Shulou(Shulou.com)11/24 Report--

This article comes from the official account of Wechat: knowledge Automation (ID:zhishipai), author: Lin Xueping

Industrial software is an industrial product, which naturally comes from industry. So why do Germany and Japan, as manufacturing powers, do not seem to have grown up in industrial software? In fact, Germany has also developed a variety of CAE software, such as professional fluid-solid coupling software, but has not developed a large general-purpose CAE software vendor like Ansys in the United States. In terms of CAD software, if you put aside the acquisition of American software company UGS by Siemens, then the German software list looks quite inconspicuous. The situation of industrial software in Japan is similar to that in Germany. In the current industrial software field, it is mainly dominated by the use of software from the United States and France, supplemented by many special tools for secondary development.

Japan and Germany dominate the market in the fields of industrial embedded software robots, automobile control modules and CNC machine tools, which is directly related to their strong mechanical engineering. The "mechatronics" proposed by Japan in the 1980s has played a great role in promoting the development of its embedded software. But why can American General Motors bring about the rise of CAD software, while Japanese Toyota does not catalyse an international general CAD tool software? In fact, Japan has successfully shaped some international brands, whether in iron and steel, shipbuilding, automobile, home appliances and other industries. Japanese industrial software is one of the few industries that have failed to move towards internationalization, which should be directly related to Japan's "hardware first" tradition.

Before 1985 and after 1980, the five major manufacturers that led the Japanese computer hardware market were IBM, Fujitsu, Hitachi, NEC and Univac, accounting for a combined market share of about 90 per cent. Since Fujitsu and Hitachi introduced IBM-compatible computers in 1974, IBM-compatible computers have occupied 59% of the Japanese computer market since 1980. IBM series products dominate the hardware market, which attracts a lot of computer software supported by IBM operating system to enter the Japanese market.

Of the IT spending of Japanese companies, only 7 per cent is spent on outsourcing customized software development and less than 1 per cent on commercial shelf software, most of which are self-developed software. External software vendors rely heavily on computer mainframe factories. During this period, the main business of most software companies was to provide services to computer mainframe factories, and revenue from computer mainframe factories accounted for about 60% of the revenue of such companies.

Before 1980, the price of Japanese software was determined almost entirely on the basis of development cost or workload. This means that most software companies do not sell technology and rely only on human resources to provide technical services.

In the process of the development of Japanese software industry, this is the beginning of a workshop. The software industry largely depends on individuals, and the improvement of the level of programmers takes precedence over the improvement of software products. Practitioners in the Japanese software industry are only seen as a source of programming labor, rather than skilled professional designers or systems engineers.

A failed state attempt with the substantial improvement of Japan's mechanized and industrial production, Japan's computer hardware industry is booming and its output is increasing rapidly. The software still stays in manual work, and the contradiction between supply and demand between them is becoming more and more obvious.

After nearly a decade of gestation, the Ministry of Trade, Trade and Industry (MITI) and its affiliated Intelligence processing Revitalization Association intend to solve this problem in one fell swoop. According to the association, there will be a shortfall of 600000 in the supply and demand of software engineers by the year 1990 and 1 million by the year 2000. Therefore, the Ministry of Trade and Industry of Japan (now the Ministry of economy, Trade and Industry of Japan) has set up a national software development project in an attempt to establish a new operating system platform for software developers to use. At that time, the Ministry of Trade and Industry of Japan, due to the great success of the very large scale integrated circuit (VLSI) project in the early 1980s, could not help but be a little hot-headed and complacent, ready to make an effort to establish a software development platform to improve the efficiency of software development.

SIGMA, a national project started by the Japanese government in 1985, is an action of "industrialization of software production", which aims to eliminate the "software crisis" and promote the industrialization of software production. It expects to establish a computer system that can provide Japanese software developers with the intelligence tools needed to produce software. The envisioned business model is that when the SIGMA system is in operation, the operating expenses can be paid by the rent paid by the user.

In July 1986, at the 15th Technical Forecast Seminar in Japan, the head of the system Development Department of the Japanese Information processing Revitalization Association made a report entitled "Japanese Software Development and SIGMA Project". At that time, more than 60 companies had already participated in the project.

However, the goal of industrialization of software production is still aimed at serving the hardware. And these Japanese companies misjudged the situation and were still tackling mainframe problems without noticing a major trend at the time that the software industry was booming independently. According to one Japanese person, the project is doomed to failure because it was developed on the premise that computers were not yet developed. Many people noticed the absurdity of this premise in the course of the project, but could not stop. Because once it is called "national project", it cannot be easily interrupted.

