Building a case base for the non-conformance problem solving in the aluminum extrusion process

Abstract

This work proposes the application of Case-Based Reasoning (CBR) to build a knowledge base system for seeking and providing solutions for non-conformances that take place in the aluminum extrusion process. CBR is an important area of Artificial Intelligence (AI) that is used to solve problems based on the knowledge accumulated previously from known scenarios, providing a solution to both recurrent and new problems. The CBR cycle is characterized by having four stages known as: retrieval, reuse, revision, and retention. These stages interact with the knowledge base in order to seek one or more solutions to the problem. The steps for the development of the relational model of the knowledge base according to CBR are as follows: (a) identification and classification of the non-conformances, (b) information gathering of situations that result in non-conformances, which include: records of non-conformances, experience of operators with nonconformances, literature on non-conformances in the aluminum extrusion process, and (c) the structure definition for the diagnosis of cases of non-conformance and support in decision-making. The final step consists in storing the cases in a relational database, which will correspond to a knowledge base. New cases may be added in the knowledge base, as they occur. The structure of the cases in the knowledge base will be important for its provision for decision-making in the aluminum extrusion process. In a further work it is intended to implement a webbased distributed system to support the inclusion of new cases (that may occur in different geographic locations and company conditions) as well as a fast search for solutions to non-conformances that occur in the aluminum extrusion process.