Fuzzy logic is at foremost a theoretical concept for handling uncertainty and converting numerical values into linguistic variables. It was described in the past in a series of academic papers. If the aim is to build a fuzzy inference system, an out-of-the box software package is helpful to demonstrate what is wrong with the idea. One of well known products is called Matlab fuzzy logic toolbox and was released recently in the version 2.5.[1]

The program is a plugin for the Matlab mathematical software. If provides a GUI interface for defining the membership function and gives realtime feedback for the values written in the fuzzy system. A possible application is to create a fire monitoring and warning system.[2] In the first step, the user enters the features, then he defines the rules and after a click on the start button, the system calculates if a fire is there or not. The problem with Fuzzy logic in general and with the fuzzy toolbox in detail is, that the prediction accuracy is low. Instead of calculating if a warning message has to be triggered or not the software shows a color gradient for Red-green-blue and it's up to the user to belief the system or not.

In a normal fire system the feature “flame is present” can be set to true or false. It's a boolean value which works great. In contrast, the fuzzy inference system is working with uncertainty in mind. That means, the system doesn't know if there is a flame or not. This prevents a clear description of the reality and a wrong alarm is the result.


  1. Release notice version 2.5,
  2. Sarwar, Barera, et al. "Design and Application of Fuzzy Logic Based Fire Monitoring and Warning Systems for Smart Buildings." Symmetry 10.11 (2018): 615.
Community content is available under CC-BY-SA unless otherwise noted.