The new industrial revolution is coming. In recent years, many countries have put forward their own industrial transformation and upgrading development strategies and names, such as industry 4.0 in Germany and intelligent manufacturing in China. However, for the basic connotation and characteristics of this global new industrial revolution marked by intelligence, there have been different opinions and different wording. There are many definitions of intelligent manufacturing, each with its own source. If we can't clearly understand the new and accurately grasp the industrial revolution  /  With the basic characteristics of intelligent manufacturing, it is difficult to really and effectively promote the capital construction of the new industrial revolution, and it is difficult for intelligent manufacturing to be implemented in enterprises.  
 
 The author combs the connotation of the new industrial revolution and summarizes its five basic characteristics:  
 (1) "human intelligence" to "wit";  
 (2) low price popularization of sensors;  
 (3) software definition and manufacturing;  
 (4) real integration of two technologies (software closed loop);  
 (5) optimize the allocation of manufacturing resources on a large scale.  The following are introduced respectively.  

1. "Human intelligence" to "wit"

Human knowledge continues to enter software, the knowledge carrier has changed from carbon based knowledge to silicon-based knowledge, and the digital productivity has increased sharply.  
 
 Intelligent manufacturing is not an unfathomable and incomprehensible term. Generally speaking, the so-called intelligent manufacturing is a process of "human intelligence changing into wit" - that is, human intelligence (hereinafter referred to as "human intelligence") is refined from tacit knowledge into explicit knowledge, modeled and algorithmized, and then various modeled knowledge (mechanism model, data analysis model, etc.) is embedded in software, software is embedded in chip, and chip is embedded in a box  /  Module, and then  Box  /  The module is embedded into the physical equipment, which gives the machine a certain degree of autonomy and makes the machine have a certain degree of "intelligence" (hereinafter referred to as "wit"). We call this process "empowerment".  
 
 In this way, the machine has a certain human thinking ability under the support of software. When the software algorithm is better, the computing power of the chip is stronger, and the industrial data is more, the degree of "wit" is higher. Therefore, when the "wit" reaches a certain degree, it can partially or completely replace the human body  /  The function of the human brain. When human body  /  After the human brain leaves the system loop of the work scene, the machine can still work autonomously, even better, and optimize the allocation of manufacturing resources without human participation. As shown in the figure  one  As shown in.
 
 
 

chartone The process from "human intelligence" to "wit"

This kind of manufacturing activity that transforms "human intelligence" into "wit" is called intelligent manufacturing by the author. It is not difficult to see from the above described process that the key enabling element is industrial software.  
 
 Industrial software is the top product of industrialization. It encapsulates industrial technology  /  Knowledge, the automatic data flow rule system is established, and the brain and nerve of the machine are created. Therefore, the machine becomes smarter and its functions can be adjusted at any time. Industrial software describes, integrates, simulates, accelerates, amplifies, optimizes and innovates the traditional manufacturing process, forming a new industrial intelligence mode - software defined manufacturing.  
 
 The first and second industrial revolutions greatly liberated the human body. The third industrial revolution has greatly liberated the human brain. The principle of liberating the human brain is that a large number of "human intelligence" enter the software. Therefore, the means of R & D and management, digital software, has revolutionized product R & D and management, and new concepts such as digital twins and data analysis model emerge one after another; At the same time, the digital software of the product itself has revolutionized the machine.  forty  Years ago, there were few wires on the equipment except the power cord. In today's equipment, there have been a large number of wires, which are connected to some boxes on the machine. In these boxes, there are multilayer circuit boards and a large number of chips. The chips run software, which contains a large number of "human intelligence", and these "human intelligence" have been transformed into "wit".  
 
 If we can turn "human intelligence" into "wit" and turn traditional machines into "intelligent machines", if we can develop more autonomous and controllable  Industrial software  , if industrial technology  /  If knowledge can be more software, it is possible to liberate a large number of knowledge-based technicians from repetitive labor, let machines generate products, let liberated technicians produce knowledge and manage machines more easily, so as to realize the re division of labor between people and machines and realize industrial technology  /  The continuous accumulation and inheritance of knowledge enable enterprises to achieve sustainable and high-quality development.  
 
 

2. Low price popularization of sensors

Sensors add "five senses" to the machine, greatly enhance the state perception ability of the machine to the environment, and accelerate the digitization of physical spatial information.  
 
