For artificial intelligence who wins

Source：Updated：2017-08-16 09:23:37

For artificial intelligence, who wins? "The age of robots is coming!" This is a theme that is getting hotter all over the world. 
 
Every day, we hear new examples of the use of artificial intelligence to assist or even replace human work. Much of the discussion revolves around the situation facing consumers. From products to the capital, however, from the factory to the global supply network, manufacturing because of its complex interactions between process and machine, in the tide of artificial intelligence might have been a greater impact. 
 
With the arrival of the sensor data deluge, the factory has the opportunity to join the information intelligentsia and effectively promote the output of the workshop. AI applications, from intelligent robot co-bots to virtual assistants, will disrupt the way factories operate. 
 
From production, distribution to sales, big data and artificial intelligence technology can help the enterprise analyze the full link in the process of production data, realize the production efficiency, inventory turnover, efficient use of equipment, such as the target. 
 
In terms of market, artificial intelligence, can be calculated to analyze vast amounts of transaction data and the flow of data gathered a large number of sensors, make the enterprise to realize automatic and intelligent production plan and market plan. In the manufacturing process, the product data can be collected, analyzed, and the data of production equipment can be used to realize automatic diagnosis and remote monitoring of the products. 
 
Artificial intelligence in manufacturing applications 
 
1. Predictive maintenance and self-monitoring 
 
Accenture's research shows that the potential benefits of artificial intelligence is far less than the automation of the impact of the past, for example, between 1993 to 2007, traditional automation system is estimated to have been produced additional annual growth rate of 0.9%, while in the developed economies grew by 0.9%. 
 
In the use of artificial intelligence for predictive and self-monitoring, Bosch, a German manufacturing company, is putting artificial intelligence at the forefront of their business, according to accenture's report. The company's "factory" of thinking in the car factory in Germany, the goal is to make artificial intelligent machines can self fault diagnosis technology, automatic order replacement parts, and predict maintenance requirements. 
 
By using machine learning to achieve predictive maintenance and self-monitoring, Bosch can save $1 billion in cost savings and is expected to generate $1 billion in additional revenue. 
 
And Bosch is not the only company that use artificial intelligence strategy, day wave co., LTD (hereinafter referred to as the wave) and the Japan electric co., LTD (hereinafter referred to as the NEC) as early as 2016 signed a cooperation, the use of artificial intelligence and the Internet of things technology, around for oil refining, chemical, liquefied natural gas, gas, power generation, resource development, and other fields to provide plant operating anomaly prediction system construction and service. 
 
Day swing with NEC together provide plant operating data analysis service, through the analysis of the operation data of covering the whole plant, can improve the completeness of data, shorten the abnormal condition prediction analysis time, improve the analysis speed. In addition, based on the overall data of the factory, the cause-and-effect relationship of abnormal situation can be determined in real time. 
 
NTT Communications Corporation, a unit of NTT group, and mitsui chemical co., LTD., have teamed up to apply artificial intelligence techniques to product quality forecasts. Them with raw materials from injected into the reactor and the reactor conditions of data set and on behalf of the relationship between the product quality of X gas concentration based modeling, with deep learning algorithm to analyze the data, a gas with highly accurate predictor of the quality of the products. 
 
American ge saw the same opportunities, the company set up artificial intelligence, is committed to using the data analysis, artificial intelligence, robot technology, such as oil and gas, transportation and energy industry with advanced testing services. For now, GE has begun testing its own drones and robotic crawlers for refineries, factories, railways and other industrial facilities. 
 
2. Improve quality inspection efficiency and yield through visual inspection techniques 
 
In the process of manufacture, the products on production line, the last is to check whether the good parts, and the work must be very familiar with products, experienced staff, thus with low efficiency, and manpower is also insufficient. 
 
However, through the image recognition of artificial intelligence, the visual inspection, the operator can replace can not only reduce the workload of workers, increase efficiency, but also can make the product quality is more balanced. 
 
NEC has launched "AI Visual Inspection" Visual Inspection, using machine learning techniques, detection of artificial resin, metal, plastic and other products processing production line of product image, quickly find out the bad products, improve production efficiency. 
 
