In the design of a manufacturing system, the design specification is often suggested by a design team. Managers are interested in verifying that this specification will satisfy the production requirements. Because the future production environment will likely differ from the one assumed, it is important to determine in which situations the suggested design becomes unacceptable. This paper suggests an approach that allows determining which uncertain parameters are important and which combinations of these parameters can lead to an unacceptable design. This approach combines several methods, namely, simulation, bootstrapping, and metamodeling. The methodology is explained and illustrated through a stochastic simulated manufacturing system, which includes uncertain parameters related to the arrival and the processing times of jobs. This example shows the conditions under which the system does not meet the requirements.
|Title of host publication||Proceedings of the 39th International Conference on Computers & Industrial Engineering (CIE 39)|
|Editors||M.I. Dessouky, I. Kacem|
|Place of Publication||Piscataway, NJ|
|Publication status||Published - 2009|