 Without "state awareness", a machine cannot become an intelligent machine  （  Even if "state perception" is realized, other conditions must be met at the same time). To realize "state perception", we need a variety of "sensing devices". Whether intelligent manufacturing, industrial  four  , industrial Internet or Internet of things are inseparable from all kinds of sensors.  
 
 The sensor is provided by the national standard  GB7665-87  It is standardized and defined as "a device or device that can feel the specified measured signal and convert it into usable output signal according to a certain law, which is usually composed of sensitive elements and conversion elements". The "available output signal" refers to the electrical signal that is convenient for processing, transmission and utilization. Generally, sensors are composed of sensitive elements, conversion elements and conversion circuits.  
 
 Sensitive element  That is, the "substance with inductive function" as the "effective substance" of the "inductive device". Therefore, the sensor is developed from the "effective material" with "material with sensing function".  Without scientific effect (physical effect, chemical effect, etc.) substances as sensing devices, there will be no sensors  。  
 
 Common sensors include thermal sensor, photosensitive sensor, temperature sensor, force sensor, gas sensor, humidity sensor, sound sensor, magnetic sensor, taste sensor and radiation sensor. Various sensors are shown in the figure  two  As shown in.  
 

(picture from network)  In recent years, the development practice of industrial Internet / industrial Internet of things has proved that the combination of high-performance machinery and equipment, low-cost sensors, industrial network, Internet and industrial big data acquisition and analysis technology can effectively improve the efficiency and benefits of existing industries and produce new technologies, new models and new formats. One of the key factors is the low-cost sensor. In recent years, sensor technology has achieved leapfrog development. New sensors emerge in endlessly, with continuous cost reduction and performance improvement, which has become the R & D norm of the sensor industry. It is precisely because the cost of sensors is continuously reduced to the cost line that can be deployed on a large scale, the large-scale application of sensors can become the choice of enterprises.  
 
 GE  The company has a lot of experience in the application of sensors. As early as  two thousand and five  In,  GE  Its aircraft engine company was reorganized into  GE  Aviation began the transformation of business model. The original business of the company was only the production of aeroengines. Later, by installing many sensors on the aircraft, the company collected various parameters of the aircraft in real time, provided the airline with a complete set of solutions for operation and maintenance management, capacity assurance, operation optimization and financial planning through big data analysis technology, and provided various services such as safety controls and navigation prediction, It has produced good economic benefits for the enterprise.  
 
 Take air Italia as an example,  GE  Hundreds of sensors are installed on each aircraft, which can collect many data such as engine operation, temperature and fuel consumption in real time  GE  After massive analysis of the software, it accurately gives the ideal control method. This is the only one, air Italia  one hundred and forty-five  A plane saves money in a year  one thousand and five hundred  Million dollars in fuel costs. Through these data, we can also predict the possibility of engine failure in advance, make forward-looking preventive maintenance, and avoid flight delay, increased cost and even greater safety accidents caused by machine failure.  
 
 two thousand and thirteen  year  one  Month,  GE  Installed in a battery manufacturer in New York  one  More than 10000 sensors are used to monitor data such as temperature, energy consumption and air pressure during production, which can be used by factory managers  iPad  Obtain these data, so as to supervise the production and effectively optimize the allocation of production resources.  
 
 The popularization of sensors at low prices undoubtedly lowered a huge cost threshold for the new industrial revolution, built a rainbow bridge from physical space to digital space, and continuously turned spatio-temporal information into bit data stream.  
 
 

3. Software definition and manufacturing

Industrial software becomes the brain of machines and algorithms  /  The computing power increases sharply, and the software defines the material  /  spare parts  /  The spatio-temporal performance of the system.  
 
 Although software is studied as a relatively independent specialty in many universities and scientific research institutions, the author believes that it is easy to separate the software from the hardware basis of software operation and software  data  and  policy decision  The physical device under control. Because software has never been an artifact with only "computational" properties,  
 
 But also has a more important "control" attribute  。  As a highly integrated digital carrier of data, information and knowledge,  Must survive and run in the chip  。  Software and chip form a symbiotic relationship  。 The function and performance of the chip restrict the running speed and definable performance of the software  “  ability  ”  The programmed instructions of the software constantly drive the gate circuit and FET in the chip to run "on" and "off",  therefore  Software has the "quasi" function of driving chip since its birth  CPS  ”Characteristics  。  
 
 The relationship between software and chip is also subtle and constantly changing. In the past, software must adapt to the chip and develop software based on the constraints of the chip; Today, it is often to design chips for better software operation and maximize software performance, that is, to develop chips based on software requirements.  
 
