The collision of the "physical world" (represented by manufacturing equipment) and the "digital world" (represented by technology representatives such as artificial intelligence and sensors) has created a huge shift in manufacturing. The integration of the two worlds will inject new momentum into the next round of economic development. The new technology represented by artificial intelligence is exerting great influence on production process, production mode and supply chain system.
According to accenture, the use of artificial intelligence technology will increase the total growth of manufacturing (GVA) by nearly $4 trillion by 2035, an annual growth rate of 4.4%. As a new "factor of production", the impact of artificial intelligence on manufacturing industry has several aspects :(1) the machine will partially replace people's work and realize intelligent automation. In China, Japan and other countries, the labor shortage caused by aging and human resource costs can be compensated. (2) artificial intelligence by enhancing labor skills to promote the efficiency of production, in order to improve the efficiency of the people, through retraining employees can perform more advanced design, programming and maintenance task or a creative work. (3) the depth of the artificial intelligence and manufacturing integration will speed up the development process of new products, not only will change the original production processes, and artificial intelligence programs not only can automatically complete the task, but also can implement new business processes. For example, customize product configurations according to the customer's personalized requirements. This will be the ultimate goal of artificial intelligence in the manufacturing sector.
, director of the Stanford artificial intelligence laboratory, a former Google brain project director, baidu's chief scientist Wu En before leaving baidu (Andrew Ng), will own entrepreneurial project focus in the field of manufacturing, hope that through transformation for the manufacturing industry to provide technology, training, operation and process management and a series of solutions, to become the AI in the field of manufacturing service provider. I hope to understand how the AI scientists understand the changes that artificial intelligence will bring to the traditional manufacturing industry through my communication with him.
Artificial intelligence will change what aspects of manufacturing.
Mr Ng believes that artificial intelligence could now be used in four areas of manufacturing.
First, many segments of manufacturing rely on visual inspection. The ability of AI devices to visually inspect samples is rapidly improving, enabling us to establish automatic vision detection systems. Ai can compare products and photos and decide whether to pass inspection. Ng's team has applied machine vision to precision quality analysis in manufacturing, and improved image comprehension by combining a camera that is more sensitive than human eyes and using AI technology. Tools, the company developed a machine vision found under outside of the realm of the human visual resolution for circuit boards and other products of the micro defects, and use machine learning algorithms for very small amounts of training sample images.
New impetus for economic development.
How do large traditional enterprises transform to artificial intelligence?
Shi ze: this round of technology wave in the accumulation of first-mover advantage will continue to widen the gap. Companies that have not been able to change in time will have to grapple with the changing business environment.
Secondly, optimize the production process. By adjusting and improving the parameters in the production process, AI sets parameters for many machines used in manufacturing. In the process of production, the machine needs to set many parameters. For example, in injection molding, you may need to control the temperature of the plastic, the cooling schedule, the speed, and so on. All these parameters may be affected by various external factors, such as external temperature. By collecting all this data, AI can improve the parameters of the auto Settings and adjustments to the machine.
Third, improve the design and manufacturing efficiency of the new product manufacturing process. Making new products is an iterative process in both design and production, full of fine-tuning. Artificial intelligence will be able to significantly shorten this process and improve the efficiency of the manufacturing industry.
Fourth, determine the source of product quality problems. The manufacturing process of many products involves a series of steps, so if the final product fails to pass the test, it is sometimes difficult to determine the source of the problem. Artificial intelligence, data science and data analysis will help automate the identification of problematic steps in production.
At present, Landing. Ai provides the cooperation partners of the manufacturing industry with solutions such as visual inspection, automatic control, intelligent calibration and problem root analysis. In addition to the applications that Mr. Ng mentioned. Future factories may use intelligent forklifts and conveyors to carry materials and finished products. In the field of machine vision, in addition to the application of quality detection at the micro level, the robot can also be trained to perceive the surrounding environment and avoid interruption or danger in the future. In addition, artificial intelligence can also be used in the field of manufacturing, automatic quality control, preventive maintenance and unmanned driving.
The challenges faced by traditional enterprises in the transformation of artificial intelligence.
