Manufacturing capacity planning and production scheduling process

To run a smooth supply chain, manufacturers need to keep track of many types of data, including order capacity, bills of resources and production schedules. In this book chapter excerpt, find out how Oracle supply chain management (SCM) software can help with manufacturing capacity planning and production scheduling processes and get tips for using Oracle SCM software.

Table of contents:
Oracle E-Business Suite Manufacturing and Supply Chain Management
Manufacturing capacity planning and production scheduling process
Manufacturing tips for Kanban production control and BOM management

Capacity Planning

Oracle provides two types of capacity planning within its planning products: Rough Cut Capacity Planning (RCCP) for evaluating the high-level capacity requirements of a master schedule and Capacity Requirements Planning (CRP) for evaluating the detailed capacity requirements of an MRP (or ASCP) plan. Both types of capacity planning compare the available capacity with the calculated requirements. Available capacity comes from your resource definitions, as described in Chapter 4; what differentiates CRP from RCCP is the method of calculating the resource requirements.

The following book excerpt is taken from Chapter 8 of the McGraw-Hill Osborne Media book, "Oracle E-Business Suite Manufacturing & Supply Chain Management," by Bastin Gerald, Nigel King and Dan Natchek. Click to download the full chapter.

Capacity Requirements Planning (CRP)

CRP calculates required capacity from the routings used to build your products. It is always run as part of running an MRP (or DRP) plan, since it needs to know the planned orders for its calculation.

In calculating resource requirements, CRP uses the primary routing for planned orders, but will use alternate routings if they have been specified for existing jobs. It recognizes modifications you have made to WIP resources (for example, specifying the use of an alternate resource). Because it is run as part of MRP, it benefits from the netting logic inherent in planning -- if your on-hand quantities don't require planning of new replenishment orders, you won't generate capacity requirements. And if jobs on the shop floor are partially completed, CRP recognizes which resources have already been used and only plans requirements for the resources that remain.

Rough Cut Capacity Planning (RCCP)

RCCP lets you check your master schedule against critical resources that you identify in a bill of resources. RCCP can be used to evaluate either master production or master demand schedules. In a manufacturing environment, it is more typical to use RCCP to evaluate production schedules because the constraints are most often manufacturing resources. In a distribution environment (or for a quick check on distribution-related resources), you can use RCCP to evaluate master demand schedules.

What makes rough cut rough? Since it does not use detailed routings, it is inherently less detailed. Though it might seem obvious, a hidden implication is in the requirements for subassemblies. Because RCCP does not depend on the netting and explosion logic of MRP, it may show resource requirements for subassemblies even if a subsequent MRP plan will find that there are no net requirements because of on-hand quantities. And there is no visibility to the progress of WIP jobs.

Bills of Resource RCCP calculates resource requirements simply by multiplying the master schedule quantity by the resource requirements expressed in a bill of resources that you define for the master scheduled item. In a manufacturing environment, you can construct a bill of resources by rolling up the routings; this process is called a bill of resource load. The result might be considered a kind of "aggregate routing" for the item; it lists all the resources you need to manufacture the master scheduled item and all its subassemblies. You can then modify the bill of resources by deleting noncritical resources or by adding resources that are not normally part of your routings -- shipping capacity, for example.

Strictly speaking, a bill of resources is a listing of the resources you want to evaluate for a single item. But, to be useful, any evaluation of capacity must look at all the demands placed on your resources. It doesn't do much good to know that item A requires 20 hours of a resource, unless you also know the demands placed by items B and C -- it's the total resource requirements that you must compare to the resource availability. For this reason, bills of resource are grouped into a bill of resource set; when you evaluate rough cut capacity planning, you must specify which bill of resource set you want to use to calculate resource requirements. Like so many objects in the Oracle Applications database you can have multiple bill of resource sets; this might be useful for simulations, or to evaluate distribution resources against your demand schedules and manufacturing resources against your production schedules.

Once you have defined a bill of resource set, you can view your rough-cut capacity plans on demand; the calculations are done when you perform the inquiry or run the report.

