Joint Maintenance and Production Planning by Maintenance-Optimal Swapping

Degradation is an inevitable course of any manufacturing tool, machine or system. The degradation of the health state of manufacturing tools results in some sort of an ineludible maintenance action which could be both costly and occurring during critical production time. In many manufacturing systems, a fleet of identical machines are assigned different tasks (or products) towards satisfying production requirements. We re-introduce the maintenance-optimal resource allocation planning scheme [1] (presented in MSEC2012) and focus on the solution of the generated mathematical model. The planning scheme, denoted as Degradation Based Optimal Swapping (DBOS), incorporates the optimal implementation of swapping scheduled tasks (or products) and allocating maintenance actions throughout a finite time horizon. The objective is to minimize projected maintenance costs and/or utilize the manufacturing productivity towards prescribed logistics and/or production goals. A DBOS-specific branch-and-bound-based optimization algorithm is developed to address the complexity in the generated model. Numerical results will demonstrate the effectiveness of the algorithm in comparison to standard optimization algorithms. DBOS planning scheme coupled with the proposed algorithm succeeds in establishing substantial savings in the simulated case studies which amount up to 70% of the estimated maintenance costs in comparison to the scenario where fixed scheduling is applied.