The following article comes from Saying East and West, written by Song Huazhen
The value and significance of digitalization lies in how to respond to VUCA's changes in manufacturing. It is to transform the complex and uncertain, ambiguous and volatile into a "stable, certain, controllable and flexible" state through digital technology.
- Article Information -
The author of this article, Song Huazhen, is the technical communication manager of B&R Industrial Automation (China) Co., Ltd. This article was originally published by the official account "Say Dongdaoxi", and is published by digital enterprises with permission.
Movie star Julia Roberts, who is also a mahjong enthusiast, talks about what she understands to be mahjong as "creating order out of chaos by drawing cards at random." This sentence reminds me that this is also the essence of digitalization – creating order out of chaos.
People call the era we live in "VUCA" – volatility, uncertainty, complexity, ambiguity. Its most striking feature is the "dramatic increase in complexity". This escalation of complexity often leads to exponential effort. And, the time granularity of the evolution of this complexity is getting smaller and smaller.
The more intuitive effects of complexity include the following:
01 The complexity of product materials and their processes;
02 The complex combination of production processes, especially those long processes in discrete manufacturing such as photovoltaics, lithium batteries, semiconductors, etc.;
03. There are more and more parameter dimensions that need to be monitored - to gain "observability" of the process and thus to gain controllability;
Figure 1 – Increased complexity in the context of VUCA
As Figure 1 shows, complexity challenges the manufacturing process in several ways:
01 Changes in production, configuration during reorganization, complexity of the combination of process parameters, if the traditional manufacturing mode, there will be a lot of adjustments, and usually the adjustment will cause more "start-up waste".
02 TTM-personalized products brought by personalized products make more time consumed in the production process for "parameter configuration" and "mechanical adjustment", and the more frequent the replacement, which will cause a decrease in delivery capacity. Because, in a day's working hours, there may not be many effective working hours for the machine.
03 The complexity of system communication and coordination in the integration of long-process production lines: the change of products is not only the change of tooling and fixtures of a certain unit, but also the change of parameters. Instead, the equipment of each process in the entire production process should be coordinated in a unified manner, and the logistics, cost accounting, and energy supply changes at the front and back ends of the manufacturing line should be changed to maintain this consistency.
Value analysis of manufacturing digitalization
The value and significance of digitalization lies in how to respond to the changes in VUCA's manufacturing industry - it is to transform complexity and uncertainty, ambiguity and volatility into a "stable, certain, controllable and flexible" state through digital technology.
Therefore, the value of digitalization includes at least the following aspects:
01
Software-Defined Manufacturing - Digitalization makes manufacturing more flexible
The significance of digitalization is its flexibility, like a traditional machine, if it wants to change the product, it often needs a lot of mechanical adjustments, such as binding a book, when we want to bind a new book, it needs to adjust the width of the book, the height of the book, the position of the glue, the pressure, etc. When it is changed to a digital servo system for tuning, it only needs to set the parameters and the system will do the work automatically. This traditionally takes 1.5 hours, but now it takes only 2 seconds – thus increasing the changeover time of the machine in the midst of changing productions.
Figure 2 - Electronic cam cutting
In terms of the most common cutting action in automation, it is widely used in carton forming, label printing, packaging, sheet metal processing, paper processing (such as A4 copy paper), corrugated paper cutting, etc. Traditionally, the mechanical tool roller needs to be replaced according to the product specifications, but after the electronic cam cutting of the servo system is adopted, this can be set according to the length by itself, and the servo motor will run the electronic cam by itself to achieve precise cutting. In this case, the change in specifications is defined by the software.
Another scenario is that in the assembly line, a large number of magnetic levitation conveyor technology is currently being used. The purpose is to change the mechanical conveyance, which could not have been controlled by software, to an electromagnetically controlled mover – and it can make the production system software-defined. In turn, change is made to respond to the need for flexibility.
