The MRP Conflict Today

Does your company work within its formal planning system or does your company work around this system? Does it try to do both at the same time? Are spreadsheets, sticky notes, and manual tracking systems still alive and well in your operations even though you have implemented an MRP or ERP system in the last 10 years?

When it comes to truly effective materials management, most Purchasing, Manufacturing, and Production Control personnel frequently feel like their hands are bound and tied. MRP's power has always been its ability to manage BOM connections in order to generate total net material requirements (demand orders that turn into manufacturing orders or purchase orders). The more complex and integrated the product structures, manufacturing facilities, and supply chains are, the more necessary MRP is for netting and getting ahead of critical and long lead time parts. Most Purchasing, Manufacturing, and Production Control personnel realize this and are forced into a set of unsatisfactory compromises that just don't work. The next section discusses the compromises that arise from this MRP conflict.

The MRP Compromises

In most cases, there are five types of compromises that frequently occur (either separately or in combination).

1. Manual Work Around Proliferation-As has been discussed already, companies frequently try to work around their MRP system by relying on stand-alone, disconnected, and highly customized data manipulation tools like Excel spreadsheets and Access programs. These tools have serious limitations and their proliferation makes the IT landscape more complicated and maintenance more intensive. Their use ultimately defeats the purpose behind the major investment in an integrated ERP package.

2. Flatten the BOM-Sometimes companies try to simplify the synchronization issue by flattening the BOM. Flattening the BOM removes levels that were originally identified to define the product and the process. The key to better synchronization is not to ignore dependencies within the product structure and across product structures. Better synchronization is possible when you know on which dependencies to focus. When the BOM is flattened, it is imperative that only those BOMs are flattened that cannot provide a leverage point. Flattening BOMs across the board can eliminate key leverage points that can provide a great deal of value. These dependencies provide an excellent way to stop variability from gaining momentum and disrupting the entire supply chain like a tsunami wave. The key to better synchronization is to understand those dependencies and control them. By flattening the BOM, companies can actually lose visibility at both the planning and the execution levels. In some cases, companies can benefit themselves by inserting an additional level in the BOM!

3. Make-to-Order Everything-Still other companies choose to place all of their cash in raw material and purchased components and embrace a completely make-to-order (MTO) strategy. In most environments, this comes with a significant price. A company either has to carry additional capacity to meet service level requirements or risk service level satisfaction with extended lead times. In some highly seasonal or short customer tolerance environments, this is simply impossible. The company just cannot supply the product in sufficient time with sufficient volume.

4. More Efficient Forecasting-Other companies implement advanced forecasting algorithms or hire more planners in hopes of guessing better. Recall that the assumption under MRP is that there is a plan or forecast that is the demand in the system to drive the MRP calculation. Even with dramatic improvements in forecasting accuracy, the results do not translate to the bottom line. Experience has shown that at best these solutions result in a 20 to 40 percent improvement in demand signal accuracy-still leaving significant room for error. Even if a company succeeds in increasing signal accuracy, it does not necessarily translate well to overall effectiveness in terms of availability and fill rates. Remember, the increase of variability and volatility (especially on the supply side) can easily offset any appreciable gain in signal accuracy. Also, remember that many manufacturers can have multiple assembly and subassembly operations that are integral parts of their overall flow. In any type of assembly operation, it takes the lack of only one part to block a complete shipment. The more assemblies there are, the more complex the synchronization and execution challenge is. Finally, even the biggest supporters of forecasting cannot argue the fact that forecasting in any form is still a push-based tactic. Yes, it can be a more educated push but it is still a push nonetheless. For companies implementing pull-based manufacturing systems (e.g., Lean or DBR), this sets up conflicting modes of operation that will simply not perform well in volatile and complex environments.

