried cold rice”, or is it really “sweet pastry”?
1. The era of AIoT is at the right time
Since ancient times, technologies and concepts have been ahead of implementation and practice. It takes a cycle from a concept to mature. In the long river of technological change, there are only a few who can seize the opportunity of technological paradigm change and persevere. .
Just like the evolution of smartphones, which gradually matured on the basis of the development of communication technology, network speed, ecosystem, etc., its popularization took 5 years.
It is not accidental that AIoT has exploded again this year.
The rise of the AIoT concept in 2017 has brought the market into the exploration period of product functions, models, technological transformation, and market education. After experiencing the care of capital and the difficult labor of landing, the current AIoT market is less impetuous and more rational.
AIoT is no longer a concept, but an industry, and it is the industry with the most landings and the largest market. At the same time, this is also an industry that cannot make quick money and needs to work hard.
After the 5-year start-up period, 2022 will become the node for the restart of the AIoT market.
And now, it is also the best era of AIoT.
-
The technical conditions are gradually in place
“There are almost no companies on the scene that have nothing to do with AI.”
This is the most intuitive feeling of an industry insider after visiting various exhibitions in the past two years.
When new AI companies emerge in an endless stream, open algorithm platforms are increasing day by day, and in the past few years, artificial intelligence has become popular, and algorithms are no longer out of reach.
Last year, a video of “restoring images of Beijing from 100 years ago” surprised the world.
In fact, AI repair is not new. The difference is that the technology behind this technology has changed from professional AI scientists to “you, me, and him” around you.
Behind the “living” of history, it is actually an explicit sign of the maturity of AI technology and tools, and the threshold for AI technology research and development is gradually lowering.
AI technology has also removed the bubble, removed the false and preserved the true, and returned to the essence of “enabling the industry”.
In fact, any technological development has several ups and downs. When a new technology appears, people will think that it can solve all the problems in the industry based on high expectations. Excessive expectations bring disappointment, but it does not mean that the technology itself does not in development.
Since AI technology was proposed in the 1950s, it has experienced at least two peaks and troughs. With the development of computer vision technology, China’s corresponding artificial intelligence technology has begun to enter large-scale commercial use this year.
One of the prerequisites for the digitization and intelligence of the industry is the richness of algorithm categories and the high cost performance of algorithms.
The AI threshold has been greatly reduced, the labor cost has dropped, and the richness of algorithms has increased, all of which have laid the foundation for the large-scale implementation of the technology.
National AI is coming quietly.
On the IoT side, technological breakthroughs have been made in the perception layer, transmission layer, and application layer that support the development of the Internet of Things.
In the perception layer, the breakthrough of sensors in material technology, biotechnology, manufacturing process, information processing, etc., especially the emergence of intelligent sensors, has enhanced the ability of IoT perception, and has wider and deeper application scenarios;
At the transport layer, advances in communication technology, such as 5G, can support massive connections while greatly reducing transmission delays;
At the application layer, breakthroughs in big data, edge computing, and artificial intelligence technologies can process the massive data perceived by the Internet of Things more efficiently, turning the data into “mineral resources” that can be used for intelligence in all walks of life. upgrade.
Technologies such as AI, 5G, cloud computing, big data, smart sensing, and blockchain are accelerating their integration with IoT.
Yes, every era has its own voice. Maybe the technology landing is not poured down by torrential rain, but a long stream of water.
The industry has come to the eve of the outbreak, and basic technologies such as AI, chips, cloud computing, and communications have gradually matured, making the connection between people and things, and between things more and more intelligent.
-
Market sophistication increases
In the past ten years, the biggest new species in the 4G era is the smartphone, which has brought the connection between people and detonated the mobile Internet.
With smartphones as the core, mobile Internet applications in various industries such as shopping software, taxi platforms, and map query have emerged.
Similarly, due to its long industrial chain, low industrial concentration, and fragmentation of scenes, AIoT was mainly used in large enterprises in the past. Therefore, AI, as an aristocratic technology, is difficult to acquire in long-tail scenarios, and public enterprises have no technology and cannot retain people. If there is no data, it is impossible to make delicious food. The layman looks lively, but in fact, “I’m all overwhelmed.”
Today, AI and IoT technologies are mature, and in addition to security and traffic scenarios, there are endless demands for countless scenarios, and the massive waist-tail market has begun to rise and fall.
The cost of education in the market is gradually decreasing, and companies almost no longer spend a lot of energy on popularizing what AI is and why it needs to transform.
