Modelo estadístico para pronosticar el porcentaje de éxito en una implementación de herramientas Lean Manufacturing: 5 s, kanban, gerencia visual y trabajo estandarizado

Abstract

This academic essay is carried out with the objective of creating a statistical model to predict the success of an implementation of Lean Manufacturing tools: 5 S, Kanban, Visual Management and Standardized Work, in the city of Cuenca. Due to different circumstances such as cultural change or commitment throughout the organization, among others, the required results have not been achieved therefore companies waste time and money. This research provides guidance to companies planning to adopt the Lean Manufacturing philosophy through a statistical function that forecasts the success or failure of implementation. In this quantitative study the binary logistic regression method is used. After a bibliographic review, the variables that in most cases guarantee the success of an implementation were selected and validated. Then, through the Chi square dependency test, it is determined that 10 independent variables have correlation with the dependent variable binary. But, through the method of Steps forward of Wald, there are 3 variables that contribute statistically to the model. With the Hosmer-Lemeshow test of goodness of fit we conclude that it is an acceptable model. Finally, the overall prediction grade is 92.3%, demonstrating that the model is reliable for predicting the success of an implementation. Data processing was performed using the IBM SPSS® statistical package.