5. Manual Reorder Point Systems-With the implementation of kanbans, supermarkets and three-bin systems manufacturing have come full circle. Unable to overcome the shortcomings associated with MRP, some companies have abandoned it completely. It is essentially throwing the proverbial baby out with the bath water. In many environments, it is devastating. These systems tend to be manually intensive and very difficult to make responsive to changes in the environment. There is almost no ability to see either the truly available stock or the total net requirements picture (all demand allocations in relation to all open supply orders). Real data is masked in traditional systems by requirements coming from forecasts or other false demand signals. In fact, by definition, each parentchild relationship in the BOM is managed independent of any other connection. MRP consolidates the total requirements for each child part and only rarely can even an experienced planner understand why that quantity is being ordered. In environments with high variety and options, it often requires massive amounts of inventory on the floor in order to be able to provide components and parts when necessary. A 2007 AMR Report (Masson et al., 2007, 6) came to two important conclusions. First that, "Kanban cards and heijunka boards become unmanageable when there are hundreds or thousands of products and components." Second, and most interesting, is that in large global manufacturers with many manufacturing sites and lines, "The pragmatist needs software to support lean manufacturing." Remember that simply knowing the stock on hand cannot provide the information to know what to order unless the on-hand position plus the open supply orders minus the demand allocations is considered (this is called an available stock equation). This is just not possible with manual reorder point systems like kanbans.

Actively Synchronized Replenishment-the Way Out of MRP Compromises

For those who are familiar with constraints management and its thinking processes, the dilemma that manufacturing companies find themselves in can be seen in the conflict cloud in Fig. 12-2.

There are essentially two critical needs (B: Produce to demand and C: Visibility to total requirements) coming into contention behind the compromises (made at D and D', the pull or push choices). From a manufacturing perspective, we must have a realistic way to respond and produce to demand. This way must include both capacity and materials. MRP tools simply do not create the correct "demand signals," nor do they facilitate materials availability within increasingly shorter horizons that are inherently more variable and volatile. Additionally, many pull-based manufacturing implementations (e.g., Lean and DBR) are effectively blocked by this lack of material synchronization. In most cases, due to the shortcomings listed previously, this leads many manufacturing personnel within companies to think that they should ignore MRP. In fact, a frequent milestone for a Lean implementation is that the computer planning system has been eliminated!

FIGURE 12-2 The conflict in utilizing MRP.

On the other hand, from a Planning and Purchasing perspective we must have a way to effectively see, plan, synchronize, and manage the availability of all materials, components, and end items, especially critical and long lead time manufactured and purchased parts. With increasingly complex planning scenarios, it leads Planning personnel to insist on utilizing MRP.

The more complex the manufacturing environment, the more acute this conflict tends to be. The inability to effectively reconcile the dilemma in those environments leads to the ineffective MRP compromises listed previously and can also essentially relegate TOC, Lean, and Six Sigma implementations to lip service. The requirements must be achieved without the conventional inaccuracy, inconsistency, and massive additional efforts and waste associated with the current set of compromises. MRP, as previously noted, has some very valuable core attributes in today's more complex planning and supply scenarios (BOM visibility, netting capability, and maintenance of sales order/work order connection between demand allocations and open supply). The key is to keep those attributes but eliminate MRP's critical shortcomings (listed previously) and use the pull-based replenishment tactics and visibility behind TOC and Lean concepts all in one system in a dynamic and highly visible format.

ASR builds upon the traditional replenishment approach of replacing what was taken or used to create a dynamic and effective pull-based solution to answer the challenges of today's manufacturing landscape. Through new approaches in inventory and product structure analysis, new pull-based demand planning rules, and integrated execution tactics, ASR is designed to directly tie material availability and supply to actual consumption throughout the BOMs, thus removing the "islands of MRP" obstacles that most supply chains face. Additionally, this approach is a prerequisite to effectively utilize pull-based scheduling and execution methods like Lean and DBR in more complex manufacturing environments. Additionally, ASR has a unique way to incorporate required elements of strategic planning with little or no exposure to the variability and volatility that gets companies in trouble with traditional forecasting techniques.

ASR has five main components.

1. Strategic Inventory Positioning 2. Dynamic Buffer Level Profiling and Maintenance 3. Dynamic Buffers 4. Pull-Based Demand Generation 5. Highly Visible and Collaborative Execution They are discussed in the next sections.

1. Strategic Inventory Positioning

The first question of effective inventory management is not, "How much inventory should we have?" The most fundamental question to ask in today's manufacturing environments is, "given our system and environment, where should we place inventory to have the best protection?" Think of inventory like a break wall to protect boats in a marina from the roughness of incoming waves. Out on the open ocean, the break walls have to be 50 to 100 feet tall, but in a small lake the break walls are only a couple of feet tall. In a glassy smooth pond, no break wall is necessary.