 Firstly, the chip has the function of "shelter, calculation and storage" for the software, and the software has the function of chip  “  Assignment, empowerment and wisdom  ”  The role of. Only by comprehensively matching the respective advantages of software and chip can we give full play to the best benefits. Software and chip together constitute a fusion, which is a "quasi integration"  CPS  ”, it's hard to say who "defines" who. But there are two obvious points: software is the object of the powerful computing power of the chip. Without the support of the powerful computing power of the chip, it is difficult for software to give full play to the "empowerment" of physical entities  /  "Endowing wisdom" function; Human knowledge contained in software is the source of "intelligence" of man-made system. Without the logical guidance of various model and algorithm knowledge in software, the powerful computing power of chip has lost its place, and it is also unable to form the "empowerment" of physical entities  /  "Endow wisdom" function.  
 
 Secondly, the decision and data generated by the software can accurately control the physical equipment.  
 
 with  CPU  Take the FET in the gate circuit as an example, driven by software digital instructions, the high level (on  /  On current) is  one  , low level (off)  /  Breaking current) is  0  Therefore, the FET is constantly turned on at a high speed  /  "Off" state for calculation, and finally a series of gate circuits are output on a group of gate circuits, which are formed after calculation“  1/0  ”The arranged binary bit data can be used to drive the display device (such as various display screens) or the controller in the physical device to make the machine or product operate accurately, thus reflecting the software's control over the chip, peripherals and physical devices.  
 
 The early software was only used on computers, and the hardware scope only included some external devices such as display and printing. The results of software operation only need to be displayed on the hardware screen, which may be display data, curves and surfaces, images or sounds. However, these calculation results are not required to form a "closed loop", as long as they can assist people in decision-making. Today, the scope of hardware has been extended to all industrial equipment connected to computers. The result of software operation is to drive the physical equipment and form a "closed loop" with the physical equipment (see the discussion in the next section), that is, every subtle action of the physical equipment is sensed and fed back to the software through sensors, The software calculates according to the "right now" working scenario of the physical equipment, makes decisions according to the embedded mechanism model or reasoning rules, and gives the next optimized and most accurate action instructions of the physical equipment. Therefore, the relationship between software, chips and industrial equipment becomes real-time, close and colorful.  
 
 Software slave driver chip computing software    "Chip" drives computer peripherals and develops into "software"    "Chip" drives the precise operation of industrial physical equipment and its digital twin, which makes the software have a strong "definition" effect on physical equipment.  Software definition has now become a technical phenomenon in the manufacturing industry: software defines not only parts, materials, but also products, tooling, process, assembly, production line, production process, supply chain, product use scenario, product maintenance and upgrading, customers and sales, Defines the enterprise and everything that can be defined.  
 
 

4. True integration of two technologies (software closed loop)

Bits embrace atoms,  IT  hand in hand  OT  , cyber integrates the digital instructions given by physical and industrial software to accurately control physical equipment across time and space.  
 
 "Three body intelligence model" in the three body intelligence revolution  ”  The two big cycles in give us good enlightenment: we should re understand software from the basic role of knowledge generation and knowledge flow, and understand why today's software is "closed-loop".  Physical entity → conscious human body → digital virtual body → physical entity  This big cycle has clearly told us that human beings have accumulated a lot of knowledge to understand and transform the world in the process of interaction with the physical world in the past hundreds of thousands of years. After the evolution from the interaction of two bodies (physical entity and conscious human body) to the interaction of three bodies, the monotonous situation of "single interface of two bodies" has changed greatly into a complex situation of "three bodies and three interfaces".  
 
 Physical entity → conscious human body → digital virtual body → physical entity  The big cycle, in fact, is a particularly important knowledge cycle. It is not only the only way for software to form a closed loop and move from conventional software to industrial software, but also the origin of the connotation of software definition.  
 