Five years later, the national project became unsustainable, so 50 computer manufacturers and software companies funded a commercial company with the same name as the project, which was the last job for the project. In 1991, Japan reached a general specification with the UNIX International Organization for Standardization, which marked that Japan gave up the route of resisting software standardization. SIGMA came to a dead end.

This failure is a major setback for Japan's attempt to seek software autonomy. The SIGMA project is intended to counter a global trend of standardized software programs, and this vanguard project that goes against the trend is quickly turned into a pile of abandoned stones. Japan lost 23 billion yen in investment, so much so that it was no longer able to organize decent actions to revitalize the software country.

Lack of strong general software, but the enterprise self-research ability is very strong, Japan's foundry industry is strong, but the foundry simulation software is not strong. Japanese foundry simulation software JSCAST is jointly developed by the research group of Osaka University and Komatsu Machinery of Japan, and it is still the first selling foundry simulation software in Japan. The core of casting simulation software is the solver of hydrodynamics and heat transfer. Compared with German MAGMA casting simulation software, American FLOW-3D software and Korean AnyCasting software, Japanese casting simulation software has a certain gap, and its main disadvantages are unstable solver, slow calculation speed, low ability to deal with complex castings and so on. There are very few companies using Japanese JSCAST casting simulation software in China. In the world, in addition to Japanese enterprises, enterprises in other countries rarely use this Japanese industrial software.

The weak foundry simulation software is a microcosm of the weak strength of Japanese general industrial software. On the contrary, the self-developed software of Japanese enterprises is often very developed.

Japan Electric Co., Ltd. (NIDEC) mainly produces small and medium-sized motors, bearings and other parts. Since 1994, Nippon Electric has begun mass production of liquid bearings (Fluid Dynamic Bearing, referred to as FDB) for hard disk drives. With the amazing progress of hard disk drive in data storage density, the traditional ball bearing manufacturing has been unable to make every ball size exactly the same in principle. The aperiodic vibration caused by the size difference between the balls will make it impossible for the magnetic head to accurately record / read data on the high-precision and high-density track. In order to explore the best structure of liquid hydrodynamic bearings, Nippon Electric has gradually embarked on the road of internal development, independent research and development of modeling technology and simulation software.

Japan's first mass-produced hydraulic dynamic bearings have developed at an astonishing speed, and the monthly output has achieved a leapfrog breakthrough, and the product range has been greatly expanded around 2000. By 2002, liquid hydrodynamic bearings have become the mainstream products of bearings for spindle motors. With the output rising from tens of thousands and hundreds of thousands to millions, it is obviously out of date to adopt the method of quality management in the production of a small number of trial products. Because in order to detect abnormal phenomena, it is often necessary to make more than 1000 trial pieces to represent different modes, which is time-consuming and labor-consuming, and the cost is also quite high.

In order to shift this work from the workshop to the design room, Nippon Electric decided to use CAE software for simulation, thus changing the trial-and-error testing process of "creativity, modification and redesign" for all verification cases. However, the general simulation software on the market is weak to cope with this special product, so Nippon Electric decided to develop unique analysis software within the company. The biggest advantage of this is that the range can be locked as the unique function of the hydrodynamic bearing, the three-dimensional model can be defined by the two-dimensional model, the degree of freedom of the vibration model can be reduced, and only the phenomena needed in the design can be taken as the object. modeling and calculation. As a result, the necessary simulation calculations of rigidity and attenuation can be carried out within the limited computing power.

By using the simulation calculation, the clearance parameters between the shaft and the bearing can be changed to 1000-2000 different modes at any time, and the calculation results can be obtained quickly in tens of minutes. This technology is now also used in hydrodynamic bearings for fan motors. With the development of electrification, automobiles, household appliances and other products have higher and higher requirements for high efficiency, mute, low vibration and so on. This highly targeted CAE technology is playing a more and more important role.

The CAE software department of Nippon Electric is also an integrated part of design, process and manufacturing system. It belongs to the same organization as the design department and carries out software development and upgrade together; when encountered with difficult projects, the CAE software department will also cooperate with the Central Motor basic Technology Research Institute of Nippon Electric to carry out joint development.

From part design to module and unit design, to the structural design of the shell side with modules or units installed, Nippon Electric can be completed through independent CAE technology. Nippon Electric has gradually evolved from a supplier of motors and bearings to a key strategic partner that OEM manufacturers rely on, in which the self-developed CAE software has played an important role.

