Buffer Sizing, Profiles and Dynamic Adjustments

Hello, this is Ken Titmuss again with the third in the DDMRP Podcasts. Today we will be taking a brief look at the second and third components of DDMRP, namely DDMRP Buffers and Buffer Profiles as well Dynamic Buffer Adjustment. I say a brief look as these podcasts can’t really go into the depth of understanding required, but hopefully inspire your curiosity to do further research and maybe take the Demand Driven Institutes Demand Driven Planner course. If you have been looking at the DDI website you will see that there is also a Demand Driven Planner Professional certification exam that you can take, there are plenty of details on the website on how to go about this and what is required.

So, recapping, we have looked at why we need to seriously look at changing the way we have traditionally been planning our businesses and what becoming Demand Driven really means. Last time we went through the first component of DDMRP, Strategic Inventory Positioning and now we will continue with the second component, Buffer Sizing and Buffer Profiles.

In this section we will look at when inventory is regarded as being an asset or a liability, list some of the factors to be considered when constructing DDMRP buffers as well as understanding the logic and purpose of the three difference zones in the buffer.

Firstly, when do we consider an item of inventory to be an asset or a liability? Well, if we have very little inventory or we are out-of-stock, or worse still we have back orders, this is a liability. We miss sales, short deliver or expedite deliveries to replenish the stock. These all cost us money and hence are a liability. At the other end of the scale if we really have too much inventory for our immediate needs then we have spent money on something that is not giving a return in the short term and hence is a liability. Not only that, stock could become obsolete or expire and will have to be destroyed, also a liability.

So, between these too high and too low points, inventory is regarded as an asset. However, as we progress to the zero point or point of too much the return on that investment will diminish. Therefore, at a mid-point between zero and too much we will get the greatest return on our inventory investment.

In practice when we analyse the complete stock in a warehouse, we typically find that too many of the SKU’s have low inventory or are out-of-stock and an equal amount is at a level where there is too much. Generally, we find only a small number of SKU’s sitting in the ideal position for a maximum return on the inventory investment. We also find that, using traditional MRP systems, an item can quickly move from having too much to too little with a requirement to immediately replenish the item due to it violating safety stock levels and hence putting it back into the too much category. So, we end up with what we call Bi-Modal inventory distribution. We need to get to a point where the majority of our SKU’s are sitting in the optimal range for maximum return on investment and those out on the tails being investigated for the reason why and changing their operating model parameters to bring them into line using a process of continuous improvement.

In order for this to happen, we need to create stock buffers at our strategic inventory positions. These stock buffers consist of three zones: Red, Yellow and Green. Each of these zones have a specific purpose in the buffer. Let’s take them one at a time.

The yellow zone is the area in the middle of the buffer. We usually calculate this one first as it is the easiest and is the heart of the demand coverage in the buffer. It is always calculated using the lead time and multiply it by the Average Daily Usage, called ADU. Maybe at this point we need to address the issue of which lead time to use and how do we determine the ADU.

Lead time, in days, is the total length of time it takes for inventory to be replenished in the buffer. For the buffer of a purchased raw material or component it will be the purchasing lead time. If it is a manufactured item, then we would use the Decoupled Lead Time or DLT mentioned in the previous podcast. For an SKU at a distribution centre it will be the transit time to replenish the stock from the supplying location.

With regards to ADU, this can be calculated in three different ways. For most items in a location we would probably use a backward looking ADU. Here we might take the last 90 days of demand and work out the ADU by dividing the total demand over this period by the number of days. For a new item that has no past demand data we would probably use a forward looking ADU, known as an FDU, taking the total of say 90 days of forecast and dividing by the number of days, again. There is another option of a Blended ADU, in this case we take some past data as well as some forecasts to come up with a number. What is best for each SKU at a location will need to be determined by the project implementation team, maybe using some simulation tools.

The top zone of the buffer is the Green zone, which is the heart of the order generation aspect of the buffer. It determines the frequency of reordering and the typical size of the orders. We calculate the size of this zone in three possible ways and typically select the number which is the largest.

The first method of calculating the Green Zone is to see if there is a significant Minimum Order Quantity, or MOQ, when this item is manufacture or purchased. Secondly, we determine if there is a cycle to purchasing or manufacturing the item, so, for example we might want to make an item in the factory every two weeks, or we might want to buy products from a supplier once a month. We then take that cycle in days and multiple it by the ADU and we end up with a number. The third method is to take the size of the yellow zone and multiple it by a Lead Time Factor, or LTF, which we will define shortly. This method again will give us a number. We compare these three numbers and generally take the largest as the size of the Green Zone.

The Lead Time Factor, mentioned previously, is determined by the length of the lead time to replenish the item of inventory. For short Lead Time items, we would use a factor between 0.6 to 1.0, for medium Lead Time SKU’s a factor between 0.4 and 0.6 and for long Lead Time items a factor between 0.2 to 0.4. This is not intuitive as we use a small factor for long lead times and a larger factor for short lead times. The reason is that for long lead time items we want to generate a smaller Green Zone and hence more frequent orders to promote flow, reduce risk in the supply chain and give us better cash flow. The factor you use for your SKU’s again will be determined by you project implementation team.

