Recently, the Electronic Information Department of the Ministry of Industry and Information Technology and the Science and Technology Company of the State Administration of Radio and Television announced a list of typical ultra-high-definition video application cases to be selected. 15 typical cases have been selected in the industrial manufacturing field, fully demonstrating the wide application space of ultra-high-definition video technology in industrial visualization, defect detection, robot inspection, human-machine collaboration and interaction, etc.
Greatly improve the level of industrial visualization
In modern industrial production and manufacturing, industrial visualization can allow managers to discover problems in the production and manufacturing process at the first time and solve them in a timely and efficient manner. Therefore, building an industrial production visualization system has become the best choice for modern industrial enterprises, and ultra-high-definition video technology is the key to achieving visualization.
If a worker wants to do his job well, he must first sharpen his tools. Mass production and flexible automated manufacturing have increasingly placed demands on machine vision, and industrial lenses are a key component of machine vision in the industrial field. Compared with ordinary lenses, industrial lenses need to have higher definition and higherSpectral correction capabilities, stronger spectral transmission capabilities, etc. The 150 million ultra-high-definition pixel large-target lens launched by Hunan Changwei Optical Technology Co., Ltd. fills the gap in the small-magnification high-resolution lens market. This lens can be widely used for OLED detection, ranging from watch screens to 100-inch large screens; it can also be used for 3C and mobile phone lighting detection. "With a 150 million ultra-high-definition pixel large-area industrial lens and a 151MP camera, even micron-level screen defects can still be easily detected, making all defects invisible and greatly improving the production accuracy of the panel industry." Wu Peilin, CTO of Hunan Changwei Optical Technology Co., Ltd., told a reporter from China Electronics News.
The production scenes of high-end manufacturing are complex, industrial visualization is difficult, and the demand for ultra-high-definition video technology is also stronger. For example, in the chip manufacturing process, traditional wafer defect detection methods mainly rely on direct detection with microscopes and human eyes, which are costly, inefficient, and prone to subjective errors. Therefore, the application of machine vision technology to wafer inspection has gradually become a mainstream trend. The introduction of ultra-high-definition video technology can make machine vision inspection more efficient, more accurate, and more cost-effective.
The 8K ultra-high-definition wafer inspection image collection platform launched by Shanghai Jiqing Huishi Technology Co., Ltd. integrates a self-developed 8K ultra-high-definition camera, supporting lenses and customized light source systems, which effectively improves the accuracy and efficiency of machine vision technology in wafer defect detection and reduces costs. Extremely clear and wiseLin Lianghuang, director of the Visual Design Department, said in an interview with a reporter from China Electronics News: "The domestic integrated circuit and chip manufacturing industries are booming and require a large number of supporting domestic chip manufacturing equipment and testing equipment. This solution serves AOI visual testing equipment and has broad application prospects."
The application of a large number of ultra-high-definition video equipment has forced the data collection capabilities of machine vision systems to significantly improve. The application of ultra-high-definition video technology enables the machine vision system to obtain a larger amount of data, which requires higher efficiency in data analysis by the data application system. The efficiency and accuracy of the data application system in analyzing bad analysis are the focus of attention. BOE, which is active at the forefront of the global display field, extensively uses machine vision systems to control the appearance quality of products and equipment. It combines ultra-high-definition video and data analysis technology to launch an "industrial vision + data application" solution. BOE uses the computing power of the big data platform, analysis algorithms based on meeting business needs, and analysis methods that utilize experience in the display panel industry to form a set of root cause analysis systems (RCA) that solidify empirical knowledge and intelligently mine root causes. It has now been implemented in multiple BOE factories.