Symbiotic development and highly embedded strategy Japan is the world's third largest economy and has a world-class manufacturing industry. Japanese software is second only to American software in terms of sales. Embedded software in the machine tool, robot and automotive industries is the only one in the world, and the quality and productivity of Japanese software are not below that of American software. However, Japanese software products and services are increasingly lack of global competitiveness and lose their sense of global presence. There is a huge gap between the strong software development ability and the weak software ability in the Japanese software industry, which is a great myth left by the Japanese software industry.

Many scholars have also noticed this phenomenon. Professors Cole and Nakata of the University of Berkeley analyzed this in detail in 2014. Two conclusions are impressive: one reason is that there are a large number of IT software outsourcing companies in the Japanese software industry, which create "blue-collar" jobs for employees with moderate software skills, while there is a lack of demand for good software architects / designers, which leads to a lack of software innovation. Another reason is Japan's worship of "creation".

Hardware engineers have higher income and social status than software professionals. The "path dependence" formed by Japan's excellent hardware manufacturing makes it impossible for the software industry to become an industry that attracts talents. 1/5 of software developers in the United States have received graduate education, compared with 1/10 in Japan. In terms of doctorates, there is an even greater gap between software developers in the two countries.

In the early 1990s, the information technology capabilities of large Japanese manufacturing and service companies weakened. Large companies spun off their IT divisions into subsidiaries and began to rely more on these subsidiaries and other systems integrators and their subcontractors. Over the next three decades, although electronics companies have evolved into IT companies, traditional electronic engineer thinking still dominates senior positions in the company. Japanese companies value hard rather than soft, and their organizational structure is often mechanical engineers and chemical engineers at the top decision-making position, while electronic engineers and software engineers at the bottom. The hardware-centric approach continues for a long time, while the role of software is still regarded as a functional assistant and controller.

Professor Nakata mentioned in his report that 20 years ago, some scholars warned the United States that small Silicon Valley companies dotted around the United States could not compete with large Japanese high-tech integrated manufacturers in terms of financial resources. The "factory production" (factory approach) software development method favored by large Japanese companies is superior to the "craftsman led" (craft approach) software development method dominated by the United States. In other words, following the success of the manufacturing industry, Japan is becoming an important pole of the world software industry.

Now, it seems that this prediction has not happened. In fact, the nesting relationship between large enterprises and small software constitutes a prominent feature of Japanese industrial software. In order to understand the close relationship between Japanese industrial software and the end-user industry, an industrial software company born out of Nissan is an excellent case study.

Nissan began to develop its own CAD and CAE software as early as 1987, and ten years later they set up an independent Nissan software company to apply the products to the development of Nissan and Renault models. Fujitsu took a full stake in the software company at the turn of the century. Subsequently, as a representative of mainframe computers and software systems, Fujitsu put various mechanical and automotive CAD software, as well as other manufacturing-related software technologies, such as lightweight 3D design, process table management, preview tools, into its company, while product data management used open source software. In the automotive industry that the company is most concerned about (sales account for more than 70%), its brand products, coupled with the Siemens software represented, are closely embedded in the process of the automotive design industry. it fully embodies the "integrated design of machinery, electricity and operation". After a comprehensive understanding of all the software of the enterprise, we can find that they are completely accompanied by a complete set of business processes for the development of a factory. This is very impressive.

Another Japanese CAE software company is Cybernet, a subsidiary of Japan's Fujisoft Software. Fuji Software employs more than 20, 000 people worldwide, and the CAE software company specializes in simulation. The American scientific computing software MATLAB was introduced into Toyota by Cybernet and quickly became the mainstream software. Later, MathWorks, an American company to which MATLAB software belongs, set up an office in Japan, and Cybernet acquired its competitor MapleSoft Software Company of Canada in order to enrich its product line. Cybernet, like a rare Japanese active volcano that likes to devour fresh air, has been acquiring global software resources. In fact, as the agent of American simulation software ANSYS, Cybernet nests agency products, self-developed products and M & A products to form a complete solution around the needs of enterprise users.

There is also an electrical design software company called JMAG in Japan, but it is difficult to say that this company is completely independent. Consortia can always be seen behind independent companies in Japan, and JMAG has the background of Sumitomo Bank.

Most Japanese like to work in big companies because they are more decent, while fewer people work in small companies. This has also led to a small number of start-ups in Japan, limiting the vitality needed by the software industry.

Japanese industrial software, after the failure of national action, chose the strategy of highly embedded globalization. Based on general software, loading one's own industry manufacturing expertise into it, and then closely binding with users, this strategy makes a strong Japanese manufacturing industry. However, for home-made CAD software, CAE software and EDA software, I am afraid Japan has already lost its goal and motivation.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.