Omron Corp, the Tokyo electronic components manufacturer which in January launched a UK operation in North London (CI No 1,839), has decided that fuzzy logic could be important here after all. After reeling off its many product lines at that press briefing, and describing the promotion of fuzzy logic in the UK as not a priority, the company recently hosted a briefing aimed at the more technical press, to expound the virtues of the technology in the computing environment. We at Computergram explored the concept of the fuzzy vacuum cleaner and fuzzy dishwasher a year ago (CI No 1,605); now the time has come to wander through the possibilities of using the technology in a commercial environment. Old concept
To re-cap, fuzzy logic is quite an old concept, historically accredited to Professor Lofti Zadeh of the University of California at Berkeley, who introduced the notion of fuzzy sets in the 1960s. The idea behind fuzzy technology was to introduce to machines the capability to make calculations when the values involved are ambiguous or fuzzy. Fuzzy logic differs from binary logic in that binary logic can only cope with a crisp 0 or 1 value, not allowing for anything in between; with fuzzy logic, there is scope for the concept of overlap, nearly true or depends. Where binary logic has to look for exact matches between what it encounters and what is in its memory – fuzzy logic can work with the concept of similarity, or near matches. A real life example of fuzzy inference would be inferring, from a general rule that if the water is very hot, add plenty of cold water, to add only a little when the water is moderately hot. One of the biggest potential markets for fuzzy technology is in control systems where, using fuzzy inference, a robot or machine can adjust to varying physical conditions. In fact, the first applications of fuzzy theory were primarily industrial, such as process control for cement kilns. And, in 1987, the first fuzzy logic-controlled underground railway system was opened in Sendai in northern Japan, where fuzzy controllers are used to make tube journeys more smooth-running, adjusting the braking and acceleration, the driver having simply to press the start button, leaving fuzzy logic to do the rest. A similar application is in the control of lifts, to minimise the waiting time.
Motorola has embraced the technology, and Omron has been into it for ages – what is it? Fuzzy Logic, and it may be heading your way. Sue Norris reinvestigates the technology, as it gears up to take on Europe and rediscover the US.
According to Omron, other major areas of research today include automation of steel and iron manufacturing, water purification, and manufacturing lines and robots; instrumentation and pattern recognition and analysis; design and judgement in train scheduling and trouble-shooting; and in computers and information processing for retrieval from databases and systems modelling. Omron got into fuzzy logic in the early 1980s, when the company was intro-duced to a Professor Yamakawa from Kumamoto University. An ensuing research effort yielded a fuzzy controller chip, now sold outside Japan as the FP-3000 Digital Fuzzy Processor-Controller, used in control mechanisms and simple pattern recognition. Omron has built on this technology to produce a robot which can grasp soft and squidgy materials as well as hard solids, based on inference. Omron also offers the FB-30AT fuzzy inference board, which can be inserted into an IBM AT expansion slot. It comes with Omron driver software which enables fuzzy inference to run with the user’s software. The FZ-10AT fuzzy software tool is an MS-DOS AT-bus-based personal computer software package designed to create rules and membership functions for fuzzy inference. The firm also makes fuzzy temperature controllers, for use in physical and chemical kit, industrial furnaces and semiconductor manufacturing equipment. Fuzzy applications available from Omron include a failure diagnosis and prediction system for machine tools, developed in conjunction with Komatsu Ltd,
which is reckoned to reduce servicing time by 24%; and a bank note feeding system for automatic cash dispensers, which keeps the gap between the rollers at an optimum level, to maintain a stable note-feeding process that isn’t affected by ambient humidity in the atmosphere conveyance conditions.
20% of products
Professor Zadeh, founder of fuzzy technology in the US, now acts as an senior advisor to Omron in Japan, which in its time has applied for some 700 patents for its fuzzy applications. By 1994, Omron says fuzzy logic will play a part in 20% of its product line. The first fuzzy product from Omron to be sold outside Japan was the fuzzy temperature controller, launched at the Hannover Fair last March. Omron’s Masayuki Oyagi cites the use of fuzzy logic in nuclear power plants, and NASA’s efforts to implement fuzzy logic control in space environments, as evidence that there is a market for fuzzy logic in the US and Europe. And, he says, if firms like General Electric Co Inc take up the challenge, we may see a consumer boom even larger than the one in Japan a couple of years ago. Whether that’s likely to happen is anyone’s guess, but Motorola’s pledge of commitment to the cause (CI No 1,865) is a sign that the US has now recognised a potential market for the imprecise technology.
