Company Hopes Its Fuzzy Thinking Pays Off : Computers: Firm is first in United States to market products using fuzzy logic, a form of artificial intelligence, to solve problems.

You wouldn’t expect a company with a moniker like Togai InfraLogic to be in the business of making, say, pencils or pin cushions. And, as a matter of fact, it isn’t.

The fledgling Irvine technology company develops “very-large-scale integration processors for embedded real-time fuzzy logic applications.” In plain English, that means the company does complicated things with computers that ordinary folks can’t even begin to comprehend.

Togai InfraLogic is the first company in the United States--and until recently, the only one--to market computer products incorporating the so-called fuzzy logic technology. A branch of artificial intelligence, fuzzy logic is an esoteric mathematical concept for solving problems that deal with imprecise reasoning.

“Fuzziness deals with those shades of gray problems, as contrasted with the traditional yes or no ones,” said Carl Perkins, a co-founder of Togai InfraLogic. “It deals with ambiguity and vagueness.”

The fuzzy logic concept was developed in the mid-1960s by Lofti Zahdeh, a Russian-born computer science professor who teaches at the University of California at Berkeley. Zahdeh, who has become something of a cult figure among fuzzy logic followers, is an adviser to Togai InfraLogic.

Zahdeh invented fuzzy logic as a way for computers to deal with imprecise information such as “maybe,” “probably” or “most”--the kind of reasoning that humans do every day.

“A large part of artificial intelligence research is concerned with how to reason with information you aren’t certain of, something human beings do all the time,” said Nils J. Nillsen, chairman of the Stanford University computer science department. “There are many approaches, and Zahdeh’s is just one of them.”

And it’s the one with which tiny Togai InfraLogic hopes to prosper. Founded in 1987, Togai InfraLogic is a privately held firm funded by corporate and private investors in Japan and the United States.

The company is named after co-founder Masaki Togai, a Japanese-born engineer who in 1985 co-invented the world’s first fuzzy logic computer chip at AT&T; Bell Laboratories. Togai tells the story about how he and other Bell engineers tested their invention by trying to program a robot to play Ping-Pong using fuzzy logic theory.

“We spent a lot of time on that, but couldn’t do it,” he said. “It’s very hard to do.”

Togai later teamed up with the firm’s other founders: Carl Perkins and Rodney J. Corder, ex-engineers at Rockwell International’s semiconductor division in Newport Beach, who left to start a company specializing in fuzzy logic after failing to interest Rockwell in pursuing the technology.

Togai InfraLogic has developed a fuzzy logic chip and software tools for developing fuzzy systems. So far, its products have been used primarily by Japanese customers in applications involving cameras, air conditioners and automobiles. Most of the applications are in the research rather than production stage, however.

Although the concept originated in the United States, the Japanese have taken a big lead in fuzzy logic research and in bringing the technology into commercial products. Perhaps due to its oddball name, fuzzy logic is still considered a bit kooky in some scientific circles in this country.

“The United States is now where Japan was about 10 years ago,” Togai said.

Fuzzy logic is hot in Japan.

Entire Japanese television programs have been devoted to fuzzy logic, and Hitachi Ltd. has built a fuzzy system that runs subway trains in Sendai, Japan, 200 miles north of Tokyo. Yamaichi Securities, a Japanese investment firm, recently introduced a “fuzzy fund” in which stocks are traded using a fuzzy logic-based computerized system. One company has even developed a fuzzy logic technique for improving the swings of golf-crazy Japanese.

Many large Japanese electronics companies, including Hitachi, NEC Corp. and Toshiba Corp., are conducting research in fuzzy logic. It is also being studied in the Soviet Union, Europe and, especially, in China, which is said to have about 10,000 fuzzy logic researchers. But only a handful of companies in the United States are involved in fuzzy logic research, including, now, Rockwell.

“We’ve applied fuzzy logic for controlling a flexible aircraft wing,” said Allen Firstenberg, director of information sciences at Rockwell’s research center in Thousand Oaks. “There are lots of applications where you’d like to have improved control of the wings so you can fly faster. Fuzzy (logic) is an excellent controller in laboratory situations.”

Still, fuzzy logic has its skeptics in the United States.

“I feel it’s overrated,” said Judea Pearl, a computer science professor at UCLA. “In applications such as medical diagnosis and prediction, it has failed miserably. . . . If a sprinkler is on, fuzzy logic will suggest it has rained, simply because the ground is wet. It has no context.”

Bart Kosko, an engineering professor at USC, says fuzzy logic’s oddball name has given it a reputation akin to “a boy named Sue.” But, he sees great promise for fuzziness, calling it “one of the fundamentally new mathematical concepts of the 2Oth Century.”

Togai’s first major commercial project is a joint development program with Japan’s Mitshubish Heavy Industries to develop the world’s first air-conditioning system using fuzzy logic. The system is designed to function more like a human’s sense of temperature, making adjustments based on more variables than just the temperature being more or less than a certain degree.

Togai is working with a major Japanese electronics company, whose name it is won’t disclose, to develop an automatic-focusing camera using fuzzy logic, said Perkins, the firm’s vice president of business development. Another customer, which the Irvine firm also won’t disclose, is a large U.S. aerospace company that is using fuzzy logic equipment to diagnosis mechanical problems in jet engines.

Masaki Togai hopes that more U.S. companies will begin to see the benefits of fuzzy logic. He believes that Japanese firms are more willing to invest in fuzzy logic research because they generally take a longer-term view of business than their American rivals.

“Japanese industry is very competitive and more likely to look at new technologies” even if the bottom-line return is years into the future, Togai said.

But USC’s Kosko offers a cultural explanation for the resistance to fuzzy logic in the United States. Eastern culture is more rooted in religion and philosophy that stresses vagueness and uncertainty--the essence of fuzzy logic theory.

In the Buddhist and Taoist religions, “there is a simultaneous occurrence of something and its exact opposite,” Kosko said. “This type of thought permeates Eastern belief. What you are seeing in fuzzy theory is the clash of Eastern and Western belief systems at the level of science.”