FLEXIBLE PLANNING METHODS AND PROCEDURES WITH FLEXIBILITY REQUIREMENTS PROFILE
Uncertainties in supply and/or demand combined with rolling horizon planning necessitate flexibility in a dynamic production planning process. In rolling horizon planning, production plans are revised when new information becomes available after time rolls forward on the planning horizon. Frequent adjustments to production plans can lead to the increase of instability in the production system, and result in a surplus or deficiency in production resources. These frequent replanning adjustments and extra efforts to cope with uncertainties in the system lead to syndrome referred to asnervousness . Frozen horizon and other planning approaches attempt to provide insights on how to mitigate nervousness. However, most of the existing studies do not consider the flexibility aspect in production plans, or provide only partial flexibility to handle the fluctuating demand. In this research, we propose to study mathematical optimization for Flexibility Requirements Profile (FRP) that is designed to mitigate nervousness by enforcing bounds on production plans in order to maintain a desired degree of flexibility. Instead of 0% flexibility in the frozen horizon planning and 100% flexibility in the make to order planning, the proposed FRP optimization model allows the trade-off between conflicting planning objectives, stability and responsiveness of the production system.In this dissertation, we first evaluate the effectiveness of the proposed mathematical optimization having FRP constraints by comparing its performance with that of ad-hoc implementation of FRP under a variety of experimental scenarios when conducting aggregate planning. In specific, we compare the production plans in a rolling horizon environment by evaluating the total costs and plan stability over the evaluation horizon. Then we extend our research to a mathematical optimization model that simultaneously optimizes conflicting objectives. Although FRP has been discussed in aggregate planning problems without optimization, none of the existing studies analyze the tradeoff between cost and plan stability under the presence of FRP. We fill these gaps by developing a bi-objective mixed-integer linear programming model using a compromise programming approach. Finally, we utilize these mathematical optimization models to demonstrate how stability in planning can facilitate leanness in system operations.The numerical results show that aggregate planning with FRP can consistently identify stable production plans without significantly sacrificing the cost objective. Flexibility bounds increase the responsiveness to demand fluctuations, provide manufacturers and suppliers a better visibility in forecasting, and have a smoothing effect on production and inventory levels. Overall, this dissertation research aims to contribute to the production planning area by introducing new optimization models to mitigate nervousness and help practitioners and researchers to build optimal and responsive planning systems by creating balanced trade-offs between those conflicting planning objectives.