Development of a decision support system for decision-based part/fixture assignment and fixture flow control.

Abstract

An intense competition in a dynamic situation has increased the requirements that must be considered in the current manufacturing systems. Among those factors, fixtures are one of the major problematic components. The cost of fixture design and manufacture contributes to 10-20% of production costs. Manufacturing firms usually use traditional methods for part/fixture assignment works. These methods are highly resource consuming and cumbersome to enumerate the available fixtures and stabilise the number of fixtures required in a system. The aim of this study was to research and develop a Decision Support System (DSS), which was useful to perform a decision-based part/fixture assignment and fixture flow control during planned production periods. The DSS was designed to assist its users to reuse/adapt the retrieved fixtures or manufacture new fixtures depending upon the state of the retrieved fixtures and the similarities between the current and retrieved cases. This DSS combined Case-Based Reasoning (CBR), fuzzy set theory, the Analytic Hierarchy Process (AHP) and Discrete-Event Simulation (DES) techniques. The Artificial Intelligence (AI) component of the DSS immensely used a fuzzy CBR system combined with the fuzzy AHP and guiding rules from general domain knowledge. The fuzzy CBR was used to represent the uncertain and imprecise values of case attributes. The fuzzy AHP was applied to elicit domain knowledge from experts to prioritise case attributes. New part orders and training samples were represented as new and prior cases respectively using an Object-Oriented (OO) method for case retrieval and decision proposal. Popular fuzzy ranking and similarity measuring approaches were utilised in the case retrieval process. A DES model was implemented to analyse the performances of the proposed solutions by the fuzzy CBR subsystem. Three scenarios were generated by this subsystem as solution alternatives that were the proposed numbers of fixtures. The performances of these scenarios were evaluated using the DES model and the best alternative was identified. The novelty of this study employed the combination of fuzzy CBR and DES methods since such kinds of combinations have not been addressed yet. A numerical example was illustrated to present the soundness of the proposed methodological approach. Keywords: Decision support systems, case-based reasoning, analytic hierarchy process, fuzzy set theory, object-oriented methods, discrete-event simulation, fixtures.