Figure 3 - Traditional mechanical conveying mechanism
In the diagram, the mechanical conveying of chain blocks, belts, or indexing plates is widely used. The problem with a single product is the potential impact of mechanical wear on quality consistency, but even without taking into account product variations, it is still inefficient because the cycle time of each station is different, which makes the system run at the slowest station cycle. If product variations are taken into account, the cycle time of each station changes again, and the machine may need to be significantly adjusted for the alignment of the machine (e.g. to keep the center point matched). As a result, when the product changes, or even when the production line needs to be reorganized, it is necessary to encounter major mechanical reinstallation and adjustment problems.
Figure 4 - Multi-dimensional motion control
However, with the use of magnetic levitation conveyor technology, the position, spacing, velocity, and acceleration of the processed product can be defined by software when it is conveyed. It is a conveyor technology designed by combining electromechanical objects, so that it can be used to implement multiple track mixing operations on its own, as well as scheduling algorithms for intelligent scheduling – and the simplified machinery also makes it easier to reorganize.
As shown in Figure 4, in fact, the implementation of software-defined motion control includes not only servo motors, linear motors, and magnetic levitation, but also a large number of robot integrations – it also configures components to the desired position in the space through its own parameter configuration. It's all about making manufacturing "software-defined."
This is the only way that complex changes can be handled simply – through fast software configuration, automatic parameter generation, and network distribution to the process equipment on the line. And it can realize the mixed line scheduling of combined products, and realize diversified product production in an efficient way.
02
Collaboration - Make it more efficient
In fact, the continuity of process industries, such as petroleum refining, metallurgy, pharmaceuticals and other process industries, makes it inherently necessary for better automation control. Discrete manufacturing, in the past, has been really "discrete", with equipment moving from one unit to another according to a certain layout. Therefore, if these productions can be made as continuous as processes, then production efficiency will be higher. Therefore, the first step is to use a conveyor belt to form a continuous production line with robots, but later found that the production line of these machinery can not be effectively organized, and people have developed a flexible conveyor system, the essence of which is to digitize the production process, and turn the original mechanical dumb system into a digital manufacturing system that can be collected, transmitted, and programmed by software.
Collaboration on machinery includes the introduction of robotics and flexible conveyor technology. At the software level, it is the integration of communication, which often represents a connection in a physical sense, while communication represents a connection in the sense of software. Therefore, digital communication, which is also a collaboration problem, is achieved through information modeling, in which modular electromechanical systems will be driven by state machines. This is also the meaning of collaboration in industrial communication systems.
Figure 5 - Industrial communication is all about collaboration
As shown in Figure 5, the purpose of the communication protocol is to enable the system to be quickly adjusted in the event of engineering changes through information modeling. Parameters can be parsed and delivered in a unified manner. In the equipment room, state-based "logical orchestration" can be realized, so that the production system can quickly enter the production of new products.
The strategy for solving complex problems is to "simplify the complex", with each independent production module acting as a highly cohesive software module that achieves "rapid reorganization" through low coupling. In communication such as PackML and SEMI, the mechanism of collaboration through state is to transform complex collaboration into "logical" orchestration.
03
Knowledge discovery and reuse
The difference between knowledge and experience is that knowledge can be described by mathematical formulas, while experience is hidden in the brains of engineers and technicians that cannot be described. Therefore, for knowledge to be reused, it needs to be softwareized, and for experience to be reused, it needs to be mined through data-driven modeling—this is the mathematical meaning of AI.
Figure 6 - Translation of knowledge and experience into software reuse
As shown in Figure 6, the purpose of industrial software is to encapsulate "knowledge" and "experience" through physical modeling and data-driven modeling, respectively. This knowledge can then be reused so that the production system can be reorganized at the software level based on these modules.
Software is the key way to maintain knowledge inheritance, and by "digitizing" knowledge, it can be flexibly integrated, quickly migrated, and reused across different platforms. This is also about the complexity of the manufacturing process – modular machines and, above all, the modular configuration of software. In addition, the cost of knowledge reuse is reduced.
