What is the original driving force of "intelligent manufacturing"?

2021-02-05 14:57:30 浙江铭君自动化设备有限公司 Viewd 2258

The origin of smart manufacturing is smart factory, and the concept of smart factory was first proposed by IBM in 2009, which belongs to the application practice of IBM's "Smart Earth" concept in manufacturing. As intelligence has become the development direction of the manufacturing industry in various countries, great changes have taken place in corporate management methods, personal life concepts and international competition patterns. The construction of intelligent manufacturing has received unprecedented attention and is expected to become a new model of manufacturing in the future. So, what exactly is it? What about smart factories or smart manufacturing?


The definition of smart manufacturing is slightly different in each country, but the degree of importance is self-evident. According to Deloitte’s ranking of the importance of advanced manufacturing technologies in the future, the importance of "smart factories" in the three major regions of the United States, Europe, and China The rankings are 4th, 1st and 2nd respectively, highlighting its important position in the future intelligent manufacturing technology:


China: Intelligent manufacturing is defined as an advanced manufacturing process, system and system that is based on a new generation of information technology and runs through all aspects of manufacturing activities such as design, production, management, and service. It has the functions of in-depth self-perception of information, intelligent optimization, self-decision, and precise control of self-execution. The general term for the model".


United States: "Intelligent Manufacturing Innovation Institute" defines smart manufacturing as: smart manufacturing is a combination of advanced sensing, instrumentation, monitoring, control, and process optimization technologies and practices that integrate information and communication technologies with the manufacturing environment , To realize real-time management of energy, productivity, and cost in factories and enterprises.


Germany: The connotation of "Industry 4.0" is digitization, intelligence, humanization, and green. The mass production of products can no longer meet the needs of customers' individual customization. If you want to make single-piece and small-batch production to achieve the same as mass production The efficiency and cost required to build a smart factory that can produce high-precision, high-quality, and personalized smart products.



On the whole, smart manufacturing does not have a strict definition. It is generally understood that on the basis of factory control, information technology, Internet of Things technology, equipment monitoring and other technologies can be used to effectively control factory planning, resources, execution and other businesses, and to understand production status in time , To reduce the phenomenon of idle work caused by poor communication of information, and reduce the risk of quality data errors caused by manual input in paper documents. In addition, green and intelligent means and emerging technologies such as intelligent systems are integrated to build a modern factory with high efficiency, energy saving, green environmental protection, controllable cost and traceable quality.


The promulgation of the "13th Five-Year Plan for Intelligent Manufacturing in China" established that by 2020, traditional manufacturing will complete the digital transformation, and by 2025, key enterprises will achieve intelligent transformation. Driven by policy guidance, China's smart manufacturing industry is playing an increasingly important role in the manufacturing industry. Since the importance of intelligent manufacturing is self-evident and imminent, where is the original driving force for intelligent manufacturing? effectiveness? Environmental protection? Or quality?


At present, most of the domestic manufacturing industry belongs to the low-end manufacturing industry, and most of them are sweatshops that eat demographic dividends. However, due to the relatively low labor costs, factory owners do not need to build so-called smart factories. It is better to devote their energy to Sales end pulls more orders.


In addition, it is also because the investment in a production line or digital system is as little as millions, and even if the factory does not have an automated production line or information system, the cost of all manual operations is lower than the cost of building a production line. Accumulated labor cost for several years <production line construction cost. Because of this, we have seen too many scenarios in reality: a factory needs one million to establish a management and control system, and even if a few interns are hired to collect data manually and summarize by Excel, the costs of a few workers will add up within a few years. There is not one million, so the choice is clearly on paper for the boss. Therefore, the biggest obstacle to smart manufacturing is not foreign competitors, nor technical difficulties, but the domestic demographic dividend.