 In combination with the views put forward by Dr. an Xiaopeng, executive member of the informatization hundred people's Committee, in the article "future industry from the perspective of software", The author gives the logic closed loop of software enabling and enabling based on the large cycle of knowledge flow of three-body intelligent model: physical world operation → operation regularization (mapped to human consciousness activities) → regular modeling → model algorithmization → algorithm coding (entering digital virtual body) → code software → software optimization of physical world operation (and human behavior), as shown in the figure  three  As shown in.
 
 
 
 

chartthree From three agent intelligence model to software definition model

In Figure  three  In the logical closed loop, sort out and summarize the knowledge accumulated by human beings in the long-term interaction with the physical world, find the operation law of the physical world (including materials, equipment, etc.), then digitize these knowledge, establish the rules of automatic flow of bit data based on industrial software, and then carry these knowledge with the automatically flowing bit data Deliver it to any place where it is needed, so as to guide machines or people with correct and ubiquitous knowledge. Under the uncertainty inside and outside the system, complex work scenarios and given resource constraints, through the continuous repetition of the intelligent process of "state perception, real-time analysis, independent decision-making, accurate implementation, learning and improvement",  Give the right data, the right version, to the right people and machines at the right time, so you can do things right once, get better once, and even do better next time  。  (author)  one thousand nine hundred and ninety-seven  Boeing Company  stay  DCAC/MRM  Project information, including“  four  individual  right  ”：  right data  ，  right version  ，  right time  ，  right person  ，  (compiled from its original intention)  
 
 The core content of intelligent manufacturing is to enter the digital virtual body with human knowledge, take the digital virtual body as the basic connotation of software definition, and emphasize the "closed-loop" characteristics of today's software different from traditional software - accurately control the shape of machinery and equipment and the microstructure of materials with the bit data flow defined by software. Today's software no longer only appears as an aid to human decision-making, but surpasses the human body in an autonomous way  /  The human brain directly drives the operation of physical equipment.  
 
 The software closed loop has formed a real integration of two technologies and created intelligent manufacturing  /  Enabling technology of industrial Internet  CPS  (cyberphysical system). Under the function of software definition,  CPS  It can accurately control the position of each material atom and the relative position (lattice) between atoms, accurately control the relative position of each formed part and each part, and accurately control the operation action, cycle, time and energy consumption of each machine and equipment. This has become the most important purpose of the new industrial revolution.  
 
 

5. Optimize the allocation of manufacturing resources on a large scale

Based on the industrial Internet, optimize the allocation of manufacturing resources in multiple domains rather than single domains and in a large range rather than in a small range.  
 
 stay  two thousand and seventeen  year  eleven  month  twenty-seven  On deepening the "Internet"    "Advanced manufacturing industry" points out in the guiding opinions on the development of industrial Internet:“  Industrial Internet is a new industrial revolution characterized by digitization, networking and intelligence  Critical infrastructure  And accelerate its development  It is conducive to accelerating the development of intelligent manufacturing  , more extensive, more efficient and more accurate  Optimize the allocation of production and service resources  , promote the transformation and upgrading of traditional industries, give birth to new technologies, new business forms and new models, and provide new momentum for the construction of a manufacturing power.  
 
 ”  in the past  twenty  In, enterprises implemented a lot of digitization  /  Many enterprise information systems are used in information transformation projects, but its scope of action is often "one plant, one workshop", and the optimal allocation of manufacturing resources is limited to local  /  Within the small scope of the enterprise. For manufacturing resources that need large-scale configuration optimization, there is often no good solution due to serious space-time constraints.  
 
 The purpose of industrial Internet is to achieve "greater scope, higher efficiency and more accuracy"  Optimize the allocation of production and service resources  ”The new industrial network is the key infrastructure for the implementation of intelligent manufacturing. Industrial Internet helps enterprises to break the barrier between  The invisible, untouchable and insurmountable "wall" between the physical world and the digital world has broken through the time and space constraints that seriously constrained enterprises in the past.  
 
 Industrial cloud platform is the initial stage and basic form of industrial Internet.  Gradually running the enterprise's R & D tools, core business systems and key equipment on the cloud can maximize the integration between various systems under the unified Cloud Architecture, make the data flow smoothly and automatically between various systems through the cloud, and promote their interconnection and interoperability. This can not only effectively improve the operation efficiency and implement fine management, It can also optimize the allocation of enterprise resources in a large range, resulting in economic benefits and comprehensively improve the overall management level of the enterprise. Now most industrial Internet platforms are installed on private cloud, public cloud or hybrid cloud.  
 
