Stochastic discrete-event simulation is in fact a statistical experiment (Banks et al. 1996), thus it was possible to analyse the generated schedule in order to determine the expected value of the given planning considering possible delivery delays. This method allows for a much easier approach to deal with the dynamic aspects of SPSCs as its stochastic and deterministic features were considered in two different steps, through the before mentioned MILP and simulation, respectively (Frazzon et al. 2014).
The simulation does not only enable the planner to determine the expected outcomes of possible decision alternatives, but also makes it possible to conduct sensitivity analyses of the scenario’s input variables. The creation of what if scenarios (e.g. machine breakdowns), can give insides on the outcome of the expected cost and service level of a given planning.
This set of advantages make the combination of simulation and MILP an important tool in the decision making process (Frazzon et al. 2014).
Through the application in a use case, it was possible to verify that the proposed procedure allows for the minimization of total costs while ensuring service levels in a predefined range for the operational planning (Frazzon et al. 2013; 2014).