At the same time, the gusty winds of digitalization, the repeated epidemics, and changes in the international situation are all forcing digital transformation of enterprises.
The AIoT industry is transitioning from a period of accumulation to a period of rapid development. Up to now, tens of billions of smart devices and edge devices have spread across all fields of society, driving industrial upgrading.
Today, the penetration rate of AI in IoT devices is increasing.
According to IoTAnalysis data, the number of global IoT connections will reach 11.7 billion in 2020, surpassing the number of non-IoT connections for the first time.
When AIoT terminal shipments will surpass traditional IoT terminals, it is not only hardware, but also the carrier of operational services, and the most important entrance for enterprises to enter the digital world.
In the next 5-10 years, it is an indisputable fact that the AIoT market will reach hundreds of billions and trillions.
The next ten years will be the ten years of AIoT, the core of which is to drive the digitalization of the industry.
2. Why can security companies transform and occupy the C position in the AIoT market?
An interesting phenomenon, security companies have become the main force of AIoT in recent years.
In April, Hikvision released its 2021 annual report, which for the first time refreshed the positioning of enterprise intelligent IoT AIoT in the annual report. At the same time, it took AIoT as the direction for the next ten years.
In the same month, Dahua also emphasized the business core of “Intelligent IoT AIoT” in its annual report.
The earlier Uniview, in the 2021 annual report, has long used “intelligent IoT” as the key word.
In addition to security companies, AI companies are also rushing to AIoT with great fanfare.
At the Megvii Enterprise Business Partner Conference in April, Megvii co-founder and CEO Yin Qi made it clear that AIoT is the next wave of the Internet and also the strategic direction of Megvii.
However, in the eyes of AI Nuggets, compared with the AIoT at the C-end with a thunderous explosion and a few climaxes, the AIoT at the G-end and B-end was relatively gentle in the early stage, and now it is fierce, but it is also more in line with technology and industry development laws.
In the early days, AIoT on the G-side and B-side was not paid attention to by the outside world, but companies in the field have been on the road to AIoT evolution.
Becoming the hottest company in the AIoT field this year may be the choice of the market.
There are two facts that are now indisputable.
First, intelligent security is an excellent channel to enter the AIoT market.
Second, under the transformation of de-security, traditional security companies have successfully turned into intelligent IoT companies.
Look at the first point first.
A few years ago, AI officially left the laboratory and chose security in thousands of industries.
Before the arrival of AI, the main body of service in the security industry was more of the user’s security department; after the arrival of AI, the perception function and market of AIoT have been magnified hundreds of times, bringing users new desires, new needs, new direction.
Security has become the largest intelligent gathering point in urban infrastructure. All kinds of intelligent sensing devices correspond to the ubiquitous connection of digital cities. Digital cities connect all nodes in urban space with the network as a medium, so as to achieve maximum understanding and optimal scheduling. , to provide people with more efficient and comprehensive services.
Video AI + big data will be the connector of the entire digital city in the future, and also the basis for the healthy circulation of the entire city’s data.
Smart security can enter a more advanced stage in this process.
The realization path of AIoT is to collect massive data generated by IoT terminals, store them in the cloud and edge, and then analyze big data and use artificial intelligence to realize the digitization of everything and the intelligent connection of everything.
The IoT Internet of Things collects various properties of the material world through various sensors and uploads them to the Internet. After massive data is processed by big data technology, the material world can be described based on the data, that is, the information world can be generated.
In the world of AIoT, video is a strong entrance to data, and the continuous high-quality data generated will become a new energy in the digital age.
Video capabilities are the strengths of security companies that have been “hardened” in the market.
AI makes security more than security, and makes AIoT the “next stop” that security manufacturers are arriving at.
Using technology as a blade, cut a wider space. Taking Hikvision as an example, Hikvision is now not only in video perception capability, but its technical means have expanded from visible light to millimeter wave, infrared, X-ray, sound wave and other wider fields, coupled with the accumulation of AI and big data technologies. , the ability of IoT perception has been very broad, and the industry needs of these technologies are also beyond the scope of security. From the Internet of Things to the intelligent Internet of Things, it is a matter of course.
Look at the second point.
Since computer vision began to fully penetrate the security industry, the boundaries of security have become increasingly blurred.
With the blessing of artificial intelligence and cloud computing, the value of the entire security industry has increased rapidly, front-end and back-end products with deep learning functions have been launched continuously, and the back-end portrait big data platform has begun to penetrate.