04
Ongoing cost reduction
Digital design and digital operations can identify potential waste, improve and improve cost efficiency. Digitalization can help us reduce costs because there is a huge amount of waste hidden in the production system that we take for granted, and from the perspective of lean production, this waste is ubiquitous, so a lot of cost-cutting solutions are needed.
1). Reduce the cost of physical test verification
2). Problems that reduce operational efficiency
3). Continuous waste improvement
05
Resource sharing
It includes code and algorithm resources in the digital world, network resources (open), and knowledge resources (such as openAI for generative AI). Digital resources are easier to share than a loaf of bread because a loaf of bread will spoil, while digital resources can be easily maintained and replicated.
Digital assets are also easier to work with, just as film can be used to produce photos, but photos in film are not as easy to process as digital photos with software such as Photoshop.
Figure 7 - Generative AI can share a wider range of resources for engineering development
Therefore, digital information is easier to store and share over a long period of time.
Digitalization has made the "sharing" economy possible – for manufacturing, resources from the IT world, whether network or algorithmic, can be shared. can be used for remote diagnostics and maintenance (via ubiquitous cloud services) – with the benefit of eliminating the need for a dedicated self-built LAN. Algorithms in the IT world, such as open source code, reduce the amount of work in code development, especially the use of generative AI in programming like today.
06
Smarter systems
In the manufacturing system, initially, the human brain realizes the control of production. It was later replaced by a controller with a defined program, which had the advantage of setting up the potential possibilities, using the program to implement an "If-Then" statement to judge various scenarios, and then invoking the existing processing unit. Therefore, in the past, in large-scale standardized production, it was a production process under established rules.
However, with the change of product variety, materials and their processes, it is no longer possible to make all the rules, or the existing knowledge is not enough to define the parameters of the various changes. Then, people need this system to have the ability to "learn", to evolve better production adaptive parameters, and to evolve better recognition and judgment capabilities. In place of the powerful uncertainty of human empirical judgment, as well as the inconsistency that changes with people.
Figure 8 - 8 major scenarios of AI in the manufacturing industry
Figure 8 lists the so-called "eight scenarios" of AI in manufacturing—in fact, we can see that it is all about making the system smarter. Visual defect analysis, machine failures, intelligent sorting – these are all things that are learned to enable the machine to adapt to changes.
Take intelligent sorting as an example, in the past, it was usually necessary to cooperate with a mechanical mechanism at the front of the production line to sort out the product queue, such as bottles with unsorted bottles and metal parts with mechanical ways to keep them in a consistent direction. However, when there are more types of products, the adjustment of this machinery will be complicated, and the robot will identify the products of different shapes under the guidance of vision, and confirm the center point or center of gravity of the grab, which can be learned to achieve the sorting of any product - this is a "smart" production line. There is no need for complex machinery and a constant teaching process.
Therefore, the future of intelligent systems is to allow machines to iterate on their own in order to gain evolutionary capabilities that do not depend on human input.
07
Faster responsiveness required
Digital systems are able to provide faster response times for several reasons:
01 Information interaction speed: Through faster information transmission, problems can be quickly collected, and then quickly analyzed and judged. This time granularity is smaller, resulting in improved responsiveness.
02 Processing capacity: If a task is orchestrated, then for people, it can be calculated, but through the digital system, the calculation and re-planning of this arrangement can be quickly realized.
03Faster iteration: Continuous improvement is extremely fast for digital systems. Because, in the iteration of the digital system, it can be improved in the next step.
This is the problem that the real-time network needs to solve - the bottom layer of digitalization is, of course, the support of network communication, whether it is within the equipment, between equipment, production line and production management, and cloud systems, all need a faster time-granular network to support.
Figure 9 – High-speed real-time communication networks support digitalization
As shown in Figure 9, the TSN network and OPC UA can achieve a high dynamic response capability of the entire manufacturing system. In short, the significance of digitalization is of great significance in the manufacturing industry - in order to give enterprises the ability to respond to changes in a complex and changing environment. Flexible, efficient, and smart.