Use the Load Bill of Resource concurrent program to calculate bills of resource from your routings. Parameters include

• The bill of resource set you are loading.
• Whether you want to preserve any manual additions or modifications you may have made to the bill of resources, if it already exists.
• A resource group, to limit the creation of the bills of resource to items within a specific resource group. (Resource groups are discussed in Chapter 4.)
• Whether you want to create a bill of resource for all selected items or for Model and Option Class items only.
• How you want to deal with other master scheduled items that might be structured into bills of material at lower levels.

This last option, labeled Rollup MPS Items on the Parameters window, deserves explanation. Consider the example in Figure 8-8; if the only master scheduled item were A, you would typically want A's bill of resource to show two hours of labor -- one hour to assemble C and one more hour to put B and C together to make A. But if C were also a master scheduled item, you would not want to roll its time up into A's bill of resource because you could be overstating the resource requirements. If your master schedule includes both A and C, a rough cut capacity plan would calculate one hour per unit for C and two more hours for A. This would overstate the capacity requirements -- one of the two required hours was already planned for by C's bill of resource, and only one more hour of labor is needed to assemble product A.

You avoid this problem by setting the parameter Rollup MPS Items to No; this "breaks" the rollup process at each master scheduled item and ensures that a bill of resources does not contain resources that were already accounted for on other master scheduled items lower in the bill of material. Note that the parameter does default to Yes. While it's possible to construct a scenario where this could make sense (e.g., you plan to do rough cut capacity planning only against your top-level items), it would be inappropriate if you plan to do rough cut planning for all of your master scheduled items. Consider this carefully when loading bills of resources.

You can also build bills of resource "from scratch" by listing the resources you want to check in your RCCP plan. If you define bills of resource manually in a manufacturing environment, you may want to include the critical resources you need to build subassemblies of the master scheduled item.

A bill of resource for an item consists of one or more rows of data that specifies the following attributes for each resource you list. On the Main tab, enter the following information:

• Department or Production Line and Resource critical to the production or distribution of the item These are often the same departments and resources you use in your production routings, but you can also define other resources that might not traditionally be part of a routing. For example, if shipping capacity is critical to your business planning, you could define it as a resource within a department and use RCCP for a quick evaluation. The key thing to remember is that you must define these resources and assign them to departments, defining their available capacity just as if you were going to use them in a routing. This process is described in Chapter 4.
• Source Item That is, the item responsible for the demand. In a rolled-up bill of resources, this will be the item whose routing contributed the resource usage, either a subassembly or the item itself.
• Setback Days The number of days prior to the schedule date that the resource is required. For example, if the resource is required two days prior to the schedule date of the item, enter 2.
• Usage Quantity The exploded quantity from the bill of material. This is the quantity of the source item required to make the item whose bill of resource you are defining; if the source item is the item itself, this quantity will be 1. Note that this is not the quantity of the resource required; you enter that information on the Discrete Resource tab.

On the Discrete Resource tab, enter the following information:

• Required Hours and Basis Type Just as on a routing, a Lot basis means that the resource hours are required only once per lot (in this case, master schedule entry); for example, a setup. An Item basis means that the required hours are need for each unit of the item on your master schedule.
• Op Seq The operation sequence where the resource is used. This will be populated for you by the bill of resource rollup; if you are defining (or modifying) a bill of resource manually, you may specify the correct operation or enter 1.

The bottom of the form displays the source of the entry: Rollup means that the entry is the result of a routing rollup. Manual addition or Manual update indicates that you have created or modified this entry manually; this distinction allows you to re-roll routings and optionally preserve manual modifications to your bills of resources.

Planning for Engineering Changes

A key element of planning is the ability to recognize anticipated bill of material and routing changes. This allows you to phase in changes to the products you manufacture and have planning react appropriately -- planning will stop generating orders for the obsolete components or resources and begin ordering the new components or resources.

Effectivity Dates

Both bills of materials and routings use effectivity dates, indicating when a component or routing operation begins to be effective, and optionally when the component or operation is disabled. Effectivity is determined by the start date of a planned order; planning (and WIP, if you manually create a job or schedule) will select only those components that are effective on the date the job is scheduled to start.

The effective date, labeled "From," defaults to the current date when you create a bill or routing; you can select a future date if the element is not yet effective. You cannot select a date in the past. If the ending date, labeled "To," is left blank, it indicates that there is no planned end to the effectivity of the component or operation.

