Tell us what you need...
We want to know how to serve you better. Through the development of additional eBooks, Video Downloads, and Direct Interactive Problem Solving Methods, we hope to do just that. But to do this we need information from you! Fill out our survey or send us an email so we can get directly involved in what you need & how best to provide answers:
 
Go to MCTS Home Page

ROUGH CUT CAPACITY PLANNING
Summary

In this section, three approaches to rough cut capacity planning are examined. The least detailed, the capacity planning using overall factors (CPOF) approach, is quickly computed but is insensitive to shifts in product mix. A second approach, bill of labor, involves multiplying two matrices, the bill of labor and the master production schedule. This approach picks up shifts in product mix, but does not consider lead-time offsets. The third approach, resource profile, takes lead-time offsets into account. Both the bill of labor approach and the resource profile approach implicitly assume a lot-for-lot policy for setting lot sizes. If some other technique, such as economic order quantity or the Silver-Meal algorithm is used, then either approach is a very rough estimate. For that reason, the bill of labor approach is recommended because it is easily implemented on a microcomputer and is just as accurate as the more cumbersome resource profile approach. In any event, rough-cut capacity plans should be used only to determine if sufficient capacity exists over broad time frames such as a month or a quarter.

Drum-buffer-rope is an emerging procedure (discussed below) that eliminates the need for iteration found in all three RCCP approaches. It is presently used by a small but growing number of corporations.

In this section, we examine the pro­cess of RCCP. First, we examine the role of RCCP in the overall production planning and control system. Then we look at three RCCP techniques and discuss the selection of a technique. Third, we examine the various decisions that are based on RCCP. Finally, we look at two alternative approaches to capacity management, line balancing under the Just-in-Time philosophy, and the drum-buffer-rope technique of the theory of constraints philosophy.

What is RCCP?

Capacity planning and control techniques were introduced briefly in previous sections. Material requirements planning (MRP) uses a master production schedule (MPS) of end items to determine the quantity and timing of component part production. MRP is capacity insensitive; it implicitly assumes that sufficient capacity is available to produce components at the time they're needed.

A problem commonly encountered in operating MRP systems is the existence of an overstated MPS. An overstated master production schedule is one that orders more production to be released than production can complete. An overstated MPS causes raw materials and WIP inventories to increase because more materials are purchased and released to the shop than are completed and shipped. It also causes a buildup of queues on the shop floor. Since jobs have to wait to be processed, actual lead times increase, causing ship dates to be missed. As lead times increase, forecast accuracy over the lead-time diminishes because forecasts are more accurate for shorter periods than for longer ones. Thus, overstated master production schedules lead to missed due dates and other problems. Validating the MPS with respect to capacity is an extremely important step in MRP. This validation exercise has been termed rough cut capacity planning (RCCP).

There is no general agreement on the level of detail that should be incorporated in the MPS validation. An APICS monograph (Berry, Voliman, and Whybark 1979) presents case histories of several companies, including details on the capacity planning process. Some companies used very crude techniques, other used detailed, time-phased, methods.

THE ROLE OF RCCP IN THE PRODUCTION PLANNING AND CONTROL SYSTEM

Figure 1 shows an overview of the entire production planning and control (PPC) process under MRP. Capacity management techniques usually are separated into four categories: resource requirements planning (RRP), rough cut capacity planning (RCCP), capacity requirements planning (CRP), and input/ output control. These represent the four time horizons considered. In an MRP system the typical sequence is to create a master schedule, use rough cut capacity planning to verify that the MPS is feasible, perform the MRP explosion, and send planned order release data to capacity requirements planning. Plossl and Welch (1979) describe the role RCCP plays in the overall PPC system:

In an MRP system, the functions of capacity planning and control are separated from the functions of priority planning and control. As Figure 1 illustrates, the capacity planning functions consist of resource requirements planning, rough cut capacity planning, and capacity requirements planning. Capacity control is usually performed by input/output control. Priority planning is the task of the MRP system. Priority control is determined on the shop floor by the use of a dispatching technique to sequence specific tasks on specific machines.

A common criticism of MRP is that it does not manage capacity well. This is a somewhat ironic criticism because few firms fully use the capacity management techniques described here. Two reasons that the techniques are not utilized are: (1) data requirements are quite high and (2) the process is designed to re iterative and therefore, is time-consuming. A third reason is that many companies do not have a stable MPS. If the MPS is unstable, capacity planning is a futile exercise.

