Flexible Job Shop Problem with Variable Machine Flexibility

Abstract

We analyze machine flexibility in the context of the flexible job shop problem. In general, the decisions about machine flexibility are taken in advance (the data sets already contain this information). Based on a mathematical model from literature, we proposed an extension that, aiming to minimize the makespan, considers the possibility of investing in flexibility. In the proposed mathematical model, the investment in upgrading a machine for a specific product becomes a binary decision variable and there is a global budget on the investment decisions. Computational results, using a data set from literature, are presented and shown that the same makespan is obtained with a 20% of reduction on the level of flexibility. For 40% of reduction on the level of flexibility some instances become unfeasible and for some instances the makespan increased. However, an interesting aspect is that, for some instances the makespan decreased, because the solver could reach a feasible solution with better quality.