In the solution application process of the actual project of intelligent security, the operation mode of the inherent players has changed from the previous hardware service to the combination of software and hardware; what they enjoy is not only the original value of surveillance cameras, but also the potential gold behind IPC. mine.
The essence of de-security is the physical medium that carries security. With the blessing of AI, it has surpassed its original capabilities.
When this industry moves from traditional security based on safe cities, to intelligent security based on AI and big data, to today’s intelligent IoT where everything is interconnected, corporate positioning is also evolving.
As early as 2019, the security industry began a de-security revolution.
Since 2016, Hikvision has transformed into a video-centric intelligent IoT solution and big data service provider, and has gradually improved and upgraded its own intelligent architecture, broadened the ecological boundaries, and gradually separated its product applications from traditional security products to open wider channels.
The corporate positioning of traditional security giants has also taken three steps according to their own capabilities and the market:
“Security Manufacturer”, “Internet of Things Solution Provider Based on Video Capability”, “Intelligent IoT AIoT”
Security is a small market with serious fragmentation. It is a very complex system from the development response to meet customer fragmentation needs to the construction of rapid delivery capabilities.
Security manufacturers have long since moved from the small battlefield of security to the further urban AIoT and industrial AIoT markets with unlimited opportunities.
They have honed their greatest capabilities in the security field and will become their weapons in this trillion-dollar market.
Hu Yangzhong wrote in his New Year’s greetings to all employees at the beginning of 2022: If Hikvision has become the world’s largest security manufacturer in the past 20 years, Hikvision will become one of the world’s largest AIoT manufacturers in the next 10 years.
3. A promising world of AIoT
In the Internet era, IoT is a PC; in the mobile Internet era, IoT is a mobile phone, and in the intelligent IoT era, IoT is all hardware.
What is “all hardware”?
Everything you can imagine: cars, industrial robots, industrial cameras, cameras, smoke sensors…
IoT is a super perception system, rich in IoT sensing devices, which can collect various types of data such as vision, hearing, temperature, and environment. With the blessing of AI, IoT devices can be widely connected. With the brain, perception and analysis , execution, and even prediction, processes are digitized and intelligent, and AI is also more intelligent under the feeding of data.
AIoT is based on the interconnection between things and things and people and things in the traditional IoT to realize the interconnection and intercommunication of equipment and scenes, as well as connection and data interaction.
New terminal connections will generate new and massive amounts of real-time data. With these massive amounts of data, data analysis and mining can be carried out. This trend contains a lot of business value and opportunities.
The integration of AI and IoT aims to form an intelligent ecosystem, and within the ecosystem, interconnection between different intelligent terminals, different system platforms, and different application scenarios.
AI+IoT is not simply the interconnection of everything, but the expansion and explosion of core terminals.
According to IDC data and forecasts, the global AIoT market will reach US$226.4 billion in 2019, and is expected to reach US$482 billion by 2022, with a compound growth rate of 28.65% from 2019 to 2022. According to iResearch, the overall AIoT market size will reach 654.8 billion yuan in 2021.
In the era of intelligence, the fields of To B, To G and To C will all be the battlefield of AIoT.
Smart travel, smart wear, smart healthcare, and smart home in the consumer field; smart city, smart meter, smart security, smart energy, smart firefighting, smart parking, and smart disaster prevention in the urban field; smart industry, smart logistics, smart Smart Retail, Smart Agriculture, Internet of Vehicles, Smart Community, Smart Park.
Intelligent IoT is a bigger market than urban AIoT, which includes urban AIoT, industrial AIoT and personal AIoT.
With the blurring of boundaries and constant cross-border, there are veterans, rookies, and spoilers who are desperately rushing forward in the blood pool.
The technical wind blows, with whistling and sand, trying to change the industry situation that has been dusty for many years.
For all AIoT entrants, it has caught up with the arrival of a big change, and everything is reborn.
And what the future needs is the vision and ambition to break the game. On the way to breaking the game, look more at the road under your feet, and often look at the outside world.
After all, most of the time, victory or defeat is not decided when Wang is appointed, nor is victory or defeat determined when the results are announced. Leifeng Network Leifeng Network Leifeng Network
This article is reprinted from: https://www.leiphone.com/category/iot/LPdiSL2ZNKCoSeBA.html
This site is for inclusion only, and the copyright belongs to the original author.