All firms running MRP should use capacity management. Not to perform capacity management is to invite extremely wasteful manufacturing and inventory management. Companies that have unstable master schedules should recognize that the instability is a symptom of inadequate safety stock at the MPS level. The inadequate safety stock may itself be a symptom of the lack of an appropriate forecasting system or failure to measure forecast error. Once a system is in place that does an adequate job of forecasting demand, measuring forecast error, and providing adequate safety stock, the capacity management process can begin.

The capacity management process should begin by insisting on a stable master schedule, Last minute schedule changes are very expensive. Few companies that permit frequent schedule changes make any attempt to measure the cost of these changes to the company's profitability. If they did measure cost, the premium charged for making such changes would likely be much higher. The topic of the cost of MPS instability will be discussed later.

The next section develops a simple example that is used to perform rough cut capacity planning. After the master schedule is developed, a discussion of MPS validation is presented.

Developing the Master Production Schedule

In the Master Scheduling section, the development of an MPS is discussed. An example is presented that discusses the assembly of a lamp. For this section, we create a hypothetical lamp manufacturer that we will call Al's Lamps. Figure 2 shows the planning bill or “super bill.” Al's Lamps has developed an MPS, as shown in Table 1.

Al's runs only one shift, 40 hours per week, with a maximum of 10 hours per week overtime permitted. Given present employment levels, monthly production averages approximately 15,000 units without overtime and 19,000 units with maximum overtime. As Table 1 reflects, demand for lamps is quite seasonal, with a major peak during the winter holiday season and a minor peak at income tax return time. Al's wish was to have stable employment in order to produce superior quality, so they choose to operate a level production strategy, with overtime as needed.

When a level production strategy is used, finished goods inventory must be built in advance of the seasonal peak (recall this type of inventory is known as anticipation inventory). The minimum level of inventory planned is the safety-stock level, which provides protection against forecast error. Because Al's has little excess capacity to permit recovery from poor forecasts, they carry a fairly substantial safety stock, 8,000 units. The maximum level of inventory planned is the amount needed at the end of October to meet the holiday peak season. For 1991, that maximum is 47,900 units.

Because Al's produces a single product family and because Al's strategy of stable employment limits the options available for aggregate planning, the process of preparing the MPS is quite simple. Current inventory is known. The desired year-end inventory is known (15,000 units, enough to provide anticipation inventory for January, 1992). Minimum production is 15,000 units per month, or 180,000 units per year. Minimum Production yields 17,000 too few units for the year; total demand is forecast to be 197,000 units. Because a maximum of 4,000 units per month can be built using overtime, and because; anticipation inventory should be built as late as possible to minimize holding costs, the 17,000 units to be built on overtime are allocated to December, November, October, September, and August, in that sequence. (The observant reader might note that inventory for the year could be lowered by planning to end with 11,000 units rather than 15,000 and planning to use overtime in January 1992. Al's prefers to use overtime in January only if the forecast for a given holiday season was pessimistic.)

Validating the Master Production Schedule

Al's wishes to validate the master production schedule at five key resources: lamp assembly, oven, base forming, plastic molding, and socket assembly. Lamp assembly assembles the base, the socket, and the shade to complete the lamp. The oven bakes the ceramic clay created at the base-forming department. Plastic molding creates the plastic lampshade. Socket assembly assembles all socket components except the power cord. The RCCP technique is used to verify that adequate capacity exists at each of the five stations. The technique consists of comparing a machine load report of capacity required to planned-available capacity at each work center. An example of a machine load report for plastic molding is shown in Figure 3. As Figure 3 clearly shows, plastic molding has adequate capacity. There are 400 standard hours available without overtime (500 with overtime) during each month of the year. Capacity required varies from 320 standard hours to 380 standard hours. Thus, plastic molding should be able to meet the production schedule all year without overtime.

The next section presents three techniques for developing the machine load report to determine capacity required. The following section discusses the determination of capacity available and decisions to be taken if available capacity is inadequate. Both discussions assume a stable MPS. The three techniques are known as capacity planning using overall factors, the bill of labor approach, and the resource profile approach. The section ends with a discussion of selection of an appropriate technique.

