Due to recent focus on resilience many supply chain leaders are using inventory in order to mitigate the risk of undesired disruptions in their supply chains. Although this is one of the more reliable techniques to improve reliability of deliveries and improve customer experience, however it should be weighed against other less expensive measures which might provide equally or even better results at reduced cost.
We all know that the more inventory we keep downstream the more expensive it is. On the other hand, keeping inventory at the earlier stages of manufacturing can result in long leadtimes and therefore potential loss of revenue. What is the “Goldilocks” amount of inventory to be kept at each stage of the supply chain to satisfy both cost and delivery performance? But more importantly, is there a need to keep inventory at every stage as long as we have “enough” of raw material and purchased items to build what we need in a timely manner? For the latter to work, one needs to keep inventory in the form of capacity. Capacity and inventory are really the two sides of the same coin except that capacity is perishable inventory. We use both capacity and inventory to respond. Both are needed. There is no point to have too much of one and too little of the other. Capacity is essentially virtual inventory to be used when needed.
Lets say you plan only 70% of your available capacity and keep 30% to respond to urgent orders and disruptions. This would be a lot less expensive than building ahead to keep the resources fully utilized and keeping the goods in FG inventory. Or not having enough capacity to respond when a disruption occurs. You may ask keeping 30% of capacity idle may result in losing current orders and increase the cost due to low utilization. That can certainly be an issue but it can be resolved using a combination of analytical and prescriptive techniques such as Multi Echelon Inventory Optimization (MEIO) and ML algorithms.
Machine learning algorithms can be used in order to decide how much capacity would be best to set aside for a rainy day. In addition, MEIO can accurately prescribe what levels of inventory should be kept at each stage under different scenarios. For example, a disruption of supply for a week, a month or even longer can be examined to see where the bottlenecks occur regarding availability of inventory and capacity. Furthermore, ML can be deployed to determine what level of safety stock is needed based on capacity availability as well as many other factors that might influence the amount of safety stock or hedging inventory.
Factors that can influence the amount of real or virtual inventory (i.e. capacity) needed, are supplier reliability and availability of alternatives, manufacturing leadtimes, variability of demand and supply and resources availability. All such factors as well as the amount of risk that the organization is prepared to be exposed to, in terms of additional cost or loss of revenue, are taken into account in calculating the amount of inventory and capacity to be made available to mitigate the impact of disruptions. More specifically, the strategy should be having a platform that merges planning and execution such that the long range release plan would be created utilizing the lower capacity levels, then open the capacity for the execution and scheduling to allow orders to pull forward to consume based on the current state. This would avoid the “cost” of underutilization of resources but may add to FG inventory. The balance has to do with your objectives and level of admissible risk.
Whatever policies are deployed are subject to change since all the aforementioned factors are changing all the time. The key is to have analytical tools that can monitor such changes and be intelligent enough to recommend decisions as the underlying parameters are changing. For more information on supply chain resiliency and inventory optimization and how to implement the recommended approach, click Here.