Integrated capacity planning over highly volatile horizons

Today production planning has to deal with highly dynamic markets and increasing uncertainties. Moreover, it has to take into account possibilities of the surrounding production network. By combining a queueing theory model with a stochastic, dynamic optimization approach, a method to support decision making in production planning was developed. Hereby, a Markovian Decision Process is solved to find cost minimal policies as reactions to volatile market demands for minimizing costs due to capacity adaptations, changes in process steps, and locations. The method was applied at an automotive supplier to find suitable system configurations and investment decisions for an uncertain future.