RCCP TECHNIQUES

The three techniques discussed in this section are similar in purpose but have substantially different data requirements and computational complexity. All three techniques are designed to convert the master production schedule from units of end items to be produced into the amount of time required on certain key resources. Because the amount of time available on existing key resources can be determined well in advance, the use of RCCP permits planning for expansion of these resources in a timely fashion. In some instances, the RCCP process may reveal that key resources are presently inadequate and that expanding the resources requires more time and/or more money than the company is willing to invest. In these instances, the MPS must be revised.

The three techniques are capacity planning using overall factors, the bill of labor approach, and the resource profile approach. Capacity planning using overall factors requires the least detailed data and the least computational effort. Not surprisingly, it is also the approach that is most affected by any changes that occur in product volume or the level of effort required to build a product.

The bill of labor approach uses detailed data on the time standards for each product at the key resources. A time standard is the time it should take an average worker working at a normal pace to produce one unit of an item. The time standard for any part has built into it an allowance for rest to over­come fatigue, an allowance for unavoidable delay, etc. When a time standard first is set, it usually is quite reliable. Because production processes continually are improved, time standards become less reliable; a time standard that is two or three years old is probably somewhat pessimistic. For this reason, many companies are reluctant to use time standards in performing capacity management. Many companies, especially those with incentive systems, do a good job of updating time standards. However, poor standards are not an absolute barrier to capacity planning; the capacity planning process contains an adjustment factor known as “efficiency” that may be used to correct for outdated time standards (although efficiency is not intended to reduce the need to keep standards up to date).

The final, and most detailed, technique is known as the resource profile approach. Like the bill of labor approach, the resource profile approach re­quires time standard data. In addition, the resource profile approach requires the lead-time required to perform certain tasks.

Capacity planning using overall factors can be performed using a calculator. The bill of labor approach and the resource profile approach both can be performed using an electronic spreadsheet on a first generation microcomputer such as the original IBM PC. Computational complexity is therefore not currently a barrier to capacity management, although prior to the introduction of the electronic spreadsheet about 1980, computational complexity was a barrier to capacity management for many companies.

Capacity Planning Using Overall Factors (CPOF)

CPOF requires three data inputs: the MPS, the time the total plant requires to produce one "typical" part, and the historical proportion of total plant time required by each of the key resources. If more than one product family exists, one "typical" part time is required for each family. CPOF multiplies the "typical" time by the MPS quantity to obtain total time required in the entire plant to meet the MPS. This time is then prorated among the key resources by multiply-mg total plant time by the historical proportion of time used at a given work center. Table 2 shows the RCCP for Al's Lamps produced using CPOF.

Al's Lamps has only one product family, lamps. A typical lamp requires 0.22 standard hours of labor/machine time, as shown in Table 3. Multiplying each monthly MPS quantity shown in Table 1 by 0.22 yields the Total Capacity Requirements row of Table 2. The remaining rows of Table 2 are found by multiplying the total capacity requirement for the month times the historical proportion for the work center. For example, the value 1,501.5 standard hours for lamp assembly for January is found by multiplying the historical proportion, 0.455, by January's total capacity requirement, 3,300 hours. This computation can easily be performed for all work centers for all months using an electronic spreadsheet.

Bill of Labor Approach

A good definition of the bill of labor (also known as bill of resources or bill of capacity) is given in Conlon (1977):

The bill of labor is a listing by item number of the amount of labor required by a major labor category to produce that item or group of part numbers. It is not intended to be a routing, but merely a means of estimating the capacity requirements for a particular item. The bill of labor (BOL) may be compiled for every distinct item or for groups of similar items, and extended by the scheduled quantities to determine capacity requirements.

In order to illustrate the concept of the bill of labor approach, we will use data from Al's lamps, introduced in the previous section. The bill of labor for lamp LAXX, a "typical" lamp is shown in Table 3. (The time standard data is taken from engineering files. For this hypothetical example, we elected to use the same data used for the CPOF approach.)

To determine capacity required, the number of lamps to be assembled each month must multiply the time per piece shown in the bill of labor. The assembly requirements are taken from the master production schedule, Table 1. To determine the total time required by a department in a given month, the number of lamps to be built during the month multiplies the time per lamp in the department. For example, it requires 0,1 hours to build one lamp at lamp assembly. January's MPS quantity is 15,000 lamps. There­fore, 1,500 standard hours are required at lamp assembly during January. All other entries in Table 4, the RCCP using BOL, are calculated in a similar fashion. Perhaps you have noticed that these repetitive computations are an ideal electronic spreadsheet application.