So, now we have calculated the size of the Yellow and Green Zones, we now need to determine the size of the Red Zone. The Red Zone is the safety in the buffer, but it is not safety stock and doesn’t behave in the same way as safety stock. It is really the ‘shock absorber’ for the demand or supply variability at this point in the supply chain. The Red Zone is broken down into 2 parts; the Red Zone Base and the Red Zone Safety. These parts are calculated separately and then added together to determine the total size of the Red Zone.

The Red Zone Base is calculated as the Lead Time, times the Average Daily Usage, times the Lead Time Factor. In other words, it is the same as the Green Zone if we used the Lead Times Factor option. The Red Zone Safety is calculated by taking the Red Zone Base and multiplying it by the Variability Factor and adding the two portions of the Red Zone together to get a total. The Variability Factor, as opposed to the Lead Time Factor, is intuitive. With high variability we use a high factor of between 0.6 to 1.0, medium variability 0.4 to 0.6 and for low variability 0.2 to 0.4. The variability of each SKU needs to be determined by the implementation team, but one can use a Coefficient of Variability factor to assist in the process. The Coefficient of Variability, or CoV, can be calculated by dividing the SKU’s Demand Standard Deviation by the ADU.

So, now we have the size of the three zones in the DDMRP buffer. In DDMRP ‘speak’ we talk about Top of Red, which is the size of the total Red Zone. Top of Yellow which is the Red Zone plus the size of the Yellow Zone. Top of Green is the Top of Yellow plus the size of the Green Zone. Over this number we talk about Over Top of Green, or OTOG for short.

To make things easier for the system to complete the calculations we need to set up Buffer Profiles to which we will assign each and every SKU. The Buffer Profiles matrix is initial sectioned into major categories of inventory: Purchased, Manufactured, Intermediate/sub-Assemblies, Finished Products and Distributed Items. Within each of these inventory categories we divide the matrix into Long, Medium or Short Lead Times. These Lead Times are then divided into High, Medium or Low variability Categories. This gives us about 45 buffer profiles to which we allocate all our SKU’s, the system then knows how to perform the buffer sizing calculations for each item. So, for example if we have a manufactured item (M) with a short lead time (S) and medium variability (M) it might be allocated the buffer profile code of MSM. In the two-day DDI Demand Driven Planner course we go through a number of exercises to ensure participants are proficient in calculating buffers.

Now we have the buffer profiles and buffer sizing organise, we need to Dynamically Adjust the buffers to take account of constant changes in the external trading and business environment. So, now we will look at the third component of DDMRP, Dynamic Adjustment. Under Dynamic Adjustment there are two issues we need to discuss: Recalculated Adjustments and secondly, Planned Adjustment Factors.

Let’s first discuss Recalculated Adjustments. These adjustments to the buffers are done automatically by the system when any of the key parameters in the part change such as Lead Time, Average Daily Usage, Minimum Order Quantity, Buffer Profile, Lead Time and Variability Factors. So, buffers are keeping up with changes to your business environment virtually in real-time.

Over and above these automatic adjustments to the buffers, we have the ability to apply manual Planned Adjustment Factors for various circumstances. Here we basically have three types of adjustment: Demand Adjustment Factors, Zone Adjustment Factors and Lead Time Adjustment Factors.

Demand Adjustment factors are used mainly for SKU’s that have strong seasonality. In these cases, the Average Daily Usage is multiplied by a seasonal index to increase the size of the buffer into the future to take account of greater sales in the high season. These seasonal indexes are then reduced after the high season to bring the buffers back to the normal low season size.

Demand Adjustment Factors can be used to ramp up demand for a new product or ramped down for products at end of life, or a combination of both for a product transition.

Zone Adjustment Factors can be applied to each of the three zone individually for various reasons. For example, we could adjust the Green Zone to manipulate order sizes and frequency of ordering. The Yellow Zone could be adjusted because of a temporary change in lead time, for various reasons, but not warranting a change in the parts basic planning data. The Red Zone could be adjusted due to a temporary change in variability which could be due to strikes at ports or transportation, or in manufacturing plants, not warranting a permeant changes in buffer profiles.

The Lead Time adjustment factor could be used for single or groups of items where lead times increase, or decrease, due to the season or event. For example, lead times tend to extend during Chinese New Year so for SKU’s coming from this region the buffers should be expanded during this period. The same thing can happen when it is summers holidays in Europe during July and August, business tends to slow down somewhat.

The DDMRP implementation team will need to determine how and when to use these adjustment factors which will vary with different businesses and products within the company.

We have now completed the first three components of DDMRP: One, Strategical Inventory Positioning. Two, Buffer Sizing and Buffer Profiles and Three, Dynamic Buffer Adjustment. In the next podcast we will go through, what I think in the most elegant portion of DDMRP, Demand Driven Planning.

Up to this point we have basically covered the first day of the DDI Demand Driven Planner course. Bearing in mind that these first three podcasts have taken about 40 mins, one can appreciate that there is a lot more to learn and know on the subject. It is therefore important that if you want to start a Demand Driven journey you need to attend the Demand Driven Planner course and ultimately take the Demand Driven Planner Professional certification exam. More information on this course and other Demand Driven education offerings are available on the DDI website: www.demanddriveninstitute.com. If you sign up on this website, it is free, you will get extra Demand Driven material on the OnDemand Tab and don’t forget to look at the case studies and presentations which are available from recent Demand Driven World conferences.  You will find a list of global DDI affiliates on the website. Contact one today in your area and see what they have to offer. I am Ken Titmuss and you can get me on ktitmuss@mweb.co.za. Until next time, goodbye.