Following the inauguration of the Siemens Global R&D Center for Industrial Software in Chengdu Hi-Tech Zone a year ago, the Siemens Innovation Center for Smart Manufacturing in Chengdu was officially launched on May 21st at the same location.
The innovation center will integrate innovation and R&D in smart manufacturing, bring together innovation in discrete and process manufacturing hardware and software technologies, and become a demonstration base for innovation centers and digital factories in the same city.
Speaking of Siemens' digital factory, it is necessary to mention that in 2011, Siemens (China) launched the construction of Chengdu production and R&D base for industrial automation products in Chengdu Hi-Tech Zone. This base is the largest digital factory of Siemens in China, the third R&D center for industrial automation products in the world after Germany and the United States, and a typical representative of Siemens' digital factory practicing Germany's "Industry 4.0".
Rarely seen forklifts, the workshop downstairs hides the "conveyor belt"
Visiting Siemens Chengdu plant, where the whole process is not allowed to take photos, all the doors need to swipe the card to enter and exit, the management is very strict.
Li Yongli, the person in charge of the research and development base, first led us to visit the first floor of the small flat, a bit like an underground garage, can not see a person, only the conveyor belt, the product in the integration of the tireless flow of the railroad.
Upstairs is the production workshop, almost all the materials are transported through these conveyor belts on the first floor, and those that need to be sent to the workshop are transported by a straight elevator.
Therefore, in the entire Siemens Chengdu plant, there are few forklifts, and few piles of materials and semi-finished products in the workshop, all waiting or processed materials are quickly transferred to the conveyor belt downstairs.
On the wall of one side of the production workshop, various products of Siemens factory are displayed, including industrial control products, PLC, HMI, etc. One of them is a small box, which is not too conspicuous, and it is mainly used to collect the data of various terminals and devices in the industrial scene, and then uploaded to the Siemens MindSphere cloud platform.
Data first, every material carries an "ID"
On the assembly line, if an operator needs material, he just needs to click on the computer display on the workbench, and in a few minutes, the required material will appear in the designated place automatically.
This has been a daily occurrence at the Siemens Chengdu digital plant for more than two years. So how is this information being delivered?
On the back of each material, there is a one-dimensional code affixed to it, and there are more than 100,000 one-dimensional codes throughout the factory, which is like a cab number, how many passengers are loaded in it, and when they get on and off, and so on, all of which are recorded.
This is like an "ID card", from the date of production to enter the various aspects of production, there is a large amount of information stored in it; is also like a personal resume, lifelong traceability, and only with this data, we can follow up on the prediction, optimization to make more improvements.
The data generated by the R&D process is transmitted in real time between the various systems in the factory, and the synchronized updating of the data also avoids the errors that often occur in traditional manufacturing companies due to miscommunication, which greatly improves the efficiency.
An advanced factory is still using one-dimensional code? Li Yongli explained, in fact, one-dimensional code is enough, and the cost is lower, some things look very advanced, but one-dimensional code is actually the most effective.
The most advanced factories are not directly buy the most advanced equipment
In the view of Willy Lee, you want to rise the quality of the product, it is necessary to improve the degree of automation, because automation is not wrong.
Increase automation, to achieve state-of-the-art production means directly buy the most advanced equipment? "In fact, it is not, because next year's equipment, it will have to be more advanced than this year, then you are not immediately to change the new equipment? The right way of thinking should be to continuously optimize the original equipment."
In the Chengdu factory, there are a lot of grasping robots from Kuka, but Li Yongli said, it is not that buy the most advanced robots will be able to do the job well. It is like everyone's computer, now the production technology of the computer is not significantly different, which manipulation to do with the computer, which is the most critical.
Another example is that many factories already exist, so is it necessary to knock down and rebuild in order to realize the latest production methods? Of course not, but rather a mathematical modeling refactoring of what can be upgraded and what can be eliminated and made with processes.
In this process, it has moved the factory to the computer model to run, in layman's terms, bought a robot, there should be a mathematical model of the robot, and the machine works together collaboratively, and his every move should also be described by a digital model.
Problems occur - people solve problems - the knowledge and process of solving problems is solidified into the equipment and production line - similar problems are automatically solved or avoided. ...... This became a major thread in German problem solving.
Before the project was launched in the online simulation "run"
Next, Li Yongli introduced the "digital twin".
In the field of data storage, there is a technical term "mirror", the so-called "digital twins" and the concept of mirroring is very similar, it is the actual product manufacturing, production processes, and even the entire management of the factory for the The so-called "digital twin" is similar to the concept of mirroring, in that it mirrors the actual product manufacturing, production process, and even the entire factory management, reproducing the testing and production process in a completely virtual environment.
Li Yongli said, the first generation of product development before the first simulation test, verification, "many problems are not measured to find out, but to test almost all the problems before embarking on the project."
"How long does a screw take, the computer will be able to test out." Yong Li said.
And then, for example, the robot arm to take the material out of the box, and then put it on the stage to test the processing, and then put it back in the box, seemingly simple actions, all need a lot of optimization.
In the optimization of the discovery, the machine arm to grab the material inserted into a slot before the material with a one-time swing, the need to pause for a while, with the material still, in order to achieve accurate insertion, but this is the imperceptible pause, the production line of the seconds and seconds of the production line, it is easy to cause the subsequent production line "traffic jam", and therefore need to be optimized. Optimization.
In his view, many people are now talking about information, process automation, which is a feature of the global manufacturing and supply chain, in fact, many times the manual tedious work is constantly replaced, the development of a few years after the formation of a number of gadgets, which is only the primary stage of digitization, if upgraded to the platform, these tools can be formed to collaborate to achieve the cooperation between the platform.
Product quality problems can be traced back to the source
Li Yongli said, the manufacturing industry is the most difficult to improve the quality of many say that the quality of their products can reach 99.999%, but in fact, this refers to the quality of the product when the product is shipped, rather than the quality of the product in the production of the product, the theory of the product quality when the product is shipped should be 100%, even if the product is sold for a billion pieces of product with a single product, the quality of the product should be 100%, even if the product is sold for a billion pieces of product with a single product. Theoretically, all products should be 100% quality when they leave the factory, and even if 100 million products are sold with one bad one, it is unacceptable.
Li Yongli quoted a word called "error rate" (dpm), that is, one million components in all the errors, in April this year, Siemens Chengdu plant dpm for two million, good in all the data can be backward, such as the production line to paste a certain component when the 12 errors, other parts of the error were also respectively
"Since we have found the problem, the next step should be to optimize, so the next time is not the dpm can be reduced to 1.9, 1.8?" said Li Yongli, the error rate is so a little bit lower.
The products produced, on average, every two days to the Siemens factory in Germany, Suzhou, and these products are produced, stay in the Chengdu factory time, the longest will not exceed 12 hours.
Before shipping, these products are stored in a three-dimensional warehouse. This three-dimensional warehouse is also the last stop on our tour, up to dozens of meters of containers, a divided into seven lanes, each lane is 17,000 pieces of products, the whole more than 100,000 pieces of products, all black, as if the scene in the blockbuster.
By Gao Xiaoping
Source Southern News Media Group South + Client