Although RCCP using the BOL approach can be performed with an electronic spreadsheet, virtually all of the necessary data is probably on a corporate mainframe. For this reason, many companies prefer to use a commercial capacity management software package. Most commercial software packages have the ability to display the rough-cut capacity planning as graphic output, in a form such as Figure 3. This graphical representation simplifies identification of occasions in which capacity required exceeds capacity available.

If Al's Lamps produced more than one product, the time required for each product in each department would have to be determined. The sum of alt product times for one-department gives that department's capacity required. In matrix notation, Table 4 is a five row by twelve-column (5 x 12) matrix. Note that the master schedule, Table 1, is a one row by twelve-column (1 x 12) matrix arid the bill of labor is a five row by one column matrix (5 x 1). Students familiar with matrix multiplication will have recognized the process used to obtain the rough-cut requirements using the bill of labor approach as a matrix multiplication. (The MPS must be transposed to enable multiplication.) Multiplying the time at Work Center 1 for Product 1 by the demand for Product 1, multiplying the time at Work Center 1 for Product 2 by the demand for Product 2, and adding the two results find the RCCP value for the first work center for the first month.

Figure 4 provides a generalized example of:

Capacity Required = Sum, from k=1 to n of aikbkj for all i, j

for a two product, two month, two work center case. Figure 5 provides a specific example. As Figure 5 shows, the RCCP value for the first work center for the first month is found by multiplying the time at Work Center 1 for Product 1 by the demand for Product 1, multiplying the time at Work Center 1 for Product 2 by the demand for Product 2, and adding the two results. Further examples of RCCP involving multiple products, and multiple periods are worth exploring to verify comprehension of the concepts.

Resource Profile Approach

Neither the bill of labor approach nor the CPOF approach considers lead-time offsets. Both approaches assume that all components are built in the same time period as the end item. The resource profile technique time phases the labor requirements. Each bill of labor must be time phased for the resource profile approach to be used. The resource profile technique is the most detailed rough-cut approach, but is not as detailed as capacity requirements planning.

A resource profile for Lamp LAXX could be developed here, see Table 5. This table would be identical to the bill of labor except that the time at each department is now associated with a specific time period, reflecting the lead-time of the part. Al's lamps have a three-month lead-time. In the first month, the bases are formed. In the second month, the bases are processed through the oven, the socket assembly department creates socket assemblies, and shades are created at plastic molding. In the third month, the lamps are assembled from constituent components and subassemblies. These lead times are unrealistically long; the numbers were created to produce a useful example. To create a resource profile, the lead-time must be converted to periods prior to the period in which the order is promised. Since the last operation always occurs immediately prior to delivery, it is shown as occurring 0 periods prior to delivery.

Once the resource profile is created, the rough-cut requirements are obtained by multiplying the resource profile by the MPS. This multiplication is not the simple matrix multiplication of the bill of labor approach. Rather, the procedure must keep careful track of the hours accumulated in each period. The resource profile approach is always implemented on a computer because of the tediousness of the calculations.

A generalized example of the resource profile approach for a case involving two end products, two work centers, a three-month planning horizon, and a three-month lead-time (for the product having the longer lead time) should be developed and worked through here, see Figure 6. Let’s say a resource profile for Product 1 at Work Center 1 is split into three parts, the time required in Work Center 1 in the month the order for Product 1 is due, the time required in Work Center I one month before Product 1 is due, and the time required in Work Center 1 two months before Product 1 is due. Any or all of these times may be 0 for any given product and time until due date at any given work center, To find the time Work Center 1 works on Product 1 during Month 1, we multiply Month l's demand by the time required at Work Center I during the month the product is due, we multiply Month 2's demand by the time required at Work Center 1 one month before the product is due, and we multiply Month 3's demand by the time required at Work Center 1 two months before the product is due. This process is then repeated for Product 2. We then add all six products (i.e., results of the multiplication process) to obtain the final value for the time required at Work Center 1 during Month 1, see term C11 in Figure 6.

Because the end-of-horizon effect is unavoidable, the planning horizon for the resource profile approach must be quite long, To provide the same unbiased visibility as the bill of labor approach, the planning horizon must equal the bill of labor approach's planning horizon plus the lead time minus one period.