In the same way, inventory is the break wall against the variability experienced from either supply (externally and internally) or demand unreliability. Remember that a company has to think holistically across not only the enterprise but also across the supply chain. Putting inventory everywhere is an enormous waste of company resources. Eliminating inventory everywhere puts the company and supply chain at significant risk. Strategically positioning inventory ensures the company's ability to absorb expected variability without having to disrupt every part of the plant and the supply chain. Important factors to consider carefully in determining where to place inventory buffers include: Customer Tolerance Time-The time the typical customer is willing to wait or the potential for increased sales for lead time reductions.

Variable Rate of Demand-The potential for swings and spikes in demand that could overwhelm resources (capacity, material, cash, credit, etc.).

Variable Rate of Supply-The potential for and severity of disruptions in particular sources of supply or specific suppliers.

Inventory Flexibility and Product Structure-The places in the "aggregate BOM" structure that leave a company with the most available options (primarily key purchased materials and subassemblies/components). The aggregate BOM structure can be defined as the holistic BOM across the company with all identified product interrelationships. The more shared components and materials there are, as well as the deeper and more complex the aggregate BOM is, the more important this factor is. Through a process known as BOM decoupling, variability is absorbed, cumulative lead times are compressed and reduced, and planning is simplified by the insertion of ASR buffers at these strategic points in the BOM. What is important to note is that decoupling should not occur at every connection in the BOM, only at the connections that really make the biggest impact (more on this later). By combining the aggregate BOM concept with the BOM decoupling concept, key child components that compress the lead times of the most parents can be identified. In addition, currently stocked positions that do not truly compress lead times for parents can be identified and eliminated. We will explore this in depth in the section on ASR planning.

The Protection of Key Operational Areas-It is particularly important to protect critical operational areas from the bullwhip effect The bullwhip effect is the cascading disruptions through a dependent sequence of events. This undesirable effect of MRP and push distribution systems is well known. The APICS Dictionary (Blackstone, 2008, 15) defines bullwhip effect as "(a)n extreme change in the supply position upstream in a supply chain generated by a small change in demand downstream in the supply chain. Inventory can quickly move from being backordered to being excess. This is caused by the serial nature of communicating orders up the chain with the inherent transportation delays of moving product down the chain. The bullwhip effect can be eliminated by synchronizing the supply chain." ( APICS 2008, used by permission, all rights reserved.) Within manufacturing, it can be eliminated by synchronizing the pull across the production processes. MRP does not do this.

The longer and more complex the routing structure and dependent chain of events (including inter-plant transfers), the more important it is to protect key operations. These types of operations include areas that have limited capacity or where quality can be compromised by disruptions. In some cases, the creation of new part numbers and an insertion of an additional level in the BOM (as opposed to deleting layers) are necessary in order to decouple long and complex routings or sequences.

These factors are applied across the entire BOM and supply chain to determine positions for purchased, manufactured, and subcomponents and finished items (including service parts). Purchased parts chosen for strategic replenishment tend to be critical or strategic parts and long lead time items. Typically, this will be less than 20 percent of purchased parts. Manufactured parts chosen for strategic replenishment are often critical or strategic manufactured and service parts, at least some finished items, and critical subassemblies.

FIGURE 12-3 A supply chain for finished product A (FPA).

Typically, this will be under 10 percent of manufactured parts (for some environments with many manufactured service parts this percentage could be higher). On the fulfillment side, most parts will be strategically replenished-that is the whole point of having warehousing positions. It is important to note that on the fulfillment side, there is no difference between ASR and what is known as the TOC solution called replenishment (often referred to as the "distribution solution"). In Fig. 12-3 is an example of a supply chain for one product called Finished Product A (FPA) after the positioning has been determined. Notice that the "bucket" icon represents strategic replenished positions. Four of the ten purchased components are "buffered." Three of the ten subassembly/intermediate component positions are buffered as well as the finished product itself. Finally, the stock positions of FPA in all three regional warehouses are buffered.