However, "Nine Nights of Dragons chanting changes, and the situation will swim in shallow water." Although the labor cost in the United States is 2.57 times that of the domestic labor cost, the United States has a high degree of automation and few labor. Two domestic production lines with a total monthly output of 4,500 tons employ 250 people. With improved equipment in the United States, only 180 people are employed on two production lines with the same capacity. According to the current upward trend of domestic workers' wages, if one considers that domestic wages will be doubled in five years and wages will be quadrupled in ten years, then China will not have any advantage in labor costs. Even if the labor cost advantage relative to India, Vietnam, Thailand and other countries is gradually weakening, since 2008, the growth rate of labor costs in my country has been significantly faster than the growth rate of industrial production efficiency. According to estimates, in 2019, labor costs in my country's manufacturing industry It will be 177% of Vietnam and 218% of India. The picture below shows the rate of increase in regional wages (estimated) and nominal GDP growth. It can be clearly seen that labor costs are growing rapidly, and the difficulty of corporate human cost management and control is increasing day by day. In particular, the difficulty in recruiting labor for labor-intensive companies has never been new, such as 2020 As soon as the epidemic began to abate in 2009, the obligation was raised: Come to work on a voluntary basis and reimburse the car fare.


While labor costs have soared, tertiary industries such as food delivery and Didi are now swarming, and even the profit growth space is far greater than that of the low-end manufacturing industry, and it is subject to many technical and non-technical difficulties, and many business scenarios are difficult. Realize unmanned automation (such as express delivery, electrical door-to-door installation, unmanned driving, etc.), so it can create a large number of labor-intensive low-end jobs with higher salary levels. I have personally seen at least a dozen friends around me bid farewell to their previous jobs at BYD factories, and then deliver food and Didi on JD. On the contrary, their salary is much higher than that of BYD, and they live more decent lives.


Therefore, the development of the tertiary industry represented by the takeaway industry is equivalent to a bloodletting therapy for the manufacturing industry, which robs labor on a large scale, and forces enterprises to push for intelligent manufacturing to realize the replacement of people by machines, thereby reducing Manual participation makes cost controllable and quality controllable. Assuming that there is no large-scale development of the tertiary industry represented by takeaway, Didi, and express delivery and the encroachment on the demographic dividend of the manufacturing industry, and the absorption of excess low-end labor, smart manufacturing lacks the most fundamental and primitive driving force to develop smart manufacturing.



When you think about it, you must seize the time and opportunity, because someone may already be acting. For example, since 2011, Terry Gou has launched a large-scale machine substitution plan to solve the problems of recruitment difficulties and rapid increase in labor costs, and plans to make Foxconn's automation rate reach 30% by 2020; and now Foxconn has deployed more than 40,000 Taiwan’s self-developed Foxbot industrial robots have replaced a large number of assembly line workers with robots. Despite this, they still have to consider relocating their factories to Vietnam in the face of the sharp increase in labor costs. Because it is estimated that the average salary of employed persons in urban units in China in 2017 is 2.14/3.51 times that of Thailand and Vietnam, this has also caused some labor-intensive manufacturing industries to shift to Southeast Asian countries with lower labor costs and more relaxed environmental protection requirements.


Similar to Foxconn, Midea has invested 600 million yuan in renovation costs and 6 years of time in 2012 to build a fully automatic assembly line for central air-conditioning core components. The production efficiency has increased by 70%, the number of people on the production line has dropped by 50%, and the product qualification rate has reached 99.9%. In the polishing workshop of Dongguan Jinsheng, 60 robots polished the middle frame structure of the mobile phone day and night. One robotic hand replaces 6 to 8 workers. In the workshop that originally required 650 workers, now only 60 people are responsible for line inspections and inspections, and the number will be reduced to 20 in the future. Compared with skilled workers, although the newly launched robots are "newcomers", the quantity and quality of products produced far exceed those of skilled workers and experts.


At the same time, labor costs and wages continue to rise, while the cost of robot recovery time is decreasing year by year: in 2012, the cost recovery time of an industrial robot was about 5.2 years, and by 2018 it dropped to about 1.05 years. At the same time, under the deterministic trend of rising labor costs and falling equipment prices, the payback period of industrial robots in the future is expected to be further shortened. The replacement of manufacturing machines and the upgrading of intelligent manufacturing are the general trend. The decrease in labor force and the rise in labor costs force the manufacturing industry to upgrade to intelligent automation. Intelligent manufacturing will gradually replace human labor and improve production efficiency. This is the future development trend of the manufacturing industry.