"The application of ultra-high-definition video equipment has improved the data collection capabilities of the machine vision system, which has led to increasingly higher requirements for processing detection data." Dr. Wang Hong, vice president of BOE and general manager of Zhongxiangying, said, "The application of ultra-high-definition technology has greatly improved the clarity of pictures collected by poor equipment.After processing by the ADC system to obtain accurate production defect data, the data is then imported into the intelligent defect root cause analysis system (RCA) for defect analysis, tracing, and prediction. As a result, the data analysis process is visualized, and the analysis results obtained are more accurate, allowing us to prevent defects in the production process before they occur. "In terms of precise production, the BOE Industrial Internet Platform can intelligently calculate production schedules based on production conditions, increasing overall production efficiency by more than 30%; in terms of product quality control, by collecting equipment-side parameters, building defect models, and opening up the production scheduling system, the product defect rate has been reduced by 75%, and the new product yield cycle has been shortened by 43%; in terms of industry chain collaboration, By connecting all links in the industrial chain and managing them collaboratively, inventory is reduced by 25%, order fulfillment cycles are shortened by 30%, and industrial chain costs are reduced by 25%, achieving efficient upstream and downstream collaboration. In just one year, the BOE Industrial Internet Platform has provided digital transformation services to more than 200 companies across the country, greatly improving the efficiency of corporate operation and management and accelerating the digital transformation of the industry.
Make defect detection more accurate and safe
Can go up to the sky and into the ground, and can also go into water, electric power, petrochemicals, coal mines, security, fire protection, transportation... Especially in some manpower shortage, high-risk, and difficult industries, inspection robots have become a much-anticipated "new star". Relevant statistics show that by 2026, China's intelligent inspection robot market is expected to reach 2.6 billion.Yuan.
As the hub of communication transmission, the safe operation of the operator's communication equipment room is crucial. In the computer room of Tianjin Mobile, the inspection robot system based on "5G﹢AI" ultra-high-definition video analysis methodically conducts engineering construction safety audits, equipment inspections, and remote inspections. The communication equipment defect detection situation in the operator's computer room is complex, and there are many identification methods such as instrument pointer identification, digital identification, status light identification, and connection line identification. The "5G﹢AI" computer room inspection robot can integrate ultra-high-definition video cameras, transmit images over the 5G network and perform AI analysis to accurately locate detailed information of objects, such as categories, locations and edges, and accurately identify abnormal conditions and issue alarms. It is understood that since its application, Tianjin Mobile's security inspection costs have dropped by 32%, the coverage scenarios have increased from 12 to 41 before, and the high-quality inspection completion rate has also increased from 35% to 65%. At present, this case has been promoted to factories, parks, substations, urban underground pipe corridors, power tunnel tube wells and other outdoor places.
In the automobile manufacturing process, the application of ultra-high-definition video technology plays the role of "inspector" to check the quality and safety of automobiles. For example, fluorescent magnetic particle inspection is often used to detect surface defects on automobile shaft workpieces. After the parts with cracks and defects are magnetized, the magnetic field formed by them will be uneven due to the defects. At this time, the adsorbed fluorescent magnetic powder will be more adsorbed than when the magnetic field is distributed uniformly, and will show green fluorescence under the irradiation of ultraviolet light, making the defectsvisible. Judgment by the human eye is subjective and will be affected by extreme environments, such as darkness, high temperature, high pressure, pollution, etc. Machine vision can solve these problems very well. The robot visual recognition system independently developed by Jinshi Robot Yinchuan Co., Ltd. combines ultra-high-definition video technology and uses machines to replace human eyes. It has been successfully applied in the Changan Mazda crankshaft production line of Changan Mazda Engine Co., Ltd., and has entered the stage of competing with overseas products.
The application of ultra-high-definition video technology creates advantageous conditions for the research and development of high-end testing equipment. As we all know, defect detection of electronic components, especially high-end semiconductor chips, is a major difficulty during the production process. In comparison, my country's research and development of high-end testing equipment systems has a short start, and there is a big gap in technical level, function, and efficiency. However, this situation has improved significantly in the past two years. With high-definition and high-precision imaging of 151 million pixels and above, a defect detection sub-pixel algorithm that combines traditional algorithms and deep learning algorithms, and a flexible and open software architecture, Hikvision Robot's ultra-high-definition, high-precision flexible automatic quality inspection system has been applied in 4K display screen production, circuit PCB production, as well as photovoltaic, lithium battery and other industries.
