On November 30 in 1964, renowned Azerbaijani-American computer scientist and professor Lotfi Zadeh submitted the paper “Fuzzy Sets”. It is a pathbreaking paper that introduced the world to his mathematical framework called “fuzzy” logic. In 1965, he published “Fuzzy Sets,” which has since been cited by scholars nearly 100,000 times. This year, Google Doodle honoured Lotfi Zadeh for his massive contribution and impact in the scientific world. Fuzzy Logic has found applications in various domains, from robotics to artificial intelligence.
Image: Google
What is Fuzzy Logic Exactly?
We know that the word “fuzzy” refers to something that is vague. In the real world, we often encounter situations when we are not sure if a given statement or problem is true or false. Fuzzy logic is a type of many-valued logic where the true value of variables can be any real number between 0 and 1. It works on the concept of partial truth (the idea of truth can spread from completely true to completely false). Conversely, the truth values of variables can only be integer values 0 and 1 in Boolean logic.
Subscribe to our Newsletter
Join our editors every weekday evening as they steer you through the most significant news of the day, introduce you to fresh perspectives, and provide unexpected moments of joy
Your newsletter subscriptions are subject to AIM Privacy Policy and Terms and Conditions.
The Fuzzy Logic based on natural language processing becomes really helpful for problems that come with uncertain reasoning. Here, any logical system can be “fuzzified”. Fuzzy logic systems come with the advantage of a simple structure and flexibility, allowing the rules to be modified easily.
Architecture in a Fuzzy Logic System
Rule Base
The if-then conditions and other rules for decision making are stored here. Due to updates and improvements in fuzzy theory, the number of fuzzy sets of rules has reduced significantly.
Fuzzification
This step helps convert crisp numbers to fuzzy sets. The information is converted into a form that the system can search for in its rule base. The converted inputs reach the control system for processing.
Inference Engine
Here, the degree of match between fuzzy input and the rules is determined. Based on this information, the inference engine determines which rules need implementation. Finally, the rules are used to develop control actions.
Defuzzification
Here, defuzzification is done to convert fuzzy values into crisp values. These crisp inputs form the output of the Fuzzy Logic architecture.
From Robotics to Warfare
With a simple structure and capability of dealing with uncertainty, Fuzzy Logic finds applications in different sectors, some of which are:
Robotics
A key area of application for fuzzy logic is robotics. This method has been applied in robot navigation largely. One of the major properties of fuzzy logic is the capability to handle partial and erroneous input signals. This makes it suitable for applications in robots while navigating in unpredictable surroundings.
Automotive Systems, Control Systems
Another important area of application of fuzzy logic is in automotive systems for controlling the traffic and speed, and intelligent management of highways, among others. It helps in improving the efficiency of automatic transmissions. Fuzzy Logic gives a more efficient way to solve control systems such as temperature controllers in washing machines, ACs, humidifiers, heaters, and anti-break lock systems.
Warfare
Fuzzy logic is helpful in warfare mechanisms as well in areas like underwater target recognition. In aerospace, it is used in satellite altitude control, controlling the altitude of the spacecraft, etc.
They find extensive use in decision making processes in business, banking (fund management, stock market predictions), medicine, chemical industry, securities, electronics, manufacturing and many more.
Though this pathbreaking idea has found applications across a wide spectrum of sectors over the years, it also suffers from certain issues. The possibilities that a fuzzy logic system generates may not always be accurate. Hence, it is not suitable for problems that require high accuracy. They also require extensive testing for verification and validation.