So on the whole, the current demand for smart manufacturing in factories is not actually derived from competition or technological development. Of course, these two also have influential factors, but they are not the fundamental reason for the digital transformation of enterprises, the most primitive drive. More power is the change that has to be made in order to get rid of the labor-intensive nature and to better control the cost. So what are the differences between the manned and unmanned factory operations? In other words, what intelligent manufacturing has brought to labor-intensive others. At present, the following scenarios are more mature:


Equipment inspection


At present, most companies in China still maintain the nature of "artificial intelligence". Whatever can be done manually must be mainly manual. The reason for this situation is that the labor cost was relatively cheap or the cost of information technology was relatively high before. Now that labor costs have soared and information technology has gradually matured, and still maintain the status quo of "artificial intelligence", it is a bit of a broomstick.


For example, in the large-scale process industry, it is nothing more than equipment and various instruments that affect the delivery time and quality. However, large-scale chemical plants, even companies with pricing power such as Hangzhou Hengyi Group, still maintain manual inspections. tradition of. Regularly inspect the equipment for spot inspection, regularly inspect the instrument to check process parameters such as temperature and pressure, and then record the data in paper documents, which is time-consuming, laborious and costly. Similar to this kind of business, it can be completed through automatic data collection and alarm, achieving semi-automation and digitization, thereby reducing costs and controlling the delivery cycle.


Quality inspection based on machine vision:


At present, most of the printing quality needs to be checked manually. Long-term quality observation has irreversible damage to the eyes, and the efficiency and effectiveness of the eye detection are open to question. Many factories have to rotate jobs every two to three months because of the naked eye. I really can't stand it. In this case, the quality inspection method based on machine vision can be used. The image of the printed matter to be inspected is collected through CCD, and the image is processed. The standard is set according to the quality requirements of the printed matter (the standard is proposed by the customer). If the relevant indicators of the detected image exceed the set standard, the system will automatically alarm, and the detected product will be deemed as unqualified, and the unqualified product disposal process will be initiated automatically. Through this way of reducing human participation, costs are reduced and quality is improved.


Application of warehousing logistics


Warehousing and logistics is also one of the labor-intensive businesses. At present, traditional factories are filled with a large number of forklift drivers, manual sorting, warehouse keeper and other personnel, ranging from warehousing sorting, location management, loading and unloading, outbound sorting to material transportation , Especially the warehousing business is labor-intensive and error-prone. Computer vision is used for the perception and map positioning of the sorting robot, and machine learning and deep learning are used to realize the path planning and obstacle avoidance of the sorting robot. Through mathematical planning and other operations research optimization algorithms and genetic algorithms, the warehouse management strategy is realized. Through the multi-agent algorithm, the ant colony algorithm is used for the coordinated actions of multiple sorting robots. Based on artificial intelligence technology to realize the overall coordination of shelves, commodities, and robots, it can more quickly realize product entry and exit and efficient warehouse shelf planning. In factory warehousing, various types of automatic assembly lines, automatic distribution, warehousing and distribution robots have begun to be slowly applied. Based on artificial intelligence technology, each material can have an optimal path and be delivered in the shortest time.


Production and operation control:


At present, there are a large number of manual control in factories, and even companies with relatively high degree of automation of assembly lines also have such problems. Even departmental companies set up "data divisions" due to leadership needs, collect business management and operation data through manual entry, and then use Excel to summarize statistics. Management should not be a labor-intensive job, but because of the existence of black box factories, The number of managers has increased sharply. It is entirely possible to use digital transformation and digital-driven business management to find business sites that are inefficient, low-efficiency, and low-quality, and then optimize and upgrade to improve customer experience and reduce delivery cycles.


At present, the demand for enterprises to replace low-end and repetitive labor through machines and software has gradually shown a trend of turbulence, but the ideal is very full, and the reality is very skinny. Due to various technical and cost constraints, many businesses cannot get rid of it at all. Human intervention can only greatly reduce the degree of human participation, thereby reducing the number of people and reducing the quality risk caused by humans. The business chain of the entire enterprise is very lengthy. At present, it can only upgrade and optimize single-point businesses such as quality inspection and warehousing. Then, according to the development of technology and the continuous integration of multiple technologies, it gradually spreads and develops from point to line topology.