Modelos multivariables para la toma de decisiones en sistemas productivos: estudios de caso en la industria vitivinícola (España) y maquiladora (México) stars

Laburpena

Production processes, including acquisition of new technologies and supply chain, require the analysis of many variables for generating models in order to facilitate the decision-making process. However, the analysis of all variables is very difficult and requires too much time; therefore, in this thesis, possible techniques to deal with this problem are analyzed, and five multivariate models case studies are reported. Two models are structural equation models applied to the supply chain for wine production in La Rioja (Spain), where variables such as procurement, demand uncertainty and the importance of human factors are discussed. The information comes from 64 surveys applied to companies from that region, and the results highlight the importance of good relations with suppliers and customers, as well as knowledge management within the company. Two other multivariable models come from two independent studies in the maquiladora industry established in Ciudad Juarez (Mexico). In the first model the information comes from a survey applied to 144 managers applying Just in Time, which contained 31 benefits indentified in the literature that are obtained after their implementation. Through a factor analysis by principal components method and varimax rotation, results indicate that only four factors can explain 67.27% of total variability of all analyzed benefits, which are: inventory management, production process, human factors and economical benefits. In the second case of this sector, a structural equation model is reported, which analyzes the innovation and new product development process in this sector when demand from subsidiaries decreases. The information comes from 17 surveys applied to managers and the model integrates the product, production process and organization characteristics with the benefits obtained by companies and customers. The results indicate that these independent variables explain 31% of the variability of financial benefits. Finally, in the fifth case here reported a multi-criteria and multi-attribute model for evaluate and select an agricultural tractor in Colima (Mexico) appears. Six variables are integrated in the model: the initial cost, liters of fuel used per hour, operator safety, maintainability, and after-sales service. The AHP technique is used for weighting the attributes evaluated, and TOPSIS is used to select an alternative from a set of six. It has been demonstrated by this case studythat the model is easy to use: it does not require the use of specialized software and are farmers themselves who made the evaluation process.