Note that the To date is a disable date, the date (and time) that the component or operation is no longer used. This is important if you are replacing one component with another; you set the starting (From) date of the new component equal to the disable (To) date of the old component so that the new component becomes effective at the same time that the old component is discontinued. Unlike some systems that might term the end date an Effective Thru date, if you were to disable a component on the 15th, for example, but not start the replacement until the 16th, a job or planned order that started on the 15th would contain neither component.

When you're setting the effective dates for a new component, it's helpful to know its Total cumulative lead time. If you plan to add a component within its cumulative lead time, you might have to expedite orders in order to obtain the component in time.

Effective dates can be entered manually on bills or routings; effective dates on bills of material can also be maintained with Engineering Change Orders (ECOs). With Release 11i.5, you can also use ECOs to modify routings.

Engineering Change Orders (ECOs)

ECOs provide a more controlled environment for maintaining effectivity dates than the simple bill of material maintenance process. ECOs (described in Chapter 5) let you group multiple changes under a single control number. With an ECO, you set an effective date for all of the changes you want to make to an individual bill of material; this eliminates any possibility of "gaps" if you define a replacement component and lets you keep multiple component changes in synch with one another. ECOs allow the system to calculate the effectivity dates for a set of changes based on using up a component on a bill (or using up the assembly itself). ECOs let you control whether a set of changes is visible to the planning applications. And ECOs can base the schedule date of a change on the date on which an assembly or component is scheduled to be used up.

Schedule Date Changes

As you change effectivity dates, planning will replan accordingly. Moving a schedule date in (scheduling the change to occur earlier than originally planned) can result in expedite suggestions from planning; moving a date out could result in shortages of the old component, particularly if you have deactivated it.

If a pending change is driven by the use-up of the assembly itself, the schedule date of the change will be updated automatically if MRP calculates a new use-up date for the assembly. If the change is driven by the use-up of a component of the assembly, MRP will trigger the ECO Use Up Alert to notify the appropriate planner.

NOTE
At present, Use-Up planning is only integrated with MRP, not with Advanced Supply Chain planning.

Discrete versus Repetitive Planning Logic

Oracle Applications allow you to plan for both discrete and repetitive production methods in a single plan. Discrete production involves recording production against discrete jobs, sometimes called work orders or production orders. Discrete jobs represent production of a specific quantity of an item, completed on a specific date. Repetitive production records production against repetitive schedules, which represent a continuous rate production for a period of time. While much of the planning logic is the same, there are a few key differences between discrete and repetitive production that necessitate differences in the planning logic.

NOTE
Don't confuse repetitive production with flow manufacturing. Repetitive production is designed for relatively high volume, continuous production of standard items. Flow manufacturing is designed for rapid production of individual items, even customconfigured products, at varying rates of production.

Repetitive Production Concepts

A repetitive schedule represents continuous production over a period of time; it models an ongoing rate of production and ongoing consumption of components. Compare this with a discrete job that represents production of a specific quantity of a product on a specific date and models an individual requirement date for each component. Whereas a discrete job has a start date and a completion date, a repetitive schedule is defined by four dates:

• First unit start The date that the first unit is started on a production line
• First unit completion The date that the first unit is completed on the end of the production line
• Last unit start The date the last unit on the schedule starts down the production line
• Last unit completion The date that the last unit is completed

These dates are represented in the following diagram:

Discrete jobs have a total job quantity to be completed on a specific date. Each component on a discrete job is required on a specific date. Repetitive schedules, by contrast, represent a rate of production for a period of time (defined by the first and last unit completion dates). On a repetitive schedule, products are produced and components are required continuously over the duration of the schedule.

The nature of repetitive production requires slightly different planning logic than that described earlier; the following pages discuss differences between planning for discrete production and repetitive schedules.

Repetitive Planning Logic

The objective of repetitive planning is to calculate a smooth rate of production for a period of time. When planning repetitive production, the process calculates the rate of production by taking the average daily demand within user-defined buckets. This process requires some unique planning logic and uses some terminology that has not been discussed previously.