The results of RCCP using the resource profile approach for Al's Lamps can be developed as a table, see Table 6. This example would be a special case of the resource profile approach) because the product requires effort during only one period for all work centers. For lamp assembly, the results are identical to the BOL results, which has a time offset of 0. For the three operations having a one-month lead-time offset (oven, plastic molding, socket assembly), BOL results are shifted one month to the left. For base forming, results are shifted two months to the left. The left shift of some department times, reflecting the lead-time offset, results in an end-of-horizon effect during November and December, as predicted. Time requirements for November and December are 0 for some departments, reflecting the fact that the end-item orders that require production in these departments will not be due until the following January or February. Since the planning horizon ends at December, no requirement can be shown at present. Users of the resource profile approach must be careful that the planning horizon is sufficiently long so that the first few periods do not exhibit an end-of-horizon effect. The first few periods of a planning horizon arc the ones that must be accurate to provide MPS verification.

Choosing an RCCP Technique

Obviously, the resource profile approach requires more computational effort than the bill of labor approach. Is there a return for the additional effort of creating the resource profiles and performing the extra computations? There may be a return, but only if lead times are quite long and the shop uses a lot-for-lot policy in establishing lot sizes.

It is not uncommon to find products whose manufactured lead-time runs several months. Most large, complex items such as airplanes, machine tools, and so on have very lengthy lead times. For parts having lengthy lead times, the resource profile approach might be useful because the bill of labor approach assumption that the components and the end item are built during the same month must be wrong. The resource profile approach avoids this assumption by including the time dimension. However, be aware that although the resource profile will have different information than the bill of labor approach regard­less of how lot sizing is done, it is likely to have more accurate information only if lot sizing is lot-for-lot.

Since both the bill of labor and the resource profile results are incorrect when any lot sizing rule other than lot-for-lot is used, the bill of labor approach with time buckets as large as practical is recommended (i.e., monthly or quarterly buckets). For those companies having lot-for-jot lot sizing throughout all operations, the resource profile approach using small (e.g., weekly) buckets will provide quite a bit of additional information. Consider the profile shown for Table 7 and the master production schedule for the coming 4 weeks shown in Table 8.

The RCCP would be as shown in Table 9 and Table 10 respectively.

For an environment in which large batches lead to large week-to-week shifts in product, mix, the following analysis is correct: The CPOF approach utilizes less data than the bill of labor approach, but is insensitive to shifts in product mix. (The historical proportions of total hours used by CPOF reflect average product mix. They are insensitive to changes in product mix that occur because of seasonality and/or batching of components.) This statement explains both why CPOF has been used and why it should not be used now that microcomputers are commonplace. In the era before microcomputers (that for business purposes began about 1981) and even before inexpensive hand held calculators (that, perhaps surprisingly, began about 1976), the process of per­forming all of the multiplications required by the bill of labor approach was quite cumbersome. The CPOF approach was much simpler, requiring only that each total labor quantity, i.e., one multiplication per cell rather than several multiply each proportion. For many situations, CPOF was perhaps the only practical solution.

From the viewpoint of the traditional MRP user, the bill of labor approach is superior to capacity planning using overall factors, because it better predicts the actual change in hours required from week to week, From the viewpoint of the Just-in-Time philosophy discussed in a later section, there should not be such large product mix variations from week to week. If the lot size of each product were very small, then it would be possible to make each product every week. Since it is likely that the product is in fact consumed every week, then there is an advantage to matching the rate of production to the rate of sale or consumption so that inventory need not be stored. A secondary advantage of the "level load" philosophy of Just-in-Time is that capacity management is simplified,

The bill of labor approach captures the changing product mix since each end product has its own bill of labor. For that reason, the bill of labor approach is strongly recommended over the CPOF approach. Given modern electronic spreadsheets, the bill of labor approach is just as simple to perform as CPOF. Thus, CPOF, while a useful technique only a few years ago, is clearly outdated and soon will be discussed only for historical interest.

RCCP DECISIONS

In this section we discuss how to determine the amount of capacity that is available. How to compare capacity available to capacity required, and the options that exist for adjusting capacity available and/or capacity required.

