General Reviews:

For all the wonders of the theory of fuzzy logic, most applications would not exist without the advanced technology of sensors, chips and high speed computing.

During a recent international conference, Lofti Zedeh paid tributes to Japanese scientists and engineers for their key role in developing the practical applications that made fuzzy logic popular. Mr Zadeh believes that only a small fraction of the potential of fuzzy logic has been tapped. In the future, he sees more complex applications and an increasing role for neural networks in deriving the inference rule and membership functions from observations.

Common sense, human thinking and judgment are the lures of fuzzy logic. Combining multivalve logic, probability theory, artificial intelligence and neural networks, fuzzy logic is a digital control methodology that simulates human thinking by incorporating the imprecision inherent in al physical systems.

Fuzzy logic is only one of whole range of emerging technologies which are at the forefront of the user friendly revolution. Artificial neural networks have already been used in conjunction with fuzzy logic to offer more powerful solutions, and their mutual coexistence will lead to further developments in the coming decade. Researchers are already aiming for speech recognition and robotic 'eye-brain' applications using fuzzy logic.

The exciting thing about fuzzy logic is that is appears to have such broad applicability that future uses are likely to be surprises. Researchers talk glibly about employing fuzzy logic to recognize natural speech in which two speakers never pronounce a sort exactly the same way or to help a robot notify obstacles on a factory floor. Such applications are not what Mr Zadeh anticipated about 30 years ago.

Worldwide R&D Activities:

Around the world, particularly in Japan, intense work is being conducted in the field of fuzzy logic. One of the centres of Japanese activity is the LIFE ( Laboratory for International Fuzzy Engineering Research) institute in Yokohama. Modelled after the Fifth Generation Institute, the LIFE institute not only enjoys an annual budget of $10 million but also has many industrial member companies, including Canon, Fuji, Hitachi, Matsushita Electric, Mistubshi Electric, IBM Japan and Thomson Japan.

According to Machio Sugeno, professor at Tokyo Institute of Technology, the fuzzy computer will not replace a digital computer but will be better suited to solving real world management and decision problems that must be tackled on a grand scale. Mr Sugeno heads a five year, $44 million project, a joint venture between universities and companies devoted to basic research on fuzzy logic. In addition, several universities as well as companies are pursuing their own extensive activities in this area.

While Japanese companies have so far largely restricted themselves to their domestic market, they are now attempting to gain foothold in the European market with fuzzy controlled products. More than50 Japanese companies are developing fuzzy applications, and two multi million dollar consortia backed by the biggest corporations in Japanese technology have been formed to further the research work. In the US the aerospace industry chiefly Boeing and NASA have shown the greatest interest in the field of fuzzy control systems. In Germany, as in the rest of Europe, research in fuzzy control is being pursued.

How Fuzzy Logic Works ?

Let us examine a few applications wherein fuzzy logic has several advantages over the conventional binary logic.

In a room heater, the conventional thermostat circuit operates as a simple on/off switch. The cut - off temperature is selected at which the heater is activated when the actual ambient temperature fails below that level. Fig. (1) illustrates the operation with the cut-off temperature set at 20°C. When the temperature reaches this level the heater is turned off. This approach does not take into consideration when the temperature is between, say 15°C and 30°C. This results in the temperature in the room being excessively low or high. There is no setting for values in between these conditions.

Now using fuzzy logic controllers the system works in the grey or fuzzy areas where the definition of cold/cool/warm/hot is less clear and more open to interpretation. Temperature ranges are considered as overlapping - 20°C is described as 60% warm and 20% cool. The crisp input value of 20°C is translated to a truth value of 0.6 in the set ' WARM' and 0.2 in the set 'COOL' ( Fig. (2) ).
During evaluation, the entire set of rules is taken into consideration. The logic controllers sense the ambient temperature of the room and set the device accordingly, resulting in the heating power being progressively varied. A saving of about 20% in the consumption of electricity is obtained.

Take another example. A person who is 2 meters can be considered as 'tall', one who is 1.7 meters as ‘medium’, and a person who is 1.2 meters as ' short'. But how to classify a person who is 1.5 meters ? Using fuzzy logic, the degree of the membership of 'medium' can be evaluated easily. Fig. (3) Represents graphically how this can be achieved. A height of 1.2 meters has a 60% membership in the set 'short', while a height of 2 meters has 55% membership in set 'tall'. The richness of fuzzy logic is in its ability to deal with vague of imprecise value and in its likeness to human thinking.

Design of a fuzzy system need not have well defined set of rules; it can start with a general idea of how system should work. This idea may consist of defining input and output ranges. For example, one might specify that steering angle for a vehicle is to be in the range of +/- 30 degrees. That range would be normalized and mapped to the fuzzy system as range between -1 and 0 to +1. Then, within that range, one has to define what constitutes a small negative, large positive, zero or small positive angle. These are the individual membership functions that assign the values, say for example, -24 degrees is 0.8 of the function negative large and so on.

