SHIPPER MANUFACTURING COMPANY

Teaching Notes

Synopsis and Purpose

Shipper Manufacturing Company is a manufacturer of electrical products, laminated materials, and specialty products. The Advanced Products Division (APD), which manufactures the specialty products, has reformulated its corporate strategy. As a result, its manufacturing strategy should also be revised.

In the past, APD has manufactured custom products in low volume for its customers. The division plans to gradually add higher volume products for multiple customers. As a result, changes will be needed in manufacturing to effectively compete in the new environment.

The purpose of the case is to expose the students to the concept of a manufacturing strategy and the changes in manufacturing which are associated with a change in strategy. The case provides an effective vehicle for teaching this and also for the concept of integrating all aspects of operations.

Discussion Questions

1. What objectives should be adopted in manufacturing with respect to cost, delivery, quality, and flexibility?

2. How should the objectives in manufacturing be achieved though process, organization, equipment, work force, capacity, scheduling, quality management and production and inventory control systems?

Analysis

The impending change in business strategy implies a change in manufacturing toward higher volume and lower cost products. Presently the products are custom-designed and costs are passed through to the customer. In the future, the products will be producer-designed and costs will not necessarily be passed through to the customer. Because of more intense competition, cost becomes a more important objective, while flexibility is simultaneously reduced.

The quality objective will not necessarily be changed by the new business strategy. The division must decide where it wants to position itself on the quality spectrum.

The delivery objective will probably be changed toward shorter lead times in order to compete. Since the product is not custom designed, the customer may expect some products to be held in finished goods inventory. Customers may also insist that raw materials inventory be carried so that some products can be quickly produced.

C-1

The change in business and manufacturing strategy will cause changes in all aspects of operations.

1. Process/Equipment. As low cost and volume is emphasized, the production process will probably use more automated equipment. This will accompany standardization of the product and less flexibility for product and volume changes. As a result the production process will become less labor intensive.

Another expected change in the process is a tendency toward line flow processes and away from a job shop layout. Line flow will be a more efficient form of layout because of higher volumes and increasing automation.

We can also expect the flow of control information to change as the process shifts from make-to-order to make-to-stock. In this case the order cycle will become separated from the inventory replenishment cycle. We might also expect a slight movement toward more vertical integration, although the company is already substantially vertically integrated with its own materials division.

2. Organization. With the move toward higher volume production which is made-to-stock, a production control function will be needed in manufacturing. While the industrial group has shared services in materials management, a separate department in APD manufacturing is needed to focus attention on the production and inventory control task.

The APD should also consider moving quality (product) assurance under the manufacturing department. This will provide the opportunity to make quality more than merely "outside inspection" and to consider quality as part of the production process.

3. Work force. Changes in the work force will follow changes in equipment. The direct production work force will become less highly skilled and less well paid. There will be fewer people in the work force due to automation. As cost control becomes more important, there will be more emphasis on production standards and labor reporting. The work force will also probably become more alienated and dissatisfied with their work, unless management takes positive steps to prevent these attitudes from developing.

4. Capacity, Scheduling and Production and Inventory Control. There will be a gradual change in stocking policy until more raw materials and finished goods are carried in inventory. These changes will be needed to support higher volume production and to achieve better customer service (reduce lead times). As a result, internal changes in production and inventory control (PIC) will occur. PIC will need to upgrade the manual system to the new computer, and they will have to make the computer system effective so it is not duplicated by the manual system.

C-2

As the operation becomes more line oriented and make-to-stock, it will be possible to utilize a higher percentage of capacity. Also, scheduling will become more complex and critical to insure a smooth flow of materials and high capacity utilization.

The production and inventory control people will probably have to gain more professional knowledge. They can do this through training courses or through APICS affiliation. Some outside people may be brought into production and inventory control with the required knowledge and experience.

5. Quality. Quality control will need to become more high-volume-production oriented rather than job-shop or project oriented. This change may include the use of quality control charts, more emphasis on process control and inspection, and more careful training of employees. With high volume production, quality will need to be more preventative rather than reactive to avoid high scrap rates or too many returns from customers.

Some of the current team work may be lost as the organization becomes more structured toward mass production. Therefore, it may be necessary to institute quality teams or total quality programs to prevent deterioration of quality.

Teaching Strategy

This case can be taught in the same order as the discussion questions. It is helpful to identify the changes that will occur in manufacturing objectives before discussing detailed changes in manufacturing infrastructure. The changes in manufacturing should flow naturally from the changes in objectives.

This case serves as an overview of the entire course since it touches on all aspects of operations. It also helps emphasize the interrelationship of decision making in operations and the overriding role of strategy and policy. It should take about 45 minutes to teach the case.

