Summary:  In a recent blog on Inventory Forecasting the core challenges and business importance of estimating inventory are outlined. A projected inventory position across time (plan) is a natural co-product of most central or master planning models that match assets with the demand to create a projected supply line linked to demand.   The use of optimization methods in this matching process can be critical to ensure the most effective use of inventory balancing minimizing inventory with on-time delivery.  This is especially true where there are alternative methods to produce a product and the different methods consume different materials.

In a recent blog “What is Inventory Forecasting & Why Your Business Needs It?”, Arkieva COO Sujit Singh covers the core challenges and the business importance of estimating the anticipated future inventory positions (forecast) of the firm.  Typically a firm is focused on three types of inventory: exit or product (sold to customers), intermediate products (products produced to be used to produce final products) sometimes called Work in Progress (WIP), and raw material (typically component parts that are purchased or acquired to use in production).  For example, in producing cheese the raw material is milk, intermediate products are cheeses produced in VATS at different ages, and the final product is then packaged cheeses at the right age.  In shirt production the raw material might be cloth in rolls and dye; intermediate products dyed material and partially created shirts, and the exit product is the shirts (defined by type, color, and size).  My IBM team referred to these as top dwellers, purgatory dwellers, and bottom dwellers.  Sujit noted the “the inventory forecast is dependent on the supply and demand and their delta.”

How does Central Planning (sometimes referred to as Master Planning or Supply Planning) fit with inventory forecasting?  The purpose of central planning is the control point for the flow of material or product within an organization and focuses on:

  1. How to best meet prioritized demand which comes from the demand management application
  2. Without violating temporal, asset (WIP and inventory), or capacity constraints and physical requirements (for example, the shelf life of a product)
  3. While doing its best to meet business guidelines and preferences, one of which is minimizing inventory in all three groups

This plan is generated by the central planning engine (CPE) or model which matches assets with demand across time.  Although the two outputs that receive the most attention are:

  • Projected supply line linked to demand
  • Capacity Utilization

Other important plans generated include:

  • Projected inventory position across time for all (top, purgatory, and bottom) products or parts

What is the role of optimization? Most demand-supply networks contain alternative action options where a decision must be made.  Examples:

  1. Shared component material: the cloth is shared between various shirts; mozzarella VAT cheese consumed by different blends.
  2. Shared capacity: the cutter is shared with the production action to make different size shirts; the VAT can produce whole milk or skim mozzarella.
  3. Alternative production paths or build options (BLDOPT). BLDOPT1 can produce all sizes of shirts, BLDOPT2 can only produce large and x-large shirts.  VATS 1 to 10 can produce whole milk and skim mozzarella.  VATS 11 to 13 can only produce skim mozzarella.
  4. Alternative production paths or BLDOPT that consume different component material to produce the same product or part. For example, we have two types of material to produce shirts.  Material 01 can produce all sizes of shirts.  Material 02 can only produce large and x-large shirts.  For this example, ignore color.  We have these BLDOPTS
    • BLDOPT01S – consumes material 01 produces small shirts
    • BLDOPT01M – consumes material 01 produces medium shirts
    • BLDOPT01L – consumes material 01 produces large shirts
    • BLDOPT01X – consumes material 01 produces x-large shirts
    • BLDOPT02L – consumes material 02 produces large shirts and slower than BLDOPT1L. There is also more waste.
    • BLDOPT02X – consumes material 02 produces x-large shirts and slower than BLDOPT1X. There is also more waste.

If we are not intelligent in deciding which BLDOPTs to use to produce which shirts then we may well generate unnecessary material inventory or may fail to meet the demand to keep material inventory low. If the total demand for shirts requires 10,000 units of material 01 and we currently have 6,000 units of material 01 and 6,000 units of material 02 are arriving in 4 days, then we need to balance on-time delivery with smart consumption of material inventory.  If we have demand for 1000 small shirts (requiring 4,000 units of material 01) and 900 x-large shirts (requiring 4500 units of material 01 or 5400 units of material 02) of equal importance, it would be best to delay the production of x-large shirts until material 02 arrives.

A detailed example of this analysis is provided in the second part of the blog Best Fit Plan. Yes, it is a bit complex, but complexity exits whether you ignore it or not, best not to ignore it. The question is, can a business really afford an underperforming plan simply, so it is easy to follow?

 
Enjoyed this post? Subscribe or follow Arkieva on Linkedin, Twitter, and Facebook for blog updates.