The position of these buffers is accomplished through a combination of "thoughtware" and software. The "thoughtware "is the application of most of the above factors in consideration of the business objectives and operating rules by the people that have experience and intuition in the environment. In complex environments, software is often required to do the heavy computational lifting in order to analyze product structure, cumulative lead times, and shared components across the aggregate BOM. Finally, the importance of this step should not be underestimated. Without the right strategic positioning, no inventory system can live up to its potential.

2. Dynamic Buffer Level Profiling and Maintenance

Once the strategic inventory positions are determined, the actual levels of those buffers have to be initially set. Based on several factors, different materials and parts behave differently (but many also behave nearly the same). ASR groups parts and materials chosen for strategic replenishment and that behave similarly into "buffer profiles." Buffer profiles take into account important factors including lead time (relative to the environment), variability (demand or supply), and whether the part is made or bought. For instance, you could have a group of purchased parts that are long lead time and high variability (subject to frequent disruptions in supply) and you could have a group of manufactured parts that are short lead time and high variability (subject to frequent spikes). These buffer profiles produce a unique buffer picture for each part as their respective individual part traits are applied to the group traits. A list of both group and individual part traits that can apply to create that unique picture for every part is given in Table 12-2.

TABLE 12-2 Part Trait Examples This unique buffer picture is not just what the top level quantity should be. In Fig. 12-4, we see that ASR stratifies the total buffer level into different "zones." ASR uses a five-colored zone stratification approach. Light blue (LB; some authors refer to this as the white zone) describes an overstocked position. Green (G) represents an inventory position that requires no action. Yellow (Y) represents a part that has entered its rebuild zone. Red (R) represents a part that is in jeopardy. Dark red (DR, some authors refer to this as the black zone) represents a stockout. This color-coding system (the words used in the text and the abbreviations used in some diagrams as the figures are in black and white) will be used for both planning and execution priority and visibility and is integral to the power of the ASR solution. From a planning perspective, the color-coding will determine if additional supply is needed based upon the available stock position (on hand + open supply - demand allocations [including qualified spikes]). From an execution perspective, the color-coding will determine actions (primarily expediting or resource schedule manipulation) based on different types of alerts. This will be explained in the section titled "Highly Visible and Collaborative Execution."

Because each part within a buffer profile has different individual traits, it yields individual buffer levels and stratification zones for each part within that group (see Fig. 12-4). It is important to note that the zones need not be of equal proportions. Instead, the percentage of each zone is determined by the type of buffer profile to which the part belongs. The illustration to the right in Fig. 12-4 shows three parts in the buffer profile group "A-10." Each of the parts has a different top level and stratification levels because they have different individual part traits.

Note: companies will know if their buffer profiles are correct when the on-hand (not available stock) inventory position should average in the lower half of the yellow zone.

The color-coding also allows planners and executives to see how many overstocked as well as out-of-stock parts there are at any one time. If you combine the raw material value with the overstocked items, you can determine quickly how much excess cash is stored in excess inventory. Remember, that while being able to see stockouts is important what is really damaging is stockouts with demand allocations against them, which reinforces the need for visibility based on the available stock equation.

FIGURE 12-4 ASR stratifies the total buffer level into zones.

3. Dynamic Buffers

Over the course of time, group and individual traits can and will change as new suppliers and materials are used, new markets are opened and old markets deteriorate, and manufacturing capacity and methods change. Dynamic buffer levels allow the company to adapt buffers to group and individual part trait changes over a rolling time horizon. Thus, as these buffers encounter more or less variability, they adapt and change to fit the environment. Please note that the length of the rolling time horizon is very specific to the environment. Some companies may choose a 3-month roll while others must use 12 months. Figure 12-5 illustrates how a buffer can adjust based on actual consumption. The initial buffer size (based on its buffer profile and individual part traits) can be seen at the far left of the figure. The black line represents the available stock position, while the grey line represents demand per week. Let us say that, for this part, we were using a three-month rolling time horizon. Over the course of a 24-month period, you can see that demand rose dramatically, began to taper off, and then eventually stabilized. The buffer followed the trend.