C-3

FHE, Inc.

Teaching Notes

Synopsis and Purpose

FHE is a manufacturer of pumps and related fluid-handling equipment faced with organizational questions for management about new product development projects. There are technical program managers in the engineering department and product managers in the marketing department who seem to have overlapping job responsibilities. Questions are also raised about the load on central drafting services, since as many as twenty separate projects may all be demanding the same services at the same time. Manufacturing receives the specifications for new products only after design has been completed, preventing the consideration of manufacturing's concerns during the design stage. Finally, the case presents a scheduling method, using a Gantt chart, which provides the basis for a discussion of scheduling methods and procedures.

The purpose of this case is to examine the role of departmental relationships in the new product development process, to look at the influence that technologies such as CAD-CAM might have, and to evaluate typical problems encountered in project management. It can be used after reading Chapter 3 on product design, or after Chapter 14 on project management. This particular case does not require special computational or quantitative skills.

Discussion Questions

1. What steps should Mr. Donaldson take to improve the new-product development process at FHE?

2. What could be done to clarify the organizational relationship between product managers, technical program managers, and manufacturing managers?

3. What can be done to better manage the workload of the engineering services department?

4. Evaluate the plans and the expected results from the new CAD-CAM system.

Analysis

As far as question 1 is concerned, Mr. Donaldson should take the following steps:

1. Clarify the organizational relationship between the Technical Program Managers (TPMs) and the Product Managers. It is apparent in reading the job descriptions in Exhibits 2 and 3 that several points of conflict exist. For example, the Product Manager is given the following responsibilities which potentially conflict with the TPM:

I. C. "Monitor project activity and take action where necessary to ensure integrity of the project."

C-4

I. E. "Ensure the coordination of all technical resources related to new product development and introduction; to include engineering, manufacturing, marketing, and service.

II. C. "Monitor engineering change order activity and take action where required."

Each of these responsibilities involves some overlap with those stated in Exhibit 2 for TPMs, especially those areas where "technical" issues are involved.

2. Mr. Donaldson should develop a procedure to more accurately forecast load on the engineering services department. At the present time up to twenty different projects may require use of the engineering services department at the same time. If the load can be forecast in advance, then steps can be taken to establish priorities, authorize overtime, or take other measures to manage the load.

3. Mr. Donaldson should review the project management system in use. He needs to evaluate the cost, performance, and time results which have been achieved on past projects. This will give him a basis for deciding whether present project management systems are working or not.

Discussion Question 2 asks for specific actions which can be taken to clarify the organizational relationships between product managers and TPMs. Several approaches can be taken. Changes could be suggested in the organization chart, in job descriptions, or in "dotted-line" reporting relationships.

The organization chart has a president in charge of marketing, engineering, and manufacturing. Below that level these three organizations are still entirely separate. One step that could be taken to better integrate these three parts of the organization is to put the product managers, the TPMs, and the process design staff under a single manager who is in charge of product development. The organization chart, in part, would then be as follows:

C-5

Another approach to this problem is to clarify the job descriptions in order to reduce the apparent conflict. This will require the U.S. Marketing Manager and the Corporate Engineering and Quality Assurance Manager to get together and hammer out an agreement on how potential conflicts and overlap areas will be handled.

A third approach is to create product-based teams. Team members would include representatives from all involved functional areas: Marketing, Engineering, and Manufacturing. Team members would jointly be responsible for the development of a given new product. Such teams could be temporary -- disband when the project is done, or permanent -- be responsible for successive new product introductions. These teams members would report to their own functional areas but also have a dotted-line relationship to whomever is in charge of new product development, whether this would be the Marketing Manager or a New Product Development Manager.

C-6

President

U.S. Marketing Manager

Market Planning Manager

Product DevManager

Corp Engr & Quality Mgr.

Manufacturing Manager

Product Manager

Product Manager

Product Manager

TPM

TPM

TPM

Process Design

Regarding question three, the Engineering Services Department has a classic problem with support services being provided to several different projects from a single support department. One way to solve this problem is to ask each project manager to forecast the staff hours of support services (by type) that will be required for their projects. These forecasts should not be made too far into the future because of schedule changes, technical uncertainties, etc. Perhaps, a one-month forecast by week might be required from each project. The manager of the Engineering Services Department, Al Hanson, could then add up these estimates of workload for all projects by week and obtain a total estimate of manpower required in drafting, model shop, testing facilities and technical documentation areas. It would then be apparent where the capacity problems occur and alternative actions could be considered. These actions might include overtime, rescheduling projects, or obtaining outside resources.

