off introduction:

Producing companies face the challenge of staying competitive in terms of price, quality and delivery time while at the same time maximizing their profit. They act in a changing market environment which requires strategic decisions that guarantee the future success of the company. More flexible manufacturing technologies like additive manufacturing and the digitalization of industrial processes allow smaller production lot sizes or even individualized products. This leads to more complex production networks with a higher number of small and decentralized factories. Due to the increasing complexity and internal influences (e.g. production methods, machinery and available manufacturing time) as well as external ones (e.g. labour costs, transport times, transport costs and production) it is often not easy to identify the optimum distribution of production facilities in a global production network. Therefore a computer simulation of the production network becomes increasingly important to assure the profitability of a company.

Compared to reality, these models are always hampered by inaccuracies stemming from simplified assumptions or intrinsic correlations of model parameters that are difficult to control. An optimal solution gained from the computer model is therefore not necessarily the best for all sets of parameters. A careful study of the uncertainties in input parameters and their propagation is hence required to assess the quality of the solution. Especially with regard to the higher number of involved entities, the containment and control of uncertainties in the simulation of production networks is the key to a reasonable network design.