How decisionmaking from bottom-up put a Japanese firm on top

How useful is the Japanese style of management to an innovative company like Fujitsu, Japan's largest manufacturer of computers?

Takuma Yamamoto, president of Fujitsu, pauses and reflects a moment.

''It's not easy to answer,'' he says slowly. ''We who are now in top management grew up in this system. People often say that in the American system, decisionmaking tends to be from the top down, whereas in Japan it comes from the bottom up. But we have top-down-type decisions also.

''I think what's more important is the way in which the bottom-up decisions are made. The initiative often comes from junior executives in their 30s or 40s. On the one hand, they take pride in knowing that they are the ones who are the real movers, the motive force of the company. On the other hand, they have to convince top management that what they are proposing is sound. Therefore they have to think in broader terms than their own particular division, whatever it may be. They may be engineers, but they also need sensitive antennas. All this is excellent training on the way up the executive ladder.''

Mr. Yamamoto said Fujitsu's original decision to go into computers was a bottom-up decision. The prime mover was Taiyu Kobayashi, now chairman of the board, then just a division chief.

In 1953 the Tokyo Stock Exchange asked Fujitsu to design a computer for it. Mr. Kobayashi was then chief of the development division. He formed a project team, and Yamamoto was one of three engineers chosen to be on it. Only the United States then had been working on computers.

The three engineers spent a month at Yamamoto's house batting ideas back and forth. Finally a design was born and began to take shape. This particular effort was not a commercial success, but from it grew the know-how that eventually led, in 1967, to the installation of the first Japanese-made on-line system for the Daiichi-Kangin Bank. Yamamoto was the leader of the Fujitsu team.

Fujitsu, founded in 1935, is also one of the world's leading makers of telecommunications systems and equipment. It has installed systems in Africa, Europe, and Southeast Asia. But 60 percent of its 800 billion yen ($3.25 billion) in sales the past fiscal year were computers and data processing equipment.

Fujitsu has only 2 percent of the world computer market, compared with IBM's 60 percent. But in Japan, the company has been outselling IBM since 1980. It also claims the world's fastest, largest-scale general-purpose computer, the Facom M-382. This has a 128 megabyte main storage capacity, compared with the 32 megabyte storage capacity of IBM's 3081K, the computer whose secrets Mitsubishi and Hitachi are charged with trying to steal.

Fujitsu has subsidiaries and joint ventures in many parts of the world, and has a 24.4 percent interest in Amdahl Corporation in the US. It exports large-scale computer mainframes to Amdahl and has recently signed an agreement with the Burroughs Corporation to supply Fujitsu facsimile machines.

As a company exploring the frontiers of knowledge, Fujitsu has regularly devoted a substantial portion of its earnings to research and development. Last year, a survey by the economic newspaper Nihon Keizai Shimbun showed that Fujitsu ranked fifth among Japanese companies in research and development, allotting almost 10 percent of its sales to R&D. (Fujitsu's domestic rival, NEC, formerly Nippon Electric, spent 10.7 percent of its sales on R&D, while the US giant, IBM, spent 5.5 percent.)

One of the more exciting devices Fujitsu has been developing is the high-electron mobility transistor, which has a high-speed switching delay time of 16.7 picoseconds (a picosecond is 1 trillionth of a second). Whereas today it takes banks of computers for the National Aeronautics and Space Administration to send the space shuttle aloft, it may eventually be possible with one computer the size of a bookcase.

Mr. Yamamoto thinks it is important for morale at Fujitsu to have the reputation of being an enterprise that people look to for unique creative achievements - a sort of ''what-will-Fujitsu-do-next?'' expectation. To do this, he says, Fujitsu must be able to tolerate the quirks of creative individuals that other Japanese companies might reject because of the stress on teamwork.

''We used to have a Dr. Ikeda,'' he said, ''who in his young days often didn't even bother to come to the office, and who had other idiosyncrasies. But we were able to go along with him, because he was an authentic genius. I think the Japanese emphasis on teamwork and consensus is a good thing. But we must also be capable of recognizing individual genius and create the kind of environment in which such individuals can function creatively.''

He sees future software programs going in two directions. First, some sophisticated programs will have global appeal, such as video games shows. Second, there will be greater local differentiation so that, while companies like Fujitsu or IBM can make hardware usable anywhere in the world, software makers will tailor programs to specific markets.

He thinks the computer of the future will acquire more and more of the qualities today characteristic of human intelligence. This is the concept underlying a project of the Japanese Ministry of International Trade and Industry to begin developing the ''fifth generation'' computer. A flock of Japanese companies, including Fujitsu, have joined forces to work on the project.