Zhang Zhenhua, product director of Hikvision Robotics, said in an interview with a reporter from China Electronics News: "This ultra-high-definition, high-precision flexible automatic quality inspection system based on vision technology has the characteristics of ultra-high resolution, excellent image quality, and algorithm intelligence. It can achieveUltra-high recognition accuracy and production line flexibility are widely used in various applications in high-end electronics manufacturing. It can help companies improve the automation and production efficiency of the production process, while improving product quality and added value. ”
Opening up the “eyes, nose and ears” of smart factories
In the era of Industry 4.0, building smart factories has become a trend for the transformation and upgrading of manufacturing companies. According to the Ministry of Industry and Information Technology, my country has built more than 700 digital workshops and smart factories so far, and the production efficiency of smart manufacturing pilot demonstration projects has increased by an average of 48%. The integration and application of ultra-high-definition video and new generation information technologies such as 5G, artificial intelligence, and virtual reality have spawned a large number of new scenarios, new applications, and new models, achieving both improvements in production and manufacturing efficiency.
Walking into Anshan Iron and Steel Co., Ltd.'s smart steel plant, the modern smelting production control equipment is refreshing. Workers who were once busy on the production line can now easily have a panoramic view of the entire steelmaking production process through the ultra-high-definition video monitoring equipment. They can also perform operations such as remote crane command, intelligent scrap identification, unmanned molten steel transportation, intelligent iron mixing in the converter, and ladle positioning, realizing a centralized, less-manned, automated, and intelligent steelmaking production process.
Zhao Qingtao, Chief Information Officer of Anshan Iron and Steel Co., Ltd. said: "The steelmaking process is complex and the production environment is highly dangerous. Visual management and remote production control have become essential auxiliary systems for modern steel companies. WeWe use ultra-high-definition video combined with image recognition and other technologies to automatically collect, record, and send production process-related data, effectively improving image recognition capabilities and on-site automation control levels in harsh environments, and playing an active role in reducing manual inspection operations, reducing production safety accidents, ensuring the safety of equipment and materials, optimizing production process management, and improving production management efficiency. "
In Jingzhou, Hubei, the world's first "5G fully connected factory" in the home appliance manufacturing field built by Jingzhou Mobile and Midea has been officially launched recently. More than a thousand highly efficient automated machines and equipment, 15 application scenarios, and more than 2,500 5G devices have been launched Terminal equipment such as CPE enables the entire production line to achieve true 5G full-process management of "full process links, full business scenarios, and full production factors." To achieve true "full connectivity," machine vision is a key part. In order to allow the machine's "eyes" to see more clearly and make more accurate judgments, the factory deploys ultra-high-definition cameras to collect images of PCB circuit boards and combines them with the large bandwidth of the 5G network. The low-latency feature transmits data to the image analysis system deployed on the edge cloud to automatically determine whether the PCB is defective and feedback the analysis results. It is reported that in the future, the factory will also deploy sorting robots to replace manual selection of defective products based on the results of the image recognition system. Wang Ting, project manager of Jingzhou Mobile, told a reporter from China Electronics News: "There are many other manufacturing plants in Jingzhou based on machine vision.With applications in detection, identification, measurement, and positioning, ultra-high-definition technology will incubate more business scenarios and empower enterprises with intelligence. ”
As early as 2019, the "Ultra-HD Video Industry Development Action Plan (2019-2022)" jointly issued by the Ministry of Industry and Information Technology, the State Administration of Radio and Television, and the China Central Radio and Television Station clearly stated that it is necessary to accelerate the development of ultra-high-definition industrial endoscopes, industrial cameras, and production line automatic inspection equipment. Industrialization of equipment, etc., promote the application of ultra-high-definition video technology in scenarios such as industrial visualization, defect detection, product assembly positioning guidance, robot inspection, human-machine collaborative interaction, etc., and create a number of typical applications that can be promoted around large-scale, high-precision industrial production scenarios such as electronics, automobiles, and aerospace.
Nowadays, with the guidance of national policies and the joint efforts of industry, academia, research, and users, ultra-high-definition video technology has become a spark in the field of industrial manufacturing, and its vigorous scientific and technological power has given a pair of invisible wings to the industry. With dreams as horses, the future is promising. From "Made in China" to "Intelligent Manufacturing in China" and "Created in China", the dream of industrial power is within sight.
Contact: Mack
Phone: +8613352972563
E-mail: mack@archled.net
Add: Building A2, Mingjinhai Second Industrial Zone, Shiyan Street, Baoan, Shenzhen,Guangdong,China
ANNA