AI application exploration phase
a&s plus observed that the time for traditional security companies to really launch AI products was concentrated in the second half of 2017, slightly lagging behind the emerging AI companies in the security market. However, in general, before 2018, there were not many AI products that really landed, and AI products and solutions such as wandering, item legacy/loss, perimeter detection and face recognition launched by mainstream manufacturers were not mature.
For example, in practical applications, perimeter detection cannot filter out interfering factors such as leaves, animals, and shadows, resulting in a high false positive rate. Face recognition will still be affected by the decoration people wear, the light environment, the side face, and so on, and can not be accurately recognized. The limitations of the application of these technologies lead to a large gap between the product and the industry demand.
Since 2018, there are still many companies in the security industry that carry out product or brand packaging through the concept of artificial intelligence, but the attitude is obviously trending toward rationalization. Enterprises began to realize that AI is just a tool, to match the needs of the scene, in order to maximize the commercial value, only rely on financing can not support enterprises to go further, at the same time, the security industry business fragmentation, complexity and engineering and other characteristics, require enterprises in addition to the ability to have technology, enterprises must invest a certain amount of energy to deeply cultivate the scene, grasp the real needs of users. Therefore, most enterprises began to put their energy into the study of scene needs, and then launched customized AI products and solutions to adapt to scene applications, and achieved application landing.
After entering 2019, security manufacturers, AI enterprises, and cross-border giants have more mature responses to the fragmentation of security scenes, not only covering front and back end products in products and solutions, but also focusing more on product architecture, open platforms, and data services, and constantly jumping out of security and empowering their own technology.
AI security in-depth fragmented application scenarios
After several years of exploration, more and more enterprises have not only realized the application of AI in the industry, but also infiltrated into the traditional fragmented field that is difficult to penetrate, and launched innovative application exploration in many fields such as intelligent manufacturing, life services, and security management, such as:
(1) Under the promotion of national policies, China has moved from "manufacturing" to "intelligent manufacturing", and industrial equipment has also been comprehensively developed and applied. While promoting China's industrialization, industrial equipment maintenance has become the current problem - industrial equipment maintenance does not have online "intelligence", once the failure leads to production interruption, it will cause huge economic losses to enterprises. Although most enterprises will arrange personnel for maintenance, the traditional maintenance cycle is long, and the efficiency needs to be improved.
Some security manufacturers such as Hikvision, in response to this problem, combined with the scene problem of industrial equipment, launched an industrial auscultation microphone. The equipment receives the sound of industrial equipment in the process of operation, and after discriminating the sound based on AI algorithm, it can quickly find the fault sound in the subtle audio. By combining with the data, it can quickly inform the management personnel through the background and point out the equipment fault problem, thus simplifying the troubleshooting process of maintenance personnel and completely curbing the fault in the early stage.
(2) In recent years, with the improvement of living standards and medical conditions, the number of elderly people in China has grown rapidly. According to the seventh National census, the country has a population of 1.412 billion, with 264 million people over the age of 60, accounting for 18.7 percent of the total. Among them, 200 million people aged 65 and above, accounting for 14.1% of the total population. Compared with 2010, the proportion of people aged 60 and above increased by 5.44 percentage points - the aging rate of the population has exceeded the birth rate, and how to provide for the elderly has become an urgent social problem that needs to be solved.