Determining Capacity Available

The plastic molding department of Al's Lamps, presented in the previous section, has three plastic molding machines. Since Al's works one eight-hour shift each day and there are 21 working days in an average month, it might seem that the capacity available to the plastic molding department is 504 hours per month (3 machines times 8 hours/day/machine times 21 days/month). How­ever, two additional factors must be considered. First, the plastic molding machines may not be available all the time. The machines may break down, the worker may be absent, and the mold needed or the material needed may not immediately be available. Second, there must be an adjustment between the time standard average and the actual average production rate of the department. The first adjustment factor is known as utilization. Utilization is a number between 0 and 1 that is equal to 1 minus the proportion of time typically lost due to machine, worker, tool, or material unavailability. The second adjustment factor is known as efficiency. Efficiency is formally defined to be the average of standard hours of production per clock hour actually worked. If a time standard is exactly right, efficiency is 1. If the time actually required to perform the work is less than the standard, efficiency is more than 1. If the time actually required to perform the work is more than the standard, efficiency is less than 1. As mentioned previously, time standards tend to be slightly pessimistic due to continual improvement in production methods.

Capacity available is found by multiplying time available times utilization times efficiency:

Capacity Available = Time Available x Utilization x Efficiency

Assume that for the plastic molding department of Al's Lamps, utilization is 0.756 and efficiency is 1.05. The time available in a month having 168 working hours (21 eight-hour days) is 504 (3 machines times 168 hours/month/ machine). Thus,

Capacity Available = 504 x 0.756 x 1.05 = 400 hours

rounded to the nearest hour.

Comparing Capacity Required to Capacity Available

Most standard software packages can determine both the capacity required and the capacity available and can display them in both tabular and graphical format. The report containing capacity available and required is known as the machine load report. The graphical format for the machine load report was shown in Figure 3. Graphical formats usually are preferred because one can determine at a glance whether capacity is adequate. Tabular formats arc more precise, however, and when capacity is inadequate one needs to know the exact shortage to be covered.

When capacity is inadequate, four basic options are available to increase capacity: overtime, subcontracting, alternate routing, or adding personnel. If no combination of the four options can provide sufficient capacity, the MPS will have to be reduced. Options to adjust capacity required or available are discussed next.

Overtime

Overtime is probably the most popular solution to inadequate capacity be cause few advance arrangements must be made. Usually workers appreciate having the money provided by some overtime; however, beyond some point, the workers would rather have the time off than the extra money. For this reason, many companies have a policy that limits the amount of overtime during specific periods. Also, almost alt departments have to meet a budget for the year, which imposes a constraint on annual overtime.

Subcontracting

A second way to obtain additional capacity is through subcontracting. Arrangements for subcontracting must begin well in advance to permit time to find a vendor capable of performing quality work. Subcontracting usually is more expensive than building an item in house on regular time (otherwise, we would buy the part and not make it). However subcontracting may be cheaper than building the part in house on overtime. Disadvantages of subcontracting are that lead-time usually increases, transportation cost may increase, and it is more difficult to guarantee a quality product.

Alternate Routing

If only a few work centers have excess work, the remaining work centers will tend to have too little work during a given period. It is, therefore, possible to consider a temporary change in the routing of specific parts so that work usually performed in Work Center A temporarily is performed in Work Center B. There are two possible reasons that Work Center B is not presently used-quality and time. If Work Center B cannot achieve the needed quality, alternate routing should not be considered. If Work Center B presently is not used because of time, alternate routing should be considered. Using Work Center B during time in which it otherwise would be idle is preferable to having Work Center A on overtime. Using Work Center B on overtime is an alternative to subcontracting. However, having Work Center B on overtime probably is not an alternative to having Work Center A on overtime. Why?

Adding Personnel

Adding personnel will add capacity provided equipment is not the constraint. There are three ways to add personnel: add a shift, add new hires to an existing shift, or move existing personnel from an underused work center. The time to consider adding a shift is when the master schedule first is formulated, when the demand chase versus level production versus mixed strategy choice is made. Adding new hires to an existing shift is likely to be an option only when the budget for the next fiscal year is being approved. Thus, the only short-term way to obtain additional personnel is to shift people from an underused work center to one that is overloaded. In union shops, union rules may prohibit or limit this option. Enterprises that have restrictions on moving workers between workstations are at a competitive disadvantage to companies that have no such restrictions. The Japanese strongly encourage a worker to cross train and master several different tasks. For this reason, many companies have insisted that union rules on multifunction workers be relaxed; most unions have been willing to trade flexibility for job security or other considerations. Clearly it is not in the union's interest for the enterprise not to be competitive.

