Position tracking and motion prediction using fuzzy logic

Patrick Rodjito, Colby College

Document Type Honors Thesis (Open Access)

Abstract

Fuzzy logic (FL) is a highly debated topic among mathematicians. Some mathematicians think that FL does not really qualify as something new and is actually no different than probability theory, while others belief that it brings nothing to the table that standard binary logic has not. However, many fought for the notion that FL and standard logic complement each other, and only upon both can all scientific thought, probabilistic or precise be built. FL is the implementation of one's logic and understanding of this imprecise, subjective, human world and made very real the possibility of an AI whose ability to adapt can rival ours. It is no wonder that many industrialists and scientists embraced FL as a solution to complex problems despite the denouncement of their peers. FL is already used in some home appliances and subway systems and it's just a matter of time for FL to become ubiquitous. Unlike with traditional, computationally expensive algorithms, with FL one does not have to own a fast computer or very precise measuring instruments to get good results. At lower costs, FL provides a simple way to arrive at a definite conclusion based upon vague, ambiguous, imprecise, noisy, or missing input infonnation. This is a facet of AI that appears to resemble human intuition in an important way: the ability to make quick and urgent judgment calls, on the spot, without the need to be absolutely certain of the input. The benefit brought by a low-cost solution like FL to mankind is widespread, as not only researchers of high caliber but also everyone can benefit from it. In the pages to come, we will delve, as deeply as the diagrams and expositions will allow, into the implementation of a fuzzy system designed for the purpose of tracking and predicting the motion of light-colored objects on dark background. Before that, however, is a brief history and introduction to FL.