The new infrastructure is ushering in a new golden age of AI industry landing.
At present, AI technologies such as image recognition, speech recognition, and natural language translation have been widely deployed and widely used in mobile Internet, new industries, and even many traditional industries. The accelerated maturity of the second artificial intelligence centered on machine learning and deep learning has finally ushered in the highlight moment of AI technology.
The popularization speed of AI technology is indeed far faster than the industrial speed brought by previous new technological revolutions, but at the same time that it has brought infinite benefits to the national economy and people's lives, security incidents faced by traditional industries and manufacturing industries have begun to break out frequently, and the core of my country's industrial enterprises Most of the controllers are foreign brands, with a low degree of localization and high hidden dangers. Industrial network security has gradually become the focus of major technology companies.
Recently, the industrial security AI solution provider "Paidao Intelligence" has successively completed tens of millions of US dollars in Series A and A+ rounds of financing. The A round was jointly led by SAIF Asia and Sinovation Ventures, and the A+ round was additionally invested by the old shareholder Zhongxin Capital. .
The reason behind the two consecutive rounds of financing for Paidao Intelligence may be that in the process of digital transformation, various industries that are enjoying a new wave of intelligent technology dividends are paying attention to the issue-industrial AI security.
The "invisible" crisis in the era of intelligent manufacturing
China's industrial structure centered on industrial manufacturing in the past is changing, and it is no longer a "world factory". The country has begun to regard "smart manufacturing" as an inevitable direction for national industrial transformation and upgrading.
According to the "Intelligent Manufacturing 2025 White Paper" proposed by the National Development and Reform Commission, intelligent manufacturing is defined from the three latitudes of value chain, life cycle and system architecture. Among them, the system architecture industry chain includes equipment layer, control layer, management layer, enterprise layer, and network layer.
Smart manufacturing led by the Internet of Things and robots is changing the way of production of life, resulting in the exposure of a large number of industrial control systems and equipment on the Internet, which has also become an important threat and weakness of industrial information security in various countries. According to the monitoring of the National Industrial Information Security Development Research Center, there are more than 100,000 industrial control devices exposed to the Internet worldwide.
The Industrial Internet, as one of the important paths for traditional institutions and traditional industries to achieve transformation and upgrading through "Internet +", especially needs to implant industrial control network security genes.
The Industrial Internet lacks the ability to identify and limit malicious attacks hidden in the transmitted information. According to the attack method, the Industrial Internet faces the threat of eavesdropping, camouflage, integrity damage, denial of service, unauthorized access to services, denial of use/provision, and resource consumption. Just wait.
When these are applied to some fields with extremely high safety requirements, such as industrial quality inspection, automatic driving, financial risk control, and medical diagnosis, once the judgment deviation of AI occurs, it will bring direct economic and property losses, and even personal safety. serious consequences and health risks.
In this regard, lessons from abroad have been common: In March 2019, the Norwegian aluminum company Norsk Hydro was attacked by ransomware, which caused the host computer to crash, forcing multiple factories to close and interrupting production operations; in August 2018, TSMC’s operation Computers at the headquarters and Hsinchu Park were invaded by viruses, and all production lines were shut down. This incident directly affected 3% of its operating income in the third quarter...
AI technology, which was originally intended to be applied to improve production efficiency and safety, may go to its own opposite. The new security situation inevitably requires a "new security concept" that keeps pace with the times, and it also provides a brand new opportunity for industrial security AI solution providers such as Paidao Intelligent.
Multiple roles in security games
Founded in 2018, Paidao Intelligence is an AI company with image and video data analysis as its core.
The implementation of AI technology in industrial scenarios has become the development trend of the new industrial era, and it is also one of the few composite scenario requirements in the new infrastructure field. It can not only give full play to the advantages in the digital production of enterprises, but also help enterprises in the field of operation safety. Realize the upgrade and ultimate closed loop of the regulatory system.
However, my country's petrochemical industry has a huge volume and has a good and stable development prospect. It has always occupied the top three in the output value of the country's industrial categories. However, due to the industry's periodicity and special processes, the petrochemical industry is generally not easy to intersect with information processing technology. However, the ability of enterprises to process, analyze and quickly respond to on-site information still needs to be improved.
In addition, external security pressure is also a major reason why the petrochemical industry urgently needs the empowerment of artificial intelligence systems. Safety accidents in the petrochemical industry often cause major personal safety losses, property losses, and ecological disasters. The confrontation between enterprise safety and accidents is almost a zero-sum game. Even if no casualties are caused, the direct economic loss and indirect economic loss and market reputation caused by it are still fatal to the enterprise.
