At the base of the SC decision pyramid, which corresponds to SC operations, decisions are made on a daily or weekly basis, while, half way up the pyramid, SC managers adopt an appropriate inventory policy and network flow planning over a multi-period planning horizon. However, at the peak of this pyramid, strategic long-term design decisions (e.g., those related to facility location and customer allocation) are made for the whole period in the SC network.
When the SC network is designed, tactical decisions are made at the planning level (e.g., inventory policy) on a monthly to yearly basis, and the relevant daily operational decisions are implemented accordingly. In our recent research study, we formulate a new model aimed at integrating the inventory planning decisions and the location-allocation design decisions, and at analysing the impact of inventory optimization in a multi-period and multi-echelon SC network design.
To test our model, we consider networks scattered over the American Northeast and over a 12-month planning period.
Below are listed some important managerial insights that we draw from our study:
- The impact of the uncertain market demand on SC network design and planning :
SC managers should consider a dynamic demand forecasting approach when designing a SC network.
- When market demand is fairly constant or shows a minor fluctuation, a similar inventory policy may be adopted at the same DCs; conversely, when market demand shows some significant variations or follows a seasonal pattern, a specific inventory policy should be implemented at each DC in order to avoid overstocking or stock-out of items.
- The results of our model indicate that, when demand is constant or shows less variation over time, sourcing from fewer suppliers is profitable; conversely, sourcing from multiple suppliers is recommended when demand shows some high variations or follows a seasonal trend.
- The more uncertain the market demand, the higher the SC planning costs, including those related to inventory holding, shortage, order flows.
- The risk of running out of stock is high when demand shows some high variations or follows a seasonal pattern. Hence, to avoid shortage penalty costs due to customer unsatisfactory situations, more DCs may be opened, and located closer to customers. This measure is highly recommended when product holding costs are high.
- The impact of inventory control on SC design and planning
SC managers should take into account the inventory control and management when designing a SC network.
- If inventory costs are high, decentralized DCs and multiple-sourcing can be a good strategy to keep low inventory levels at DCs, with more product flows in transit, although this may slightly increase total transportation costs. When inventory costs are low, inventory centralization and single sourcing are suitable approaches.
- When inventory costs rise, there should be more suppliers, items should be stocked at closer regional DCs.
- The inventory policy should be well-defined at the strategic level using demand forecasting techniques in order to capture the maximum market share and to reduce the risk of customer unsatisfactory.
- When demand uncertainty is high, the (s, S) inventory control policy, also called reorder-point and order-up-to level system, makes DCs more responsive to meet the required throughput capacity.
As a conclusion, we argue that the inclusion of inventory planning and policy that consider demand forecastings at the strategic level can improve the quality of design decisions. The key findings of our study are highlighted in the original paper.
To read the original paper, click here.