Planning Periods

You establish planning periods on the Planning Parameters form for each inventory organization. You define an Anchor Date, which marks the start of your repetitive planning periods. Initially, this should be the start date of the first repetitive planning period; after that, each time you run a plan the system will roll the anchor date forward, if necessary, to keep the anchor date consistent with the start of the first repetitive planning period. You also define two horizons as a number of days; this breaks up the planning horizon into three sets of periods. Within each, you define the size of the buckets. Smaller buckets are often used in the first set of periods, so that repetitive planning is more reactive to changes in demands, while larger buckets in the later sets of periods allow planning to generate smoother production plans.

When you define repetitive planning periods, you can choose whether to use workdays or calendar days. Though it seems counterintuitive, it is generally advisable to use calendar days to determine the length of your planning periods. Using calendar days lets you ensure that your planning periods always start on the same day of the week, regardless of holidays or other non-work days. Consider Figure 8-9; if you define your repetitive periods to start on a Monday and define each period as five work days in length, the first scheduled non-work day will shift the next period to start on Tuesday.

If instead you define your repetitive periods as seven calendar days long, each period will start on the same day of the week regardless of intervening holidays, as shown in Figure 8-10. Calculation of average daily demand will always use the number of work days within a period, even if the length of that period was determined by using calendar days.

Average Daily Demand

Consider the example in Figure 8-11. This example shows a weekly planning period, with demand of 40 on Wednesday of that week and demand of 10 on Friday. Since there are five working days in that weekly bucket, the average daily demand is 10 (50/5); consequently, planning will suggest a repetitive schedule with a daily rate of 10. Note, however, that this will not satisfy the demand of 40 on Wednesday; if you produce 10 per day, you will only have produced 30 by the end of the day on Wednesday and will not satisfy the total demand. To satisfy the demand, you may need to produce at a different rate than suggested, consider reducing the size of your repetitive planning buckets, or even consider if repetitive planning and production are suitable for a product with such an erratic demand pattern.

Day	1	2	3	4	5
Demand			40		10
Production Rate	10	10	10	10	10

FIGURE 8-11. Average daily demand

An implication of this planning method is that demand is generated daily for the components of a repetitive schedule. This generally results in a large number of discrete demands. If you are not prepared to respond to these individually, you should consider adding order modifiers, for example, a minimum order quantity or fixed days supply.

Recognition of Existing Schedules

Perhaps the most unconventional aspect of repetitive planning is that it does not recognize, nor suggest changes to, any existing schedules. It always suggests new schedules. In part, this is because repetitive schedules have no identifying number (like a discrete job); they are identified by the rate and the four key dates. Suggesting a change to any of these parameters, therefore, means suggesting a new schedule. Thus, repetitive planning does not need to know the existing schedules; its job is to suggest the right rate of production.

Another implication of the fact that repetitive planning does not consider existing schedules is that any changes you made to the component requirements of a repetitive schedule are not considered in planning. If you have made a component substitution, for example, the planning process does not recognize it. Thus, you must carefully monitor the availability of substitute material (or excess plans for the original material). While this may appear problematic, remember that repetitive production is designed for high-volume production of standard items; if you are making frequent changes to component requirements, the product may not be a good candidate for repetitive manufacturing.

Treatment of Order Modifiers

Like discrete planning, repetitive planning uses order modifiers -- minimum and maximum order quantities and fixed order quantity. While these are the same modifiers that discrete planning uses, they are interpreted as modifiers to the daily rate, not to the total schedule quantity. Keep this in mind if you use modifiers on repetitively planned items, or you can dramatically inflate your schedule rates.

NOTE
One exception to the use of order modifiers is the Fixed Days Supply modifier. Because repetitive planning inherently plans for a fixed number of days (as determined by your repetitive planning buckets), the Fixed Days Supply modifier is simply ignored for repetitive items.

Planning Controls

To allow you to build stability into repetitive plans, Oracle provides several controls that determine the maximum increase or decrease in repetitive schedules at different points in time. Oracle also allows you to specify how much overproduction you will tolerate, in order to reduce suggested changes to production rates.

First, within an item's lead time, planning will not suggest any new repetitive schedules. This reflects the fact that it is often difficult or impossible to change the production rate once an item has started down the assembly line.

