A numerical example of total production maintenance and robust scheduling application for a production system efficiency increasing

In the paper, the proposition of application of two methodologies: the predictive scheduling and Total Productive Maintenance - TPM to increase efficiency of a production system is presented. In this paper, an example of problem of predicting a time of a bottle neck failure is presented. Using the Statistica program, histograms that show the graphical relationship of a number of observations and failure-free times of the bottle neck for historical periods are created. The fitting of the histograms to the theoretical distributions: normal, exponential, gamma and Weibull using appropriate tests (for example the Kolmogorov-Smirnov test for normal distribution) is researched. After finding distribution and setting parameters for historical periods, for the next scheduling horizon values of parameters are extrapolated using the regression method in the Statistica program. For the bottle neck various reliability characteristics are computed. Having the Mean Time To Failure (MTTF) and Mean Time of Repair (MTTR) of the bottle neck, robust schedule is generated. At the time of the predicted failure, preventive actions and technical survey of the machine are scheduled. The production system is modeled in the simulation program - Enterprise Dynamics 8.1.