Additionally, these individual buffer profiles can be manipulated through something called "planned adjustments," based on certain capacity, historical, and business intelligence factors. In ASR, these planned adjustments represent the necessary elements of planning and risk mitigation required to help resolve the conflict between predictability and agility. These planned adjustments are manipulations to the buffer equation that affect inventory positions by raising or lowering buffer levels and their corresponding zones at certain points in time. Planned adjustments are used for common situations like seasonality, product ramp-up and ramp-down, and capacity ramp-up and ramp-down. In the seasonality example in Fig. 12-6, you can see a product that has a substantial bulge in demand once per year. In the part ramp-up example, you see a part that is being ramped up based on a sales and marketing plan. In the part ramp-down example, you can see a part that is being discontinued. In all cases, if the planned adjustments do not follow the actual consumption (who ever heard of a Sales and Marketing Plan being accurate?), the color-coding/buffer stratification system will quickly identify that things are not going according to plan.

FIGURE 12-5 Dynamic buffer maintenance.

FIGURE 12-6 Buffer profile adjustments.

The combination of these first two solution elements (strategic inventory positioning and dynamic buffer level profiling and maintenance) of ASR creates strategically placed points of inventory that are actively managed, carefully sized, and dynamically adjusted. These buffers dampen or eliminate the effects of variation caused by the bullwhip effect and system nervousness that are passed up and down the chain of resources and dependencies (Fig. 12-7a).

4. Pull-Based Demand Generation

Most Purchasing, Materials, and Fulfillment organizations have limited capacity and trust when it comes to sorting through the current demand signals and planned orders generated by MRP. The volume of reschedule messages is impossible to work before more changes happen and the process begins again. Many times critical actions are missed or incomplete pictures are painted. A significant understanding of MRP logic is required to even begin to understand the implications of a reschedule message. Many times it is easier to just leave it alone than risk disrupting the operation. However, this short-term fix can generate the need for many expensive corrective actions later (expedites, premium freight, overtime, etc.).

FIGURE 12-7a Conceptual illustration of dampening effect of stock buffers.

Generating, coordinating, and prioritizing all materials signals becomes much simpler when the environment is modeled properly. The current inventory status is evaluated for potential negative impacts and flagged for alert against open supply orders and demand allocations, which includes future sales orders that meet specific spike criteria. Planners then have the ability to see where the signals are really coming from and react, before they get into trouble. This better matches the current intuition of the planners, but now they have the real visibility to establish correct and comprehensive priorities.

Key components of the ASR supply generation process include the following.

Demand Driven

Buffer levels are replenished as actual demand forces buffers into their respective rebuild zones. It is important to note that the buffer level driving demand generation is based on an "available stock" equation (as opposed to on-hand). Available stock is calculated by taking on-hand inventory plus open supply minus demand allocations. Actual on-hand inventory position relative to the buffer zones will provide the execution priority (discussed later in this chapter). Figure 12-7b shows the difference in relative buffer position between available stock and on-hand. The black arrows indicate the on-hand position and the white arrows represent the available stock position. This type of visibility gives relatively clear signals from both a planning and an execution perspective.

For example, part "f576," according to its available stock position relative to its defined buffer zones, clearly needs additional supply created. Additionally, part "r672" does not need additional supply but rather the existing open supply needs to be considered for expedite.

Decoupled Explosion

Component part requirements are calculated by pegging down through the BOM. However, this planning is decoupled at any buffered component part that is independently managed by an ASR buffer. This prevents the tsunami wave (nervousness) from rippling throughout the company as it does under MRP when a disruption occurs. The decoupled explosion from our earlier example of FPA is shown in Fig. 12-8. Note that whenever a buffer position is encountered, the BOM explosion stops. The figure on the left depicts the explosion for the parent item FPA after its available stock position has been driven into the yellow zone. The middle figure represents buffered children that independently explode when they have reached their respective rebuild zones. Finally, you see the explosion for Sub-Assembly B (SAB) after its available stock equation has been driven into yellow.

FIGURE 12-7b Available stock versus on-hand.

FIGURE 12-8 Decoupled explosion.

Material Synchronization

Component parts with incoming supply orders that are out of synch with demand allocations from parent work orders must be highlighted. This allows the Planners to take actions or make adjustments before work is released to the floor. This reduces the confusion in manufacturing and eliminates a significant amount of expediting.

ASR Lead Time