Second, within an item's planning time fence, Oracle makes use of the item attributes Acceptable rate increase percent and Acceptable rate decrease percent. These attributes let you specify how much of a change is allowed within the planning time fence. These percentages are applied to the current rate of production and determine the maximum and minimum rates that planning will suggest for new schedules. For example, if the current production rate is 100 units per day, an Acceptable rate increase of 10 percent means that planning cannot suggest a new rate greater than 110 per day within the planning time fence. Similarly, an Acceptable rate decrease of 10 percent means that planning could not suggest a rate lower than 90 per day. Outside of the planning time fence, however, planning is free to suggest whatever rate is required.

Note that zeros and nulls are treated very differently in these attributes. A zero means the acceptable increase or decrease is indeed zero; no rate changes are allowed. Leaving the attribute blank (null) implies that there is no limit on the acceptable increase or decrease; planning can make whatever suggestions it wants.

To dampen minor fluctuations in repetitive schedules, planning uses another item attribute, Acceptable overrun percent. This is the amount that the current daily rate may exceed the optimal rate before planning will suggest a new rate. In other words, it is the rate of overproduction that is acceptable in order to avoid minor changes in scheduled rates. The overrun percent is applied to the optimal production rate, and the result is compared to the current aggregate rate. For example, if the optimal production rate is 100 units per day, and the Acceptable overrun percent is 10 percent, your current production rate could be as high as 110 per day before planning would suggest a new schedule with a lower rate.

Repetitive Schedule Allocation

Repetitive planning begins with the calculation of the total production required; this may be accomplished on multiple production lines, so the total required is referred to as the aggregate repetitive schedule. Once the aggregate rate of production is known, planning will allocate production across all the lines designated for the product. When you define production lines for an item, you specify the maximum rate the line can produce and assign a priority to that line and item combination; planning uses these rates and priorities to allocate production across lines.

The allocation process first attempts to assign production to the highest priority line (lowest number) for the item. Allocation will load the first line to its stated capacity, then load the second line to its capacity, and so on. If there is not enough capacity across all the designated lines for an item, allocation will load the excess production onto the first line. Note that while allocation will attempt to respect the maximum stated capacity of a line for an item, it does not consider the load placed by other items that may also be scheduled on the line. Also keep in mind that it will overload the primary line for an item, if necessary.

Repetitive Schedule Implementation

You can implement repetitive schedules from the Planner Workbench in much the same way that you implement discrete jobs; simply select the job or schedule for release, and initiate the load process from the Planner Workbench. This initiates the WIP Job/Schedule Interface program to load the job or schedule into WIP. But while a discrete job may be loaded in either a released or unreleased status, repetitive schedules are always loaded with a status of Pending -- Mass Loaded.

This status avoids a restriction in WIP on the existence of overlapping repetitive schedules. WIP will not allow schedules for the same item on the same line to overlap, unless the schedules have a status of Cancelled or Pending -- Mass Loaded. As mentioned earlier, planning always suggests the needed schedule and does not suggest changes to existing schedules. If you have already implemented a repetitive schedule for the same item, line, and time period, a new schedule would overlap. Thus, schedules implemented from the Planner Workbench always carry the status of Pending --Mass Loaded. You must release these schedules in WIP, respecting the restriction on overlapping schedules, before you can report production against the new schedule.

When you release a repetitive schedule, you have the opportunity of closing or canceling existing schedules; this lets you avoid overlapping schedules. Alternatively, you can simply modify the rate or dates of the existing schedule based on the suggestions of the new schedule; then you can cancel the new schedule. This is preferable if you have issued material or charged labor to the existing schedule, as it avoids misleading variances. The use of the Planner Workbench is discussed in detail in Chapter 11 and 12; repetitive production is described in more depth in Chapter 16.

Configuring Repetitive Planning

Repetitive planning requires the same setup as discrete planning. In addition, because of its unique requirements, it requires the following:

• Define production lines on the Production Lines form.
• Associate items and production lines using the Repetitive Line/Assembly Associations form.
• Define repetitive planning buckets on the Planning Parameters form.
• Designate the desired items to use repetitive planning logic on the Master Items or Organization Items forms.
• Set repetitive planning controls, also on the Master Items or Organization Items forms.