The new CAD-CAM system could indeed serve as a way of coordinating Marketing, Engineering, and Manufacturing. However, such integration is not easily achieved; simply purchasing and installing the system will not cause interfunctional coordination to occur. Upon installation of CAD-CAM, each of the functional areas needs to integrate the system into its day-to-day operations before the benefits of improved coordination can be realized.

Teaching Strategy

In teaching this case, I start with the discussion questions which the students have prepared in advance. Question 1 allows me to get the general approach for improving new product development on the board. This approach will probably include issues from questions 2, 3, and 4 that the students may suggest.

After listing the general approach, the discussion can be turned to specific suggestions in each area. I generally take about ten to fifteen minutes to get the general approach defined on the board. And, then I take an average of about ten minutes to discuss each of the suggested improvements. The case can therefore, be discussed in about an hour with slightly more or less time possible depending on the pace and depth of analysis chosen.

It is often advisable for the instructor to provide a wrap-up and to highlight the key points or lessons taught in the case. Sometimes a student can also provide this wrap-up. In any case, students need a summary to pull things together especially after a number of different issues have been raised and discussed during the case analysis.

C-7

EASTERN GEAR, INC.

Teaching Notes

Synopsis and Purpose

Eastern Gear is a company that traditionally produced custom gears in small lots, and has recently been given the opportunity to begin producing larger volumes of gears per order. This is a typical small job shop which is owner operated. In addition, Eastern Gear has experienced a large increase in sales which is causing major problems in operations. Because of the wide range of problems presented, the case permits an overview of a job shop type of operation and practice at looking at the entire scope of operations including objectives, capacity, production and inventory control, organization, and quality.

The purpose of this case is to allow students to examine a complex manufacturing problem and the associated decisions in some detail. The case illustrates how problems are interrelated and how objectives might be clarified before decisions can be made. It also provides a fundamental understanding of what a job shop is like and the typical problems faced in job shop management.

Discussion Questions

1. What are the major problems being faced by Eastern Gear?

2. What action should Mr. Rhodes take to solve his problems?

3. How can this case be related to operations strategy and process design concepts?

Analysis

Eastern Gear is experiencing a wide range of problems including:

1. Need for objectives in operations. It is not clear at the present time whether operations should emphasize cost, flexibility, delivery, or quality. The desired emphasis on these objectives needs to be clarified.

2. Lack of an order size policy. Eastern Gear has accepted a wide range of order sizes as shown in Exhibit 2. Furthermore, Eastern Gear's President has just decided to accept a few larger orders. The two types of order sizes are best served by different operations strategies and different process designs.

3. Lack of planning for growth. It does not appear that the company has a strategy or plan for future growth. As a result, the company could experience cash flow problems, capacity problems and other problems associated with rapid growth.

C-8

4. Production and inventory control. At the present time expediting seems to be the rule rather than the exception. Twenty percent of the orders have rush tags on them. Production processing time has increased from two to four weeks, and there does not seem to be a production and inventory control system in place. Also, certain orders are being handled on a rush basis and this may be disruptive to the smooth flow of production. Finally, Joe Irvine has expressed concern about bottlenecks which are caused by the lack of effective production and inventory control.

5. Poor Quality. There seems to be a lack of quality control as evidenced by the six percent return rate. Many of the returned orders seem to be missing one or more operations.

6. Organization chart. The present organizational structure does not have a clear separation of functions. The Engineer is involved in shop floor supervision. The Controller is responsible for purchasing, which is a manufacturing function. The expediter’s job is defined too narrowly.

7. Raw material inventory. Due to lack of a raw material inventory, delivery lead time is extended by one or two weeks. Some of the high usage raw materials could be carried in order to reduce lead time.

8. Layout. The present layout of the shop results in irregular material flows and excessive travel distances.

The solution to these problems is intertwined and can start in many different places. We will discuss the possible solutions in order of the problems presented.

With regard to objectives for operations, it appears that delivery lead time is very important for the custom gear business. However, the operations objectives for larger order sizes are not necessarily the same; cost becomes more important in that situation.

The case refers to additional profits available from reduced delivery times. Based on Exhibit 4, the variable cost of material and labor is about 50% of total cost (assuming for the moment that other costs are fixed or semi-fixed). If sales can be increased by $500,000 through reduced delivery times or other means, as much as $250,000 could flow to the bottom line.

The custom gear business places more importance on delivery lead time as an objective than on cost. The gears are used for R&D where the costs of the gears themselves are probably a small percentage of the total customer's costs. However, the entire R&D project can be held up if the gears are not available in time.

Quality is important, at least to the extent that it affects the functioning of the customer product or causes a delay in schedule while repairs are being made.