Revising the Master Production Schedule

Most companies consider a revision to the MPS to be a solution of last resort in the event of insufficient capacity, to be implemented only when all other options are exhausted. MPS revision actually should be the first thing a company considers. A number of things can cause an order to be expedited. An order is rarely de-expedited. There may be several orders on the existing master schedule that no longer are needed as early as the current due date shows them to be needed. It is not uncommon to find a machine load report that shows too much work in the first two or three periods, but ample capacity beyond that. This problem may be corrected or substantially alleviated merely by putting the true need date on all orders.

If de-expediting, overtime, subcontracting, alternate routing, and shifting personnel collectively cannot provide sufficient capacity, MPS revision does in fact become the technique of last resort, It is important to understand that when an unavoidable capacity overload exists, it must be corrected. If insufficient capacity exists, it is impossible to complete all orders on time. Our choice is to have management decide whose order will be late based on the impact on the whole enterprise or to have a worker on the floor make the choice based on the convenience of one department. The worker must perform his or her tasks in some sequence. It is a mistake to Jet a more-or-less arbitrary sequencing decision determine that jobs are late. Even if the sequencing decision is ordinarily made by a rule such as earliest-due-date-first, when capacity overloads exist we may prefer to have jobs completed out of strict due date sequence.

Management must take responsibility to see that rough cut capacity planning is performed. If an unavoidable overload exists, management must take the responsibility to revise the job due dates in order to provide a feasible master production schedule. This is the meaning of master schedule validation.

DRUM-BUFFER-ROPE

The rough-cut capacity planning process just described is needed because the MRP technique assumes that needed capacity always is available. The medium-range planning systems used by Just-in-Time (JIT) and the theory of constraints (TOC) are somewhat simpler because both involve explicit recognition of capacity limitations.

The JIT approach maintains excess capacity in order to achieve very short lead times and to hold very little inventory. JIT also produces every item every day, changing the quantity per day about once a month to reflect the market rate of sales, so the matter of predicting the quantity and timing of needs for components is trivial. Medium-range planning under JIT consists merely of multiplying the daily rate of sales for the end item by the quantity of components per end item.

The theory of constraints says the constraint must pace the entire system. The constraint may be the market or it may be an internal resource (work center). If the market were the constraint, the theory of constraints schedule essentially would be identical to the Just-in-Time schedule. If an internal constraint exists, the TOC approach, called drum-buffer-rope, is to have the master scheduler sequence all jobs through the constraint resource. Since the lead time from the constraint to shipping is known, due dates can then be set for end items with the assurance that they are capacity-feasible. The resulting MPS can be fed to the MRP system, which will set operation due dates for all work centers except the constraint, (Since the process is deterministic, MRP should arrive at precisely the same due dates at the constraint that the master scheduler originally established.)

The name drum-buffer-rope is derived from the following metaphors: the schedule at the constraint is the drum, setting the pace for all work centers; the deterministic lead-time offset from the constraint to order release is the rope, pulling work into the shop at the pace the constraint is completing work; the linkage between the constraint and order release ensures that an essentially constant buffer is maintained at the constraint. If problems never occurred, the buffer would be constant. But since problems such as absenteeism do occur, the buffer at the constraint must be large enough to absorb these fluctuations and avoid idle time at the constraint. Because this technique is used only if demand exceeds capacity, any idle time at the constraint results in lost revenue.

Although drum-buffer-rope is not as yet common practice, a number of writers have recognized that the iterative process of MPS creation (RCCP to validate the MPS MPS revision RCCP repeated perhaps another MPS revision RCCP a third time etc.) simply is not an ideal way to perform the scheduling task. Because it is likely that when an overload exists one station is more overloaded than all others, and because a schedule for that one station easily can be created, drum-buffer-rope is a single pass technique that any company can use. Wahlers (1990) reports that Valmont Industries presently is using drum-buffer-rope with a great deal of success.

Back...
Keeping Promises...
Overcoming Resistance...
Turning Communication Into a System...
Identifying the Heart of What Really Matters...

ERP-MRP Evolution…
ERP & Hoshin Kanri…
ERP Implementations… 
Profit-Ability Improvement... (¬Click here to see definitions)
Profit-Ability Management Principles... (¬Click here to see definitions)
People, Empowerment & Profit-Ability… (¬Click here to access articles)
Hoshin Kanri & Deming's Plan-Do-Check-Act... (PDCA) Cycle…

Google

Valid XHTML 1.0!