In the practical application process of the traditional smart elderly care scheme, some systematic defects have gradually emerged: the way of wearing wearable devices will bring uncomfortable experience to the elderly, and wearable devices require the elderly to persist in the action of repeated charging for a long time, increasing the energy consumption of the elderly, and it is easy to interrupt wearing. The construction of video solutions in bedrooms, bathrooms and other places involves the personal privacy of the elderly, and has a negative impact on the psychology of the elderly. In order to break this embarrassed situation, Yushi Technology launched the smart health millimeter wave radar series, through simple installation, can effectively solve the problem of elderly care, not only can achieve 7×24 hours of all-round uninterrupted care, but also can avoid infringing on the personal privacy of the elderly, so as to provide a dignified, non-inductive and safe care program for the organization and elderly people;
(3) With the rise of big data, cloud computing, artificial intelligence and other technologies, digital development has become one of the future development directions of society, and innovative applications based on digital technology have been rapidly developed and increasingly popular. In the process of digitization and intelligent landing in all walks of life, identity authentication has become an extremely important part of it. However, the traditional authentication mode has many problems, such as low efficiency, high cost and difficult security guarantee. Different from the general authentication mode, the trusted identity authentication technology is compared with the data information of the national "Internet +" trusted identity authentication platform, which has higher reliability and security than the previous authentication mode. Because of this, many technological innovative enterprises have laid out this scene. For example, Entropy based Technology has launched a new generation of trusted digital identity authentication terminals, which can flexibly and efficiently meet the diversified identity authentication needs of customers.
The above similar cases are numerous, and the traditional relatively fixed mode of security business compared with the current wave of industrial digitalization, more is applicable to the customized business of fragmented scenes, which has become the focus of attention of security enterprises in recent years.
Generative AI opens a new wave of technology applications
In 2023, large model and generative AI are leading the current new wave of science and technology, and the industry has also explored new technologies. In essence, AI technologies such as GPT, AIGC and multi-modal can achieve more efficient data processing and recognition by improving the system's understanding and analysis ability of a large number of images, voice and text data. On the other hand, under the integration of multi-modal algorithms, more personalized and customized services are derived, and these two points are also in line with the current trend of active security and pan-security.
Although the application of AI in security or other projects is still in the early stage of development, driven by the increasingly rational market cognition and the increasingly mature integration of new technologies, the industrialization process of AI will undoubtedly accelerate, and the competition in the future industry will be more focused on solutions, integration and operational capabilities.
In recent years, the spread of the epidemic has had a broad and far-reaching impact on the global economy and society, as well as the development of the digital economy and the smart Internet of Things industry, with the acceleration of digital transformation and upgrading, and the rapid improvement of business complexity and differentiation. The current video technology has broken through the connotation of traditional monitoring and security, and the application of thousands of lines and industries have been linked, with the growth of product computing power and the decline of cost, the application proportion of intelligent Internet of things products in the digital transformation of thousands of lines and industries will continue to increase.
For example, in the industrial scene, in response to the security risks of gas leakage in chemical enterprises, Hikvision launched a gas cloud imaging monitoring camera, which can monitor dangerous hydrocarbon gases, locate the leak source, and display the leak gas trajectory in real time. The gas cloud imaging camera improves the efficiency and convenience of dangerous gas detection with its long-distance, non-contact VOCs gas monitoring method.
In short, the ability to mine the value of data has become the core of smart security in the future.
In the future, from security to industrial digitization
In addition, multi-dimensional perception technology, big data applications, and the digital transformation needs of thousands of industries have opened up a new market for intelligent security - AIoT, and the future industry has a huge space for development.
At the same time, with the construction of intelligent security ecosystem has become an industry consensus. Security market competition has also risen from the dimensions of simple hardware competition and solution competition to the competition of the industrial ecological chain, the core of which is technology enabling strength, platform open ability and partner recruitment ability.
The application of intelligent iot is highly fragmented and scenario-based. With the continuous improvement of the cost performance ratio of chip platform in recent years, the switching cost of software and algorithm on hardware platform has decreased, and it can be commercialized. Many enterprises have also launched AI open platform and algorithm mall, such as Hikvision on the one hand, built an AI open platform to automate algorithm training. On the one hand, it supports the replacement of adaptive algorithms and applications on the hardware.
Undoubtedly, the degree of openness is a sign of whether the industry is mature or not, and it is also the primary factor driving the intelligent security industry to the development of industrial digital transformation.