Organizing supply timing variability

From a standard order, you deal with lead time uncertainty. This can mean timing for late deliveries.

Safety stock is often required to hedge against uncertainty in delivery and manufacturing lead times. Lead time variability (standard deviation around the mean):

• Captures variations in lead times due to natural, external forces

• Does not include the effects of expediting and advance order placements (we do not plan to expedite)

• Does not include the effects of upstream material unavailability (that is a different measurement)

Inserting a number in the LT Standard Deviation field means that SAP takes into account LT variability in the calculation of safety stock. SAP then assumes a shipment may arrive late, about 50% of the time. The LT StdDev shows the system how late the shipment may arrive and the probability of lateness. The higher the LT StdDev value, the later the shipment arrives.

This potential late shipment is then combined with other uncertainties in the supply chain and this figure is used to calculate safety stock. LT uncertainty also adds to potential exposure in the supply chain.

This field accepts non-zero values and they do not need to be integers.

Forecast error variability captures errors around forecasts.

Service variability captures intentional, less-than-perfect service from internal replenishment points, due to multi-stage optimization.

Lead time variability captures variations in lead times due to natural external forces. Lead time does not include the effects of expediting and advance order placements. Lead time does not include the effects of upstream material unavailability.

Total lead time is the time between when an order is placed and when it is received.

Actual lead time varies from mean lead time due to delays or faster processing.

Lead time variability creates the risk that order receipts may be delayed.

Order delays can result in a shortfall in inventory position.

Ignoring lead time variability negatively impacts service level.

The basic approaches are as follows:

• Plan for average lead time.
• Plan for maximum lead time (the worst case scenario).

The typical trends and tendencies for both approaches can be derived from the table shown in the following figure.

In contrast to SAP IBP for inventory, pure academic formulas simply ignore many of the important complexities in planning.

Based on the following figure, let’s look at an example:

• Does lead time variability at IN affect inventory shortfall at CF?
• Is this impact affected by the inventory decision (SS) at IN?

Example: Is CF shortfall the same if SS(IN) = 0 vs. SS(IN) = 1000?

Is the shortfall at IN and/or CF the same in every period if demand is time varying?

Let’s look at the following figure and compare the three different approaches:

• Plan for average lead time
• Plan for worst case lead time
• The SAP IBP for inventory approach

Two consecutive orders arrive. Each has an LT = 2. The orders are covered by receipts using the First Come, First Served (FCFS) principle.

The first order is delayed, but the second order arrives earlier.

The lead time of the first order = 3 and the lead time of the second order = 1.

Orders are not received in the same sequence and this is termed order crossover.

There are multiple scenarios for the values of the actual lead time of the two orders.

Think about which have order crossovers and which do not.

Principally, order crossovers can occur:

• When orders are not processed, based on a First Come, First Served (FCFS) policy.
• When there are multiple transportation modes or providers.
• When orders are processed in parallel.
• When there are multiple suppliers.
• When there is variability in transportation time due to factors like geography and weather.

Ignoring order crossover causes significant overestimation of inventory shortfall and hence safety stock requirements.

This can also be seen from the total statistics in the order crossover table.

Let’s simulate the impact of a rising PBR for a crossover scenario, first we assume PBR = 1.

Next, the PBR increases to 2, all other parameters remain unchanged.

With the PBR being increased to 3, the tendency becomes clear for our example. The higher the interval between orders, the lower the chance for order crossovers.

As the figure shows, large batch sizes increase the interval between orders. This reduces the chance for order crossovers as before.

With regard to the batch size, we make the following observations:

• The average on-hand inventory increases with batch size in both scenarios.
• The on-hand inventory can be lower when the impact of order crossover is considered.

For our example, these tendencies can be derived in numbers from the table in the following figure.

Safety stock always increases with lead time variability, as shown in the following figure.

When lead time standard deviation is very small relative to lead time:

• Order crossover does not occur
• Safety stock decreases as batch size increases

When lead time variability is high:

• Order crossover occurs when batch size is relatively small.
• Safety stock increases asymptotically as batch size increases. For example, it reaches a limit.