 When more and more industrial factors enter the industrial Internet, the vast majority of supply chains will become a part of the industrial Internet. according to  IBM  A report released shows that in the environment of "software defined supply chain", in the industrial Internet platform and Industry  APP  Under the support of, the minimum production scale to ensure operation and cost competitiveness has been reduced on average  90%  , enterprise operation and management costs are greatly reduced.  
 
 Many tangible results have been achieved by connecting industrial elements on a large scale through the industrial Internet: for example, the "industrial Internet platform for new energy industry" developed by Qinghai Guodian green energy company accesses data every second in the perception layer  About 3600,  twenty-eight  The power station is connected to the platform to realize "unattended and few people on duty" through centralized monitoring, so as to save the cost of operators for power generation enterprises  40%  ；  Youye company is an annual producer in Shandong  three hundred  10000 ton iron and steel enterprises have implemented  Thingswise  After the industrial Internet platform, it saved a year  four thousand  RMB 10000;  Dongfang Guoxin company will  one  000  In a blast furnace for ironmaking  three hundred and ten  The block is connected to the ironmaking big data intelligent interconnection platform to reduce smelting energy consumption  3%~10%，  Improve labor productivity  5%  Above, reduce safety accidents  60%  ； After the root Internet implemented the root cloud industrial Internet platform for an injection molding machine equipment manufacturing enterprise, the plant's equipment failure was "discovered"  -  Response cycle from  one  Reduced to over hours  fifteen  Within minutes, reduce response time  75%  , shorter arrival time  50%  And provide customers with transparent processing process to improve customer satisfaction  50%  above; Qujing Cigarette Factory builds a cigarette digital factory based on the Jupiter industrial Internet platform of Hualong Xunda company, and the effective operation rate is from  89%  Rise to  96%  Capital occupied from spare parts inventory  one point two  Hundred million  /  Year down to  zero point three eight  Hundred million  /  Year, every  one hundred  Ten thousand cases cost less than enterprises of the same scale  zero point six  Hundred million  /  year  ； Beiming Zhitong company implemented Zhitong industrial Internet for a dairy group to improve machine efficiency  15%  , quality traceability is reduced from hours to minutes, and the effectiveness of quality analysis is increased  50%  , workshop statistics and patrol inspection personnel decreased  50%  Production records and labor volume reduction  80%  , wait.  

The connection of large-scale industrial factors makes it possible for enterprises to optimize the allocation of manufacturing resources at the large system level, so as to obtain better enterprise management and control ability, and effectively promote the transformation and upgrading of enterprises.  
 

6. Summary

The new industrial revolution marked by intelligence has brought unprecedented opportunities for global industrial transformation and upgrading, provided a huge driving force, created countless new technologies, new models and new formats, and produced many characteristics of the new industrial era.  
 
 According to the author's years of observation and patient combing of the new industrial revolution in many countries, I found many characteristics of the new industrial era, 100 and 1000, which can be summarized into these five:  
 (1) "human intelligence" to "wit" -- human knowledge is constantly entering the software, the knowledge carrier has changed from carbon based knowledge to silicon-based knowledge, and the digital productivity has increased sharply;  
 (2) low price popularization of sensors - sensors add "facial features" to machines, greatly enhance the perception ability of machines, and accelerate the digitization of physical spatial information;  
 (3) software definition manufacturing - industrial software becomes the brain of machines and algorithms  /  The computing power increases sharply, and the software defines the material  /  spare parts  /  The spatiotemporal representation of the system;  
 (4) true integration of two technologies (software closed loop) - bit embraces atom,  IT  hand in hand  OT  , cyber integrates physics, and the digital instructions given by the software accurately control the physical equipment across time and space;  
 (5) optimize the allocation of manufacturing resources in a large scale - based on the industrial Internet, realize the optimal allocation of manufacturing resources in multiple domains rather than a single domain, and in a large scale rather than a small scale.  
 
 Without clarifying the basic characteristics of the industrial revolution, intelligent manufacturing cannot be implemented in enterprises. A clear understanding of the above five basic characteristics can enable enterprises to drive into the fast lane of intelligent manufacturing as soon as possible.

This article is reproduced from“Zhao Min”, support the protection of intellectual property rights. Please indicate the original source and author for reprint. If there is infringement, please contact us to delete