Fuzzy logic works by turning the hard-edged world of binary logic in to more natural human- like reasoning, since people use rules of inference based on vague concepts and approximate knowledge. For example, when driving a car, one knows when to apply the brakes in the face of an approaching car. We make this judgment based on our power of reasoning. We don't depend on the exact distance when the car in the opposite direction approaches, nor is it possible. In technical terms, we must calculate the value of the control variable (the pressure on the brakes) from the data or input variable (the speed of the car and the head - way distance). We must do this almost continuously to account for input changes and so ensure effective control. Fuzzy logic provides us with a method of carrying out these calculations with ease.

One of the main advantages in developing the translation controllers is that the engineers need not construct a detailed mathematical model of the system in advance. Performance is perfected through simulation and experience.

In space research, NASA has found that the results of simulation have been encouraging, especially in terms of fuel efficiency. In holding position with respect to another spacecraft, the fuzzy controllers required significantly less acceleration - that is smaller increments of position change than did the human controlled simulation. In overall maneuvers, the fuzzy controller has shown 20 to 70% better fuel efficiency than the currently used auto pilot and the best simulation runs of human pilots.

One of the reasons for the increasing popularity of fuzzy logic is that it offers a very simple, intuitive way for engineers to describe a complex problem using the design methodology of fuzzy set theory. Designers are turning to the fuzzy methodology as design complexity is simplified. It typically takes only a few rules to describe systems that may require complex mathematical and software routines.

The adoption of fuzzy logic has given many companies a competitive edge in terms of time to market for products. One estimate claimed that by the year 2000 more than 90% of the embedded control market would employ fuzzy logic in one form or the other.

Applications using Fuzzy Logic:

While fuzzy logic evolved in the United States, practical applications are mushrooming in Japan. Camcorders, TVs, Cameras employing new electronic circuits that promise enhanced images, washing machines, elevators, and anti lock brake systems employing fuzzy logic have been introduced.

The first person to demonstrate the practical possibilities of fuzzy algorithm was Abe Mamdani, an electrical engineer at Queen Mary College, London. In the early 1970s, he applied his algorithms to control the pressure and speed of a steam engine.

In 1980, a Danish company F. L. Smidth, used fuzzy logic to control the operation of a cement kiln. This was the world's first industrial application of the fuzzy theory.

The mid 1980s saw the first industrial applications in Japan. Fuji Electric Company used fuzzy theory to control a water purification plant in Akita City, and Hitachi developed a fuzzy predictive system to operate the automated underground trains in Sendai City. In 1990, the fuzzy wave reached the consumer product. Cameras, washing machines, microwave ovens and dozens of other consumer goods featuring fuzzy logic began to appear.

Japanese engineer, for example now use fuzzy logic to improve the efficiency of automated transmission in cars, to control the injection of chemicals in plants for purifying water, and to simulate the shutdown of a nuclear reactor with a computer. Japanese industry now offers more than 50 consumer goods featuring fuzzy logic. These appliances are extremely easy to operate and the performance is better than of conventional models.

In the medical field, expert systems using 'fuzzy inference' help doctors to diagnose diabetes and prostatic cancer, while anaesthetists rely on fuzzy logic to control the blood pressure of patients undergoing surgery. NASA engineers developed natural language rules to run the controller and are testing in it a multi vehicle simulation by substituting the fuzzy controller for the simulator's normal human inputs. The rule base was extracted from the experience of human operators, and the efficiency of the controller was tuned based on flight profiles recorded from actual missions and simulations.

NASA is also exploring other application of fuzzy control in space. Among the projects being considered is the use of inexpensive cameras for constant tracking of objects around a space station. Fuzzy control can contribute to collision-avoidance systems, robot arm control and traffic management.

It is estimated that by the end of the century, the electronic components in automobiles will double. Among the areas of interest to fuzzy based system design are emission control system that would continuously monitor the exhaust gages and make adjustments to the carburettor and ignition system to keep levels of hydrocarbons and other toxic elements within specified limits. Interior climate control could adjust to the number of passengers, just as some fuzzy controlled air conditioners do today. Fuzzy controlled digital signal processors are being investigated for use in interior noise cancellation systems.

Today's digital anti lock braking systems work by cycling the brakes rapidly on and off. A fuzzy controller would provide smooth anti lock braking by adjusting hydraulic pressure in response to slight variations in wheel rotation.

One fuzzy item already under active development is a multiple mode automatic transmission. By sensing rotation torque, engine speed and throttle position, a fuzzy controller could determine the proper gear shift points. By selecting a different set of rules and membership functions, the driver could set the transmission for smooth shifting, optimal economy shifting or sport shifting.

Matsushita, better known for its Panasonic and Technics brand names, has several household appliances which employ fuzzy logic. For example there is the 'Aisaigo Day Fuzzy' washing machine. 'Aisaigo' in Japanese language literally means 'beloved wife'. The washing machine features no less than 600 washing cycles. The fuzzy controller relies on a battery of sensors to determine which one to choose. These sensors check the size and weight of the wash load, how dirty the clothes are and what type of detergent is being used. The machine even takes into account what type of fabrics are being washed. All the user, or Aisaigo, needs to do is press one button and the system looks after itself just as a housewife would do.

