An encounter with Fuzzy Logic

Starting off with an informal introduction to Fuzzy Logic (FL), FL is a controlling method to manage processes/problems involving simple logical decisions. FL is different from traditional binary or digital logic because it can work in a more relaxed and free environment as compared to digital world which only works on a binary logic of 1’s & 0’s. Before giving a formal definition, an example will make simplify the working of FL. Consider two water valves, one digital and the other analog. Digital valve is designed to move one inch at a time, so if it has to open 4 inches, it will open as, open 1 inch, open 2 inch, open 3 inch and open 4 inches. Thus the digital valve will not be able to open at 1.2 inches, 2.4 inches or 3.4 inches because it’s values are based on binary logic and can only open in steps of 1, 2, 3, and so on. On the other hand, the analog valve works on FL; it can have values between 1 inch and 2 inch, making it more comfortable to open at any interval.

Formally, Fuzzy Logic is an efficient controlling mechanism for problem-solving that can be embedded in any software/hardware and making it possible to implement robotic technology in such a way that can comprehend fuzzy data in comfortable way, comparing over “True” or “False” values. “FL is a form of algebra employing a range of values from "true" to "false" that is used in making decisions with imprecise data. The outcome of an operation is assigned a value between 0 and 1 corresponding to its degree of truth. Fuzzy logic is used, for example, in artificial intelligence systems” (Fuzzy-Logic). Referring to the example above, control system based on FL will be able to comprehend the command in a form like, ‘level increasing rapidly, decrease valve size’ or ‘level falling, increase valve size.’ Whereas the traditional digital logic will accept command in a very precise manner, ‘level > 3, valveValue = --’ or ‘if level <= 2, valveValue ++.’ Now FL is being used in Artificial Intelligence to make robots behave more like humans and able to interpret imprecise, incomplete, noisy, unclear commands and perform specific actions. As it is a form of control system in problem-solving, its implementation is more like traditional control systems.