NEC in NEC Industrial IoT, new AI Visual Inspection function, assist the operators in product testing. The specific process is that NEC will take a series of product images taken during the on-site inspection, and store them in the cloud of the Internet of things, NEC the WISE Iot Platform. When reach a certain amount of image data, the cloud of AI will be drawn from these data and analysis of the characteristics of a good product and bad product, and according to the analysis results, further induce of discrimination model, that is a good product and bad product, and the transmission of discriminant model from the cloud to the factory site on the equipment installed. 
 
At the end of the factory, the AI is built in the field equipment, which is based on the discriminant mode received and the discriminant product is defective. When production line to add new products, the cloud of AI also automatically learn the characteristics of the new product information, and update the discriminant model, which effectively reduce the additional design, research and development work. 
 
And abb cooperation with IBM, can record real-time image in the process of production, with the aid of abb system, and connecting with the IBM Watson AI ability and the real-time image real-time found defective products, identification of the human eye can observe question, and then analyzed using artificial intelligence system, quickly found and troubleshooting. 
 
3. Automation of robots 
 
Artificial intelligence in the factory, the most frequently mentioned is the robot. And some companies are working to make robots easier to use and smarter. 
 
Universal Logic based on the robot of artificial intelligence technology of the Neocortex can make the robot sense their surroundings, such as machine maintenance, rubbish, can rapid real-time interaction and to respond to the environment, it includes identification of various shapes of material objects, such as bottles, bags, boxes, and appropriate response, reliability of 99%. 
 
In addition to the third-party robotics platform, ge is developing its own technology in-house, such as the products that go inside the engine of an aircraft engine, as a vertical manufacturer. In addition to building software and hardware platforms, ge has also invested in the well-known Clearpath Robotics and the famous OC Robotics of the robot arm robot. 
 
As the robot's sense of touch becomes more and more sensitive, it will become more and more convenient to assemble and operate independently in the manufacturing workshop. And as the ascension of sensor and visual technology, the robot will be more intelligent, light and friendly, using robot take most of the assembly, moving, packaging, transport, or other physical labor, thus can greatly improve efficiency, reduce the cost. In the future, robots will not only be able to judge based on their procedures, but also make decisions based on experience. 
 
Artificial intelligence + manufacturing, who wins? 
 
Whether it is Germany's "industrial 4.0" or China's "made in China 2025" or the us government's efforts to bring back manufacturing, manufacturing remains the core battleground for global competition. However, it has previously been different that artificial intelligence will add to the competition for manufacturing, and the state's investment and emphasis on artificial intelligence has become a key factor. 
 
According to McKinsey, most outside investment will focus on the machine learning aspects of AI technology, such as voice recognition and robotics. Many investment company went to the United States (66%), followed by a Chinese company (17%), according to the Infosys survey found that over the past year, U.S. institutions are likely to increase the investment in the AI technology, although China AI investment amount is far lower than the United States, but its rapid growth in investment and AI deployment. "Our survey shows that the us and China are leading not only in AI investment and innovation but also in AI adoption," said the McKinsey study. 
 
Both China and the United States have built an ecosystem of artificial intelligence technologies and put AI technology at the center of their manufacturing strategy. Last year, the Obama administration released a strategic plan for AI development. China has also made AI an important part of its five-year plan to achieve significant development by 2020. 
 
"Made in China 2025" is the latest action plan of China, the purpose is to China's manufacturing from low labor costs to the manufacture of high value-added, and AI will in this play an important role in the success of the action plan. In order to complete the "made in China 2025" plan, China may need to make more acquisitions in accordance with its timetable. One example was the acquisition of German Robotics company kuka Robotics by Chinese manufacturer of household appliances, midea, in 2016. 
 
Infosys survey report, is in a leading position in China in the artificial intelligence maturity score, probably because of its "have less legal system and business procedure, facilitates the adoption and integration of artificial intelligence." But that does not mean that China will soon overtake the United States, which, after all, has many other advantages besides nearly four times as many external ai investments. For example, according to McKinsey, most of the big companies that invest in r&d are in the us. Superindustrial manufacturers such as ABB, Bosch, general electric, IBM, Siemens and tesla are all in the us or Europe, rather than in China.