Matsushita's Canister 7 vacuum cleaner senses what type of floor surface it is cleaning, whether carpets or smooth surface. Coupled with a dust quantity sensor, suction power of the system is varied accordingly, thereby saving energy.
Matsushita is already improving its fuzzy logic technology by replacing ordinary microcomputer with a special processor tied to a simple neural network. These improved devices can handle more sensors and make complex decisions. Buyers are already faced with the choice of first and second generation fuzzy logic appliances from the company.

Mitsubishi's fuzzy logic air conditioners differ from conventional models by doing away with a simple thermostat. Internal and external air temperature and humidity are constantly evaluated and applied to a set of 50 fuzzy logic rules. Cooling power is varied progressively instead of being simply switched on or off by a thermostat, with a power saving of up to 25 per cent.

Sony uses fuzzy logic to recognise handwriting and calligraphy in its pen based protable personal computers. This is an ideal application for the Japanese duetot the nature of their writing.

The most common fuzzy logic application is in camera autofocus system. Sanyo, Fisher, Richoj and Olympus video camcorders use such systems to make focusing more reliable. An example is that, once set, the subject will remain in focus even if it moves about on the screen. Conventional systems would produce a shaky picture.

One of the first fuzzy camcorders, the Canon handheld H800 which was introduced in 1990, adjusts the autofocus based on 13 fuzzy rules. Sensors measure the clarity of images in six areas. The rules take up about a kilobyte of memory and convert the sensor data to new lens settings.

One of the most complex fuzzy systems is a model helicopter, designed by Michio Sugeno of the Tokyo Institute of Technology. Foure elements of the craft the elevator, aileron, throlle and rudder respond to 13 fuzzy voice commands, such as 'up', 'land' and 'hover'. The fuzzy controler can make the craft hoer in place, a difficult task even for a human pilot.

Fuzzy Chips Detail:

At present, simple every day microprocessors are used and fuzzy logic software is designed for specific applications. However AT&T's Bell Laboratory has demonstrated the reality of dedicated fuzzy logic processor chips. Also there is much talk of combining fuzzy logic with neural network computing in efforts to mimic the human brain responses. Prospects for a fuzzy future appear bright and exciting.

Fuzzy products use both microprocessors that run fuzzy inference algorithms and sensors that measure changing input conditions. Fuzzy chips are microprocessors designed to store and process fuzzy rules. In 1985, Masaki Tojai and Hiroyuki Watanabe, then working at AT&T Bell Laboratory, built the first digital fuzzy chip. It processed 16 simple rules in 12.5 microseconds, a rate of 0.08 million fuzzy logical inferences per second.

Currently the most important manufacturer of fuzzy control hardware and software outside Japan is Togai Infra Logic Inc., a software company whose main products are a development system for efficient fuzzy control applications ( based on standard microprocessors) ad special fuzzy accelerator boards. Togai Infra Logic is the only North American company producing a fuzzy logic computer chip set. Using this set, graphics and language software, programmers can write fuzzy rules for control or analysis systems. A Togai designed system is used in Mitsubishi's air conditioner and other products.

ADDING THE POWER OF HUMAN THINKING TO COMPUTERS:

Since 1987, the Japanese city of Sendai has a driver less subway train that is controlled by a system developed by Hitachi's Dr. Seiji Yasunobo. The train ride is perfectly smooth and impressive, and the braking is so effective that the train can stop within centimeters of a predetermined spot on the platform. Grab handles are not provided inside the compartments, since there are absolutely no jerks felt by the commuters.

Toshiba has developed a control system for an elevator which keeps the waiting time for a lift to arrive to a minimum. One need not wait for more than 30 seconds for the lift, even though the building is 43 stories tall.

A washing machine from Matsushita senses the quality and quantity of dirt in the clothes, the weight of the load and the type of fabric. It then adjusts the wash cycle, temperature and detergent level accordingly.

A television set automatically adjusts its volume as the ambient noise in the room increase or decreases and alters its brightness as the intensity of the light in the room changes.

In 1965, Lotfi A. Zadeh, working as computer scientist at the University of California, proposed a mathematical way of looking at vagueness that a computer could deal with. He called the new approach fuzzy logic.

In Japan more than 2000 products ranging from rice cooker to subways use fuzzy logic. The National Aeronautical and Space Administration (NASA) has found that in simulation of space shuttle maneuvring and docking, fuzzy logic controllers perform much better than a standard autopilot or even an experienced human pilot.

To understand fuzzy logic we must begin with the working of present day computers. The operation of present day computers is based on simple yes/no logic (binary logic), which is widely different from the information processing inherent to human thinking. That is why commonsense and flexible judgment evaluations are difficult for present day computers.

Boolean logic is good for orthodox computing, based on the binary system, but it does not work for vague or imprecise values, since these computers normally solve problems by breaking them down into a series of yes - or - no decisions, represented by 'ones' and 'zeros' ( binary logic).

Fuzzy logic lets computers assign numerical value that fall between 'ones' and 'zeros', there being no clear dividing line between these values. Instead of statements being only true or false, fuzzy theory sets up conditions as slow, medium and fast. Fuzzy logic is now applied to help computers simulate the vagueness and uncertainty of our thought processes and languages.