page contents

Forecasting Cores and Spares

Posted Sunday, February 1, 2015 BY: John Barry

Filed in: Forecasting, Supply Chain Management

Tagged with: Tags: , , ,

Forecasting cores and spares present special software challenges, and may require creative thought to providing a workable approach. There is no single silver bullet process with a one size fits all methodology to address this issue. Solutions are as different as corporations, management, industries, and product.

 

Cores

 

Cores, or rebuilds require special planning to ensure replaceable components are readily available for repair, and the remanufactured item can be issued back into the field promptly. Let’s illustrate the challenge by examining an automotive alternator. If you get stranded away from home and a determination is made that you have a faulty alternator, then you may be able to get an OEM (Original Equipment Manufacturer) replacement or a rebuild. Depending on make, model, and year of your vehicle, an OEM might cost $300, and a rebuild $200, and you get a $10 to $30 credit for the core, or old faulty item.

 

If you are in the business of manufacturing and remanufacturing alternators, what are the challenges? Consider the number of automotive manufacturers, the types of vehicles, engine size, and model year. Naturally alternators can be used across a number of makes and models. In planning new production we use history and project future demand taking trend and seasonality into account. Since product is always being improved, we must keep track of sequential supersessions as the old item gets replaced with a new improved and reengineered product. It is critical to issue new part numbers with every change of form, fit, or function. Generally manufacturers receive releases from the automotive companies, and they keep you informed of requirements for new model changes and end of model runs.

 

Rebuilt core production is more challenging to plan. When a damaged core is returned, you strip, test, and inspect the device, determining which components needs to be replaced. It could be the roller bearings (large and smaller in some alternators), rotor assembly (or armature), stator, regulator, or carbon brushes and rectifier combination. It is less likely that you need to replace the pulley, drive-end shield, collector-ring end shield, or protective cap unless the vehicle was involved in a serious motor accident. Sensibly you may decide in every situation to replace the bearings as you would not want this to fail in a rebuild. That begs the question—how do you plan for replacement components?

 

Using historical information is a great help. By reviewing previous completed work orders you will have a list by date of what components went into each build for each end item part number. In effect you create a planning Bill of Material (BOM) for a generic alternator showing the percentage of each component that was replaced. Bearings might be 100%, while armatures only 35%. You now forecast end items, based on history, to determine the timing and number of cores you are likely to receive in the future. This information is used to explode the planning BOM to determine how many components you need to have on hand in each period. Your replenishment application will net requirements based on on-hand inventory to help plan new components. This planned order information is helpful in planning purchase material requirements, capacity and labor. However, it is not necessarily that straight forward. History may show an increasing trend as the model year of the automobile ages with a greater likelihood of needing a replacement. This demand could be further increased as we go through economic downturns when motorists keep their vehicles on the road for longer periods. Extrinsic factors have an important bearing on future demand and appropriate adjustments should be made. This information should be factored into future demand if looking to increase customer service levels and be responsive to need, while minimizing the inventory investment.

 

To return to the concept of an OEM supplier and remanufacturer of cores, the component demand information should be aggregated. With the OEM product, knowing the quantity and timing of automotive demand, explode the BOM to determine component requirements. Add to this the demand of components from the projected core requirements. If effectively planned you will not run the risk of robbing components for the OEM build, or for the core rebuild, but will have sufficient materials for both demand streams.

 

The core rebuild statistics should be a boon for engineering. It alerts them to when components fail as they track Mean Time Between Failure (MTBF). This may call for a redesign to further assure the public of a high quality product, and solidify the company’s reputation for building quality products. In some industries, where additional information is available, this could help pinpoint causes. If we only knew the mileage of each vehicle when the product failed this would provide an additional insight to the life of the product. Generally the only available information to use in calculations is a statistically aggregated annual national average mileage driven. With aircraft we know the count of landings and flying hours to help plan maintenance for say the landing gear.

 

Service Parts

 

A spare part, or service part, or repair part, or replacement part, not tied to a core or remanufactured item, is another challenging field for planning. Here we are looking to replace a failing part. Included in this category are consumables. Generally the expectation is that the part will not fail, and when it does could be very costly. My brother-in-law operates a highly automated sorter separating ingredients by color. The machine has a number of $2.00 plastic washers that if broken causes the entire line to stop operating. It goes without saying that he carries more than a year’s worth of spares for this specialized imported machine and its imported service component.

 

In discussing the planning of spares, I’ll use one example. Washers. Washers come in an almost infinite variety of outer and inner diameters, thicknesses and material. Aluminum, brass, copper, plastic, rubber, Teflon, steel, stainless steel, etc. Beveled, flat, round, and square. Then too we have similar products such as shims, seals, gaskets, spacers, and intricate shapes. Generally the outer diameter is twice the inner diameter. You may need to cater for metric and imperial sizes. Washers serve many industries including electrical, agricultural, and automotive.

 

If we review one stock keeping unit (SKU) we might find in looking at its history that we sold one or two infrequently over the past year. If we decided to forecast this item, it is indisputable that the forecasting engine would select Croston’s method, add aggregate demand over the prior year, divide by 12 if we are forecasting monthly, or by 52 if we are planning weekly, and draw an average straight line demand looking forward. In that case you might be wise to add in a level of safety stock if you can afford the inventory investment, establishing a reasonable customer service level factor in your calculations. With a detailed SKU you can expect a high degree of noise as you attempt to predict the future. What alternate strategy might we use?

 

We could group washers into a family of products. Let’s decide that one family may be Flat Washers made of stainless steel suitable for desks, chairs and computers. This could be the family of small washers. Another family of larger stainless steel washers to be used for vehicles, buildings and spacecraft. By aggregating a family of products consisting of a variety of sizes we should find the historical demand is sufficient to help generate a meaningful forecast providing a representative trend and seasonality. However, we do not manufacture a family of products, but individual SKUs. Examine the makeup of the SKUs within the aggregate family and calculate the percentage contribution of each SKU to the aggregate. With a believable forecast at the aggregate level, ideal for adjustments using extrinsic or econometric modeling information, we explode or prorate from the aggregate to the detailed SKU level using the planning BOM concept knowing the percentage contribution of each SKU to the total. I clearly understand that you may find fractional results at the lower level that you should round up to the nearest whole number, providing an added margin of safety stock.

 

Factory Planning or APS (Advanced Planning and Scheduling)

 

At the end of the forecast and replenishment cycle you have a picture of the inventory investment and helpful information to drive into your factory planning system to verify that you have sufficient material, labor and available capacity. Sharing this information with suppliers helps them to be in lock step with your short, medium and long term plans. Then too, where practical, you request your customers to validate your projected demand for their portion of the product range.

 

ABC 80/20 Pareto Analysis, and XYZ Frequency of Demand

 

At the outset in discussing spare parts, I deliberately made the assumption that demand was infrequent. This need not be the case. If you have high levels of demand as determined through a Pareto Analysis (80/20 ABC rule), then “conventional” methods of planning will assist in determining future demand. In addition where there is a high level of frequent demand as determined by the XYZ analysis, with “A items” and frequent orders, or “X items” there will be little need to have a safety stock buffer helping to free up cash, unnecessarily tied up in inventory. Planning these items is relatively simple, it is the remainder, the 80% that might present the greatest opportunity to creatively gain control. You may even be more fortunate and have “A” type customers that account for a significant portion of your demand. Here too you can request demand information from them using a collaborative planning approach.

 

Goal

 

Our goal in working with clients is to ensure maximum profitability through the most productive use of material, labor, and capacity, increased on-time delivery to customers with the highest inventory turns possible.

Comments are closed.