[Big Talk Intelligence] As smart factories are underway, how will artificial intelligence boost Industry 4.0?

When Industry 4.0 meets Smart City , Smart Factory, as a key topic throughout the two future projects, is naturally mentioned repeatedly by industry leaders along with the general trend of global industrialization development.

In recent years, the manufacturing industry has been exploring and practicing smart factories. Labels such as "unmanned factory, lights-off workshop, robot factory" have become representative words of this trend.

▲Tesla's fully automated robot factory

But is this really the case? Can we identify a factory as a smart factory just because it has the characteristics of unattended and fully automated ? I think it is not that simple. There is an essential difference between automation and intelligence . Therefore, artificial intelligence is indispensable to realize a truly smart factory !

▲【Dahua Intelligence】What are the characteristics and cutting-edge technologies of the world's top smart city construction?

I have previously introduced classic practice cases of foreign smart city construction (click on the picture above). Today, I will focus on the field of smart factory and analyze how artificial intelligence technology can help us build smart factories in the context of Industry 4.0 with several practical cases!

Building a smart factory means that we should first consider intelligence rather than automation , that is, giving industrial robots the ability to analyze, think, and reason like the human brain, so that even without human supervision, they can make reasonable decisions based on the current situation.

You know, in a huge factory, there are countless machines and workers working non-stop every day. It is not an easy task to keep track of the operation of the production line at all times. But now with the artificial intelligence "deep learning technology" , everything has gone from "impossible" to "potential"!

The factory builds an artificial intelligence deep learning framework to build a data interconnection, equipment Internet of Things and human-machine collaboration production system. With the support of AI deep learning technology, artificial intelligence will continue to learn knowledge and accumulate experience through big data.

Ultimately, AI will help promote the upgrade of Industry 4.0 and help factories gradually evolve into smart factories where industrial robots lead production execution, recognition sensor systems provide comprehensive monitoring, and intelligent data systems continuously optimize management .

Although the current technical conditions are not sufficient to achieve this ultimate goal, some companies are already at the forefront of exploration, such as Bosch Group, the world's largest supplier of automotive technology.

Not only does its ABS automobile brake system have a high market share in the world, its representative smart factory, the Bosch Humboldt Factory , is also a global case-level smart factory.

The special feature of this production line is that all parts have unique radio frequency identification codes that can automatically "talk" to the checkpoints . After each production link, the central artificial intelligence control center can receive information feedback. AI will use the experience accumulated by deep learning technology to process these feedbacks accordingly, thereby improving overall production efficiency.

Not only that, because each part has a unique radio frequency identification code, it means that inventory can be tracked and updated dynamically in real time, allowing resources to be fully mobilized and avoiding waste caused by backlogs of inventory.

So it seems that although the introduction of the radio frequency code recognition system costs a lot, it is not worth mentioning compared to the high returns brought by cost savings and efficiency improvements from smart factory upgrades. It can be regarded as one of the solutions with a very high input-output ratio.

Although machine vision technology is no stranger to traditional production lines, traditional machine vision systems rely solely on preset algorithms for judgment, and their application scenarios are very limited. For example, they are helpless against irregular defects such as cracks, scratches, and stains.

In this case, by introducing artificial intelligence technology into the machine vision system and building a powerful model through deep learning reasoning, it is possible to detect irregular defects of the above products. Moreover, as more and more samples are collected, AI will continue to learn to improve the detection accuracy and gradually achieve fast and stable detection, thus solving the problem of unstable accuracy of manual eye detection.

If AI deep learning technology gives smart factories a "brain", then it is AI machine vision technology that opens the "eyes" of smart factories!

Whether it is checking whether the product quality on the production line meets the standards or positioning parts with micron-level precision, with it, smart factories can perform various "high-precision" tasks and complete various tasks that are beyond human control.

To give a typical example of the use of machine vision systems in industrial inspection, ROVER Automobile Company from the UK pioneered the use of a 100% online inspection system when inspecting the contour dimensional accuracy of its 800 series automobile bodies.

The system consists of 62 measuring units, each of which includes a laser and a CCD camera to detect 288 measuring points on the body shell.

(Online reference material only)

To ensure detection accuracy, each laser/camera unit is calibrated offline. At the same time, professional instruments are available that can be calibrated at any time and adjusted online, providing a double guarantee.

During the formal inspection, this machine vision inspection system will inspect three different types of car bodies at a speed of one car body every 40 seconds. The final inspection results will also be compared with the traditional measurement results of personnel and the dimensions exported from the CAD model .

(Online reference material only)

ROVER's quality inspectors plan to use this machine vision inspection system to determine whether the dimensions of key parts such as the overall appearance of the car body, doors, and glass windows are consistent. Of course, the final practice has proved that this system is successful, and it will continue to be used in the inspection of other series of ROVER models.

Looking back, the three industrial revolutions in human history have realized the evolution of industry towards mechanization, automation and informatization respectively. Now we are gradually ushering in the fourth industrial revolution. With the continuous development of artificial intelligence technology and its continuous implementation in various industries, the future undoubtedly belongs to artificial intelligence.

▲The American TV series "Westworld" depicts a different AI future

In the intelligent era, productivity and production methods will undergo fundamental changes. As productivity continues to upgrade, we will find that the smart factory extended from Industry 4.0 complements the concept of smart city that governments are currently actively promoting .

Therefore, it is foreseeable that in the future, smart factories will work together with smart medical care and smart transportation to promote the construction and improvement of smart cities, bringing us a more environmentally friendly, efficient and intelligent smart life.

After reading this article, are you more interested in the field of smart cities? You are welcome to click on the picture below to enter the [Key Areas] of WPG World for more cutting-edge smart city information and opinions from senior industry experts. Full of useful information is waiting for you!

The pictures used in the above article are all from the Internet; if you need to reprint the article, please leave a message in the WeChat background!