/ Daily News/July 27, 2022 Guangzhou -- Unwavering in its commitment to manufacturing as a city, Guangzhou is speeding up the digital development of manufacturing, guiding industrial enterprises to strengthen cooperation with software and Internet companies, and supporting the digital transformation of traditional industries with advantages. Founded in Guangzhou in 2017, Netmind Technology Co., LTD. (hereinafter referred to as "Netmind Technology") is one of the few leading technology enterprises in China with full-stack self-developed artificial intelligence technology and whole-chain domestic adaptation. Last year, it became one of the first batch of "Guangzhou 'Specialized, Special and New' nurturing enterprises". Just last month, Netsi Technology was ranked 28th on the list of China's top 100 quasi-unicorn companies for 2022 by Internet Weekly of the Chinese Academy of Sciences. In this issue, a series of reports on "Specialized, special and new" supporting plan to cultivate enterprises entered the network thinking technology, to see Guangzhou enterprises break new ground in the industrial Internet field.
Intelligent engineering quality inspection system to the whole country < p>
In the familiar home broadband installation, door-to-door installation of broadband workers usually need to complete the installation according to certain standards in the installation manual, such as what each cable should be made, how to mark and how to handle the line. To ensure that workers follow the manual, operators require workers to take about 30 photos of each household broadband installation to be kept on file for random inspection.
"30 photos per household, assuming there are about 10 million households in, say, Guangdong Province, that's 300 million photos. Before operators can only manually to spot check, even if dozens of organizations see dizziness there every day can not see a few." Wang Huan, chairman of Netmind Technology, told reporters that in such business scenarios, AI exerts its maximum value.
Based on artificial intelligence technology, the intelligent engineering quality inspection system developed by Netthink Technology for telecom operators has sorted out more than 30 processes for installation, and each process defines standard images and videos for project implementation. "Ai doesn't distinguish very well at the beginning, but it automatically improves its efficiency by constantly learning from massive amounts of data." Wang Huan introduced that the current system can be processed within 30 seconds to 2 minutes according to the number of each work order, and the average processing time of each photo is less than 100 milliseconds. The time of single task in the execution of daily installation and maintenance tasks has been reduced by more than 90%, and the rate of finding problems has increased to more than 90%. The implementation of artificial intelligence in business scenarios has truly realized cost reduction and efficiency increase for enterprises, and this system has also been promoted from the early use in individual provinces to the nationwide use.
Digital twinning provides "optimal solution" for workshop production
Digital twinning is another key to the digital transformation of manufacturing industry, and is also another business core of Netsys Technology. What does a digital twin mean? "You can think of it as digital twins. It's a virtual representation of the physical world in the digital world. At the same time, it's a linkage. "Wang Huan explained.
According to Wang Huan, the best application scenario of digital twin is in the field of intelligent manufacturing. From 2020, Netsi Technology began to communicate with GaC Group. After more than a year of testing, it finally signed the first phase of the digital twin contract with GAC Group's auto plants. There are two specific landing scenes. The first application is to display the production floor completely in real time on the system. "For example, in the assembly workshop, the state of every manipulator and every part on every production line is projected into the system in real time, so as to solve the patrol monitoring problem of the project." On this basis, the problem of production management can be solved at the second level. When arranging the production plan, the optimal arrangement for production can be provided through the data analysis of the operation condition, working efficiency of each production link and the time arrangement of shutdown detection. The digitalization of automobile manufacturing in Guangzhou has covered all production lines, with the automation rate exceeding 95%, according to the Guangzhou Development and Reform Commission's response to the suggestions of deputies to the municipal People's Congress this year. After digital production line coverage, how to let the data speak? This is also the third level of intelligent manufacturing targeted by Wang Huan's research team. "This is a field deeply combined with informatization and industrialization. For example, a production machine may have tens of thousands of sub-parameters. Based on understanding the principle of the machine, the optimal state of the machine can be analyzed through continuous data changes and tuning tests."In Wang Huan's opinion, "a more exciting scenario of digital twin is autonomous driving." Through digital twinning, real-time monitoring monitors the actual conditions around each autonomous vehicle and its operation on the screen. "For example, if a self-driving car drives on Guangzhou Avenue, its surrounding real-time information, road conditions, traffic conditions, pedestrians, traffic lights and weather will be displayed on the monitoring screen one by one, so as to provide the basis for remote guarantee of driving safety."
< p>
< P > No matter the application of intelligent engineering quality inspection system in the field of telecom operators or the application of digital twinning in intelligent automobile manufacturing, they are all realized on the same open platform of AlphaMind® AI capability. This new generation of artificial intelligence platform is based on deep learning, integrates key AI technologies such as machine vision and knowledge graph, and has built-in models of communication industry and automobile manufacturing industry. It has full-stack AI model production and operation capability, and absorbs AI technology achievements from well-known universities.
Recalling her cooperation with Professor Chen Pengfei of Sun Yat-sen University, Wang Huan remembers that it was in the spring of 2019. "I heard Professor Chen speak at a government-sponsored forum. After the meeting, we talked and hit it off immediately." Since then, Wang Huan's team has collaborated with Professor Lai Jianhuang's team at CUHK. In his opinion, this is the industry-university-research advantages of starting a business in Guangzhou. "University research forms a technical prototype, and it needs to be verified in actual production. We have a lot of business scenarios and we need a stronger base to support our business products."
Such industry-university-research cooperation has undoubtedly set the wings for the development of enterprises to soar. Since its establishment in 2017, Netsys Technology, headquartered in Guangzhou, has set up 20 branches nationwide with more than 390 intellectual property rights, providing technical service solutions for many leading enterprises in the industry, such as China Mobile, Ping An Bank, GaC Group, Weichai Power and State Grid.
Wang Huan, who had worked in IBM for more than 10 years, realized that Chinese enterprises needed local technology companies to serve them, so that they could really listen to China's local industrial development, so he started his own business in the field of industrial Internet. After years of practice, he confessed that at present, the domestic industrial Internet has not been productized, more is the ability of middle Taiwan and business scenarios, the degree of productization is not high enough. "In the customer scenario, we still need to do development, do system coordination, and all kinds of interface modifications, but if we can productize 90% or more, then we only need to do parameter configuration."
To improve the degree of productization, not only needs the enterprise itself to improve the technical level, but also needs sufficient product iteration time to conclude, in addition, it also needs the overall industrial transformation environment. "This is also our direction, to be the core product of independent IP, through continuous technology research and development and business scenario iteration, to improve the technical level and product degree of our solutions, so as to better serve the industry and customers, so as to help customers achieve cost reduction and efficiency, and even industrial upgrading and digital transformation." Facing the future of industrial Internet, Wang Huan is full of confidence, "Our plan is to achieve a business scale of 2 billion by 2025, with self-managed delivery business and self-owned IP delivery business, and successfully listed."
