Mass customization requires frequent product changeovers thus leading to the need of manufacturing systems endowed with flexibility and reconfiguration capabilities, in order to be robust to changes in the production scenario. Therefore, manufacturing companies face a relevant risk when taking strategic decisions about which system resources should be acquired. This risk can be mitigated by exploiting performance evaluation models, such as analytical models and Discrete Event Simulation, that are effectively adopted to estimate the performance of possible system configurations. However, current decision-support tools for optimizing system configurations can be only loosely coupled with performance evaluation models, hence undermining the actual optimization of the system itself, even more if production requirements may evolve in the future. This work presents an analytical methodology to support the optimization of manufacturing systems configuration and reconfiguration subject to evolving production requirements. The methodology integrates a stochastic analytical model for performance evaluation of manufacturing lines into a mixed integer programming problem, by means of performance linearization. The advantage of using the proposed methodology is shown on a line configuration problem, where buffer capacities and machine capabilities have to be jointly optimized, in order to minimize costs and satisfy the target performance.
