Lean production and business efficiency: An artificial neural network analysis in auto parts companies

Abstract

The aim of this work is to determine if it is possible or not to identify significant differences in performance, in several and simultaneous dimensions, among players applying lean production techniques, inside the Spanish and Portuguese auto parts industry. The automotive industry is a pioneer and outstanding group in the application of the lean production techniques. In this case the techniques were associated to five Lean Dimensions: (1) Manufacturing Flow (2) Process Control (3) Inbound Logistic (4) Organizational Design and Culture, and finally (5) The Lean Metrics. An artificial neural network (ANN) was used due to its flexibility and absence of “a priori” scenarios. With the purpose of knowing and measuring the application degree of the different techniques of the Lean Production System in the auto parts industry, a survey was carried out through Internet with first tier suppliers of automobile components. The questionnaire used has four parts, the last of which made reference to each one of the five dimensions of the model of lean production. It was answered by 49 companies, although it was necessary to eliminate some of them because they were not completed. In particular the final analysis was made with only 31 complete (valid) answers. The most interesting part of our analysis is the ascertainment of the main hypothesis about the link between techniques and the best results. The finding is that only by the application of a set of techniques, you will not have a success guaranteed. To obtain a good result, you not only must be lean, you must be something else. In any case our analysis has revealed, at least for those players that don't apply the lean approach, or that not apply it consistently; that is very difficult to obtain over average outcomes.