Why AOI machine so expensive?

Let me answer first, yes. Artificial intelligence is sweeping almost all industries because its applications are too numerous and widespread. AI not only brings innovation in AOI technology, but also causes many industries to face the unemployment problem of human replacement.

Since the advent of LeNet in 1998, CNN has been constantly evolving. Many new networks such as AlexNet, VGGNet, GoogLeNet, ResNet, etc. are very well-known models. Most of the current CNN models are composed of quite deep and wide networks. Their ability to process images is already very powerful, and the success rate of identifying defects can almost reach more than 95%. For AI, processing images and classifying them is just a piece of cake, not a big problem. Where AI will really change AOI in the future is not simply defect identification.

Generally speaking, a production line will need to use AOI for screening. Most of them have a very large output, or the product quality control is very strict. Representatives of this type of industry include PCB, semiconductor industry, mobile phone parts, medical equipment, etc. Most of these production lines are products with high unit price. When the production line is in production, it is very important to improve the yield rate. The role of AOI here is not to simply inspect and eliminate defective products, but to be a data collector. We obtain a large amount of defect data from AOI, and through reasonable analysis and induction, we can find out the reasons for defective products in the process.
As for AI, data analysis and data processing are also his strengths. As long as the Recipe (Input) and defect data of the production line are input into the network, after continuous iteration, AI can analyze the key parameters or inputs that really affect the yield, help us find the problems that cause defects, and even simplify Other processes.

With the evolution of the times, the industrial application of AOI has become more and more common. If AOI is still just AOI without comprehensive analysis of AI data, it will surely be gradually replaced by other industries.