Implementation of a Fuzzy Relational Database. Case Study: Chilean Cardboard Industry in the Maule Region

TL;DR

This paper presents the implementation of a fuzzy relational database model, GEFRED, in a Chilean cardboard manufacturing company, to handle imprecise and uncertain quality control data using fuzzy set theory.

Contribution

It demonstrates the practical application of the GEFRED fuzzy data model within a real-world relational database for manufacturing quality control.

Findings

Successful modeling of fuzzy attributes in a relational database

Enhanced handling of uncertain quality data in manufacturing

Integration of fuzzy logic with traditional database systems

Abstract

The international database community refers to the manipulation of data with inaccuracy and uncertainty using the term fuzzy, which has been translated into Spanish as "borroso" and into French as "flou". Semantically, this term conveys two main ideas: first, the natural concept of ambiguity or vagueness in human reasoning, and second, its connection to fuzzy set theory, fuzzy logic, and possibility theory, as developed by Zadeh between 1965 and 1977. This article explores two key aspects: the attributes of the fuzzy data model GEFRED (GENeralized model for Fuzzy RElational Database) and their implementation in a Relational Database (RDB). The modeling of these attributes was conducted in a Chilian cardboard manufacturing company located in the Maule Region, where the described phenomena involve imprecise and uncertain attributes and values. Specifically, our focus is on the knowledge related to the manufacturing process of coated cardboard, particularly the quality control process for finished products in the company's Conversion Department. The quality of these products, categorized as either stacks or rolls, is characterized using both classical and fuzzy attributes. Classical attributes are typically measured with physical instruments, whereas fuzzy attributes are assessed through human senses, primarily sight and touch, as perceived by the operators.