Optimizing Demand Re-Allocation under Fixed Capacity Commitments

Abstract

The automotive industry is suffering from large supply-chain disruptions and restrictions, and sourcing contracts are signed during concept and design phases, fixing capacity commitments years ahead of time based on highly inaccurate demand fore- casts. Companies struggle with demand variation and the inability of the supply-chain to react, with suppliers having limited ability to respond to updated demand signals and varying market conditions. This thesis proposes an algorithmic solution to optimize allocation of demand order volumes into different product types when supply capacity is constrained, while trying to maintain the originally requested mix shares and volumes as much as possible.

The model allows the company to approach supply-chain flexibility under a different lens and the results demonstrate the volume and financial improvements to current processes, with analyzed used cases showing an increased volume output of 10% and profit increases of up to 11% over a manual re-allocation process, or 7% increases when constraining the model to a lower volume output. Finally, a few additional model improvements are suggested as grounds for future work.