Where should intelligent manufacturing be?

For any industrial product, defect is an important factor related to the survival of enterprises and even social security. Although defect detection is very important, it becomes more and more difficult with the increasing precision of products. Based on the deep understanding of the development of the industry, Intel has provided a set of effective solutions for industrial product defect detection through cooperation with Shenzhen vision technology, helping enterprises to see far and wide.

            Han feiziyun, a thousand Zhang dike, crumbles in the nest of ants; a hundred feet room burns in the smoke of a sudden gap. During the Warring States period, Bai Gui, the Xiangguo of Wei state, was good at building embankments and flood control, and attached great importance to small caves, even ant caves. Under its rule, Wei never had a flood.            The truth contained in this is also applicable in modern times - in the field of manufacturing, for any industrial product, defects are important factors related to the survival of enterprises and even social security. 

           Although defect detection is very important, both enterprises and governments have invested a lot of human and material resources in it, but with the increasing precision of industrial products, defect detection has become more and more difficult. The accuracy of traditional manual inspectors' judgment is usually between 90% and 95%. With the increasing of the number of products, the number of unqualified products flowing to the market is increasing rapidly. 

           Taking the vehicle industry as an example, according to the statistics of the Ministry of public security, by the end of 2018, the car ownership in China had reached 240 million, an increase of 22.85 million over 2017, an increase of 10.51%. It's not just cars. With the continuous improvement of mass consumption, the number of various industrial products is growing exponentially. 

           With more and more products, it is almost impossible to achieve complete and accurate quality judgment only by manpower, and more and more unqualified products flowing to the market will also affect more families, and the harm is more and more serious. Take the popular "China captain" as an example, the unexpected product damage will affect the lives of more than 100 people. Although the pilot's personal skills will eventually land smoothly, no one hopes to have an accident. 

           So, people think, with the advent of the intelligent era, can we use AI, Internet of things and other technologies to upgrade the product quality detection process?  

           Using AI to solve the problem of defects in industrial products 

           Thanks to the continuous enhancement of deep learning, computer vision and image processing technology, it seems possible to use AI to prevent production defects. 

           Industrial machine vision is widely used in industry. Its core functions include measurement, detection, recognition, positioning, etc. Among all the functions, the industrial surface visual inspection technology is the most widely used. 

           Industrial surface vision detection technology is based on machine vision. Through the use of computer simulation of human vision, it can collect, process and calculate images of specific objects, and finally carry out actual detection, control and application. At present, the industrial surface visual defect detection system has been widely used, which effectively promotes the high-quality production of factory products and the intelligent manufacturing industry.            China's industrial machine vision started with the introduction of technology in the 1980s, and then in 1998, the whole line of semiconductor factories also brought into the machine vision system. However, before 2006, China's machine vision products are still mainly concentrated in foreign manufacturing enterprises, with small scale. 

           Since 2006, the customer base of industrial machine vision applications has gradually expanded, and the detection fields such as printing and food have taken the lead in the application of industrial machine vision technology. Until 2011, the industrial machine vision market began to grow rapidly.

            With the increase of labor cost and the upgrading demand of manufacturing industry, coupled with the rapid development of computer vision technology, more and more machine vision solutions have penetrated into various fields. By 2017, the size of China's machine vision market has reached nearly 7 billion yuan.

            Up to now, there are more than 102 machine vision enterprises in China, more than 200 machine vision product agents, more than 50 professional machine vision system integrators, and the industrial chain of all visual inspection equipment from light source, industrial camera, industrial lens and processing software is becoming more and more complete.

            It is based on the deep understanding of the industry development that Intel provides a set of effective solutions for industrial product defect detection through cooperation with Shenzhen vision technology.

            Not long ago, Shenzhen vision technology and Intel used their solutions to help a leading tire manufacturer solve some defects that could not be seen by human inspectors before:

            Faster detection. Compared with the traditional manual inspection, the detection speed of the solution is significantly improved, and the average time of each inspection is less than 1 second.

            The detection accuracy is improved. Thanks to the deep vision technology solution based on Intel technology, the tire manufacturer has improved its defect detection accuracy to more than 99.99%, and can inspect more than 20000 tires in real time every day.

            Increased profitability. Based on faster speed and higher accuracy, the company's net profit of the tire manufacturer is greatly improved, and the average labor cost per production line is reduced by about $42000.

            Of course, such a real ability improvement is inseparable from the deep accumulation of deep learning and powerful computing capabilities of Shenzhen vision technology and Intel.

            Tracing back to the origin, from the focus

            As far as the principle is concerned, the solutions of Intel and deep view technology are firstly collected by the imaging system, then transmitted to the edge devices, real-time detection and analysis are carried out with deep inspect software of deep view technology, and finally the results are transmitted back to the production line for classification.

            In addition, the solution also uses the test equipment of Shenzhen vision technology to replace the original test equipment, and can directly communicate with customers through the interface configured by the equipment

            Please read the Chinese version for details.