Therefore, at the beginning of its establishment, Paidao Intelligence aimed at the petrochemical industry, which has a strong demand for production safety. Under such a market background, technologies such as real-time image processing, visual small sample analysis, AI and RPA have been developed.
The on-site monitoring platform Prime Eagle uses hardware + software to perceive and analyze people’s activities and specific behaviors in real time, and realize real-time detection of security threats and issue warnings.
Paidao Intelligent's AI system solution adopts the technology of small sample identification and abnormal detection, and applies it to various classification identification and continuous work analysis, which solves the problem of redundant normal data samples and lack of abnormal data samples in the petrochemical industry. Compared with traditional big data algorithms, the results are closer to the real situation.
In addition, Paidao Intelligent has also built a chemical safety map, which makes the system judgment more in line with the production process of the petrochemical industry, and can also more comprehensively and accurately meet the latest emergency safety supervision requirements.
If the upgrading of industrial intelligence is regarded as a river that is poised to move forward, then AI enriches and expands the breadth of industrial intelligence scenarios, its maturity and upgrade determine the depth of industrial intelligence, and the safety and reliability of AI determines the The length of industrial intelligence.
Paidao Intelligent knocks on the door of the market through technology, and continues to gain the trust of customers through reliable technology. In the field of industrial security, Paidao Intelligent is opening up a larger market step by step.
Entering the industrial metaverse, grabbing the beach and laying out the wave of digitalization
In recent years, some large industries have begun to benefit from the application of AI, such as the automobile, steel, chemical and power industries.
However, in most industries, the lower the degree of standardization of production scenarios, the less obvious the scale effect. The reuse of AI in these scenarios is difficult, and both the supply and demand sides need to pay a huge cost. In addition, the overall gross profit margin of the manufacturing industry is low, and the return on investment cycle is infinitely lengthened, resulting in a low penetration rate of AI.
Then, with the increasingly popular concept of the Metaverse, Paidao Intelligence has also begun to deploy cutting-edge technologies related to the Industrial Metaverse. Can this bring new problem-solving ideas to Paidao Intelligence?
Now the industry generally agrees that the metaverse includes six underlying technologies: Internet of Things technology, blockchain technology, interactive technology, video game technology, artificial intelligence technology, network and computing technology. IT, OT, and CT can only be counted as very small branches. The integration of these technologies to the extreme will create a three-dimensional virtual world that is completely parallel to the real world.
In the Industrial Metaverse, real-time collaboration between users and applications allows for updates, iterations, and changes without prepared data, streamlining workflows. The platform provides scalable, true real-time ray tracing and path tracing, enabling beautiful, physically accurate and realistic visuals.
This also means that industrial equipment has a more efficient and lower-cost solution to the security issues that are stuck in the industrial Internet. The relevant solutions and technologies formed by the Metaverse will bring substantial convenience in many aspects of industrial production. But is an "industrial metaverse" really feasible?
In reality, AI security incidents are not uncommon, and Metaverse will inevitably face these problems.
In the nascent virtual world of the "Metaverse", there is no need to talk about restraint. Naturally, some people have evil thoughts, and commonplace phenomena such as network telecom fraud, digital asset theft, forgery of other people's identities, and blackmail virus attacks may be transplanted.
The second is digital twin technology, which is particularly important for the industrial metaverse. Although digital twin modeling is not complicated in theory, it requires a lot of technology as support and a lot of data in the modeling process. Taking Paidao Intelligent as an example, if you want to carry out construction in the chemical industry scene, you need to carry out high-precision modeling of the scene, which requires a large amount of data collection, and it is not so easy to complete such a huge task.
Summarize
For China, the transformation and upgrading of the manufacturing industry is a historical subject that is always on the way. If it wants to seize the commanding heights in the new round of technological revolution, it must rely on more policy promotion and technical support, including AI.
Perhaps Paidao Intelligence can bring a faster and better development path for the transformation and upgrading of the manufacturing industry with the help of Metaverse, but this concept itself is actually a concentrated description and imagination of various application scenarios, not an independent technology . If these imaginary scenes are to be realized, the support of related technologies is inseparable.
From reality to virtuality means that the concept of safety in the industrial field will undergo subversive changes. Although there are not enough cases to prove it, this change in direction is still worth looking forward to.
Source: New Industry Insights