Wu En think, for large enterprises, in the application of artificial intelligence, the biggest challenge is how to use the AI technology to expand business scale, can help enterprises to complete the challenge of talent and AI tools and did not exist. For traditional manufacturing enterprises, there are several main obstacles to the application of AI, one of which is the complexity of artificial intelligence technology. Few teams have been able to understand and implement this technology effectively. At the same time, few companies outside of AI technology have access to enough AI technology talent.
The AI integration strategy itself is as complex as the technology itself, which involves data collection, organizational structure design, the priorities of AI projects, and so on. Also, good AI strategy experts are rarer than AI technologists. Is the process of stretch launched a global enterprise transformation, especially involves the labor structure adjustment, the process itself is more complex, including cultural challenges, this aspect especially the problem of labor transformation, it takes time to transition. Manufacturing companies need to provide employees with better retraining and help them prepare for future work. In addition to the traditional enterprise talents, technical reserves need to be transformation, at the same time, people need to understand the purpose of artificial intelligence is not to replace labor, but in order to enhance the capacity of employees and can assign for the enterprise, to help it succeed.
Artificial intelligence can effectively meet the challenges of today's manufacturing industry through adaptive manufacturing, automatic quality control, preventive maintenance and other solutions. The AI in manufacturing applications, some technologies are already starting to part of the application, but has not been widely used, technology of complex and resources (including human resources) the shortage of the current transformation of one of the obstacles.
Positioning of traditional manufacturing companies and AI technology companies.
Because traditional manufacturing industry lacks the corresponding talent in artificial intelligence. Artificial intelligence technology companies should play a role in training the labor force, and are the first to help traditional employees teach the necessary skills in the new wave of transformation in traditional industries. At present, the traditional company might think that AI is a hard to imagine the future, but the AI can help enterprises to realize the automation of part of the task, enable employees to take on a higher level of job responsibilities, and thought it is used to create more valuable contributions.
In addition, Mr. Ng noted that he recently attended the industrial Internet summit in Beijing in 2018 and found that many Chinese companies are already on the path to artificial intelligence. The Chinese government has been promoting the rapid development of the industrial Internet and has promoted many initiatives to transform traditional industries. By forming industrial alliance of the Internet, supporting industrial upgrading the new policy, a large number of companies are starting to fully meet the robots and machine learning, Internet of things, big data fusion of iot industry development trend.
(
)
In the global map of traditional manufacturing, the impact of transformation on developed and developing countries varies. In developing economies, the transformation of manufacturing to AI will make it easier for products to scale production and lower manufacturing costs. AI will also help small producers sell their products to global supply chains and benefit from them. In advanced economies, the deep integration of artificial intelligence and manufacturing will also pave the way for a new generation of products, equipment and experience.
Some of the challenges facing the manufacturing industry are universal, not specific to a particular company or industry. The principles and concepts behind the application of AI in the manufacturing industry are repeatable. But Mr. Ng hopes Landing. Ai will be able to work with manufacturing companies to develop technology and training for specific industries, rather than generic technology tools. Wu En da team developed including the introduction of new technology, operational process management, reshape talent strategy, organization structure, AI AI transformation plan, has cooperation with companies such as foxconn, hon hai.
In this, please try to imagine the future manufacturing forward scene: by artificial intelligence, big data, cloud computing, 5 g represented by a series of technology such as communication, the Internet of things will make the future manufacturing clusters, production from the value chain model, business model, product design aspects of changing nature. As technology advances, the product itself will carry more information or more intelligence. With the rise of consumer power, it is possible for consumers to actively participate in design and co-create products. And the increasing demand for personalization and customization will undermine the so-called mass consumer market. Due to the popularity of 3D printing and other technologies, "scale" is not necessarily "economical". The manufacturing value chain will also be redistributed. Producers bypass middlemen and attract consumers directly. The production mode of the product will also change from "production by forecast" to "production by order". From ideas to market speeds, consumers are also more direct to feed demand back to producers. With the evolution of technology, traditional manufacturing needs to embrace the future and conduct a self-revolution.![]