Sistema experto híbrido para la programación de tareas en sistemas de servicios

Abstract

This doctoral thesis studies the decision-making problem known as scheduling on service systems, which has a scope of the operational level where the decision must be made in the short term. Thus, we need methods to understand how a person approaches the problem. The process starts when a person must identify a scheduling problem, next to a performance measure and later a solution technique. In addition, given the complexity of programming problems, which is a combinatorial problem, it adds to the characteristics of service systems such as intangibility, heterogeneity, co-production and variability, which makes it difficult to classify and identify problems. In the same way, it is often found that this process does not use computer tools, professional staff and the knowledge of this process is not organized, so they do not have an adequate knowledge management of this operational function. To solve this problem, we propose a hybrid expert system that consists of two subsystems. The first one determines the type of service and its main features, according to a sector (primary, manufacturing, or service), a section, and a division of ISIC (International Standard Industrial Classification). The second one takes the output of the first subsystem, and a set of additional variables to determine the type of scheduling problem, its performance measure and a suggest solution technique, using a three rule-based system in a waterfall way.