Flexibility is also of great importance for Eastern Gear's customer gear business. As a result, Engineering and Marketing should be working closely with the customers to make sure they get what they need. Flexibility is the key to providing customer service. Generally speaking, the importance of flexibility is decreased with larger orders.

C-9

My ranking of the objectives from most important to least important for the custom gear business is: 1) flexibility, 2) delivery, 3) quality, and 4) cost. Larger order sizes might emphasize quality and cost rather than flexibility and delivery.

The decision by the President of Eastern Gear to accept a few larger orders has many implications for the manufacturing organization and its present problems. First of all, it aggravates the problem of lengthening delivery intervals, given the present job shop environment. Secondly, manufacturing in larger lot sizes is better accomplished using a line flow manufacturing process and layout.

The lack of planning for growth can be solved by developing at least a one-year plan. The plan should include market objectives, financial projections, and capacity plans, as a minimum. An elaborate plan is not needed, but the basic issues need to be addressed. As a result of this plan, a strategy for acquisition of space by operations should be determined.

Production and inventory control can be improved in several ways. First, a production load or schedule should be developed and constantly kept up to date. When a new order is accepted, it should be placed on the schedule and the delivery date promised accordingly depending on the load on the factory at the time of order acceptance. For a company of this size, the schedule could probably be kept on a personal computer. An elaborate system is not warranted for a small company such as Eastern Gear.

Rush orders should be handled through the scheduling system, and not by Fred Dirkson. The rush orders are probably causing delays in other orders, so the current rush order policy needs to be reconsidered. A premium could be charged for rush orders to make it worthwhile to save capacity in the schedule for these orders, or capacity could be saved for special customers simply to provide better delivery service.

Expediting should be limited to those orders which absolutely need it. Twenty percent of the orders with rush tags seems too high. With better scheduling, expediting might be held to a small percent of the orders.

The poor quality should be corrected immediately, especially since 75 percent of the returns are due to missing operations. An operations sheet should be prepared for each order. The operations should then be checked off as the order is routed through the factory. The operations sheet should stay with the order during production.

An inspector should be designated to check each order before it is shipped. Final inspection should include checking the operations sheet to be sure all operations were performed. Other final inspections should also be made as specified by the engineer. This amounts to "inspecting quality into the product", but is probably needed until the processes can be controlled.

The company should be reorganized to clearly separate functions. Engineering should be separated from manufacturing. With the present organization, critical engineering functions are not being done because the engineer is busy managing manufacturing.

Expediting should be eliminated and replaced by a production and inventory control manager. Purchasing should also be placed under manufacturing to form a materials management position responsible for coordinating scheduling, inventory and purchasing.

C-10

Perhaps a finance function should be developed in place of the controller and marketing should replace sales. The resulting organization chart would be as follows:

President

Marketing Engineering Manufacturing Finance

Materials Foreman Management

This organizational structure would require hiring one additional person, a manufacturing manager. It might also require training or replacing some of the present managers.

A small raw materials inventory could be carried at low cost. Since delivery lead time is of the essence, some raw material inventory would probably be a good idea. Carrying three items (E, H, J) would cover 47 percent of usage. Carrying an additional five items (A, D, G, I, and K) would raise the coverage to 82 percent of usage. If delivery lead time is as important as the case indicates, it might pay to carry all eight items. In this case 82 percent of usage is covered by 20 percent of the items -- a good illustration of the ABC principle.

The layout needs to be improved by streamlining materials flows. One way to do this is to interchange the heat treating and drilling departments. If this is done, there will be little backtracking and a smooth flow of materials. Eastern Gear should consider incorporating a manufacturing cell into its facility. The cell would utilize a line flow layout and would serve as a more appropriate method for manufacturing large orders or possibly rush orders.

If the above steps are taken, delivery lead time should be reduced to its previous level of 3-4 weeks maximum and 5 days on rush orders. Returns for quality should be reduced from 6 percent to about 1.5 percent of sales. 1988 sales should increase to at least $5.5 million and profitability should be improved.

Teaching Strategy

I teach this case by getting a list of problems on the board such as the list given above. This will usually take about 15-20 minutes depending on how much discussion develops.

I then ask for solutions in each of the problem areas which takes another 45 minutes or so. I wrap up the case by emphasizing the role of objectives and tradeoff decisions in operations. I also indicate how all the problems are interrelated and they should, therefore, not be viewed in isolation. For many students, the case is their first exposure to a job shop or manufacturing company. They consider Eastern Gear to be a good learning experience.

C-11

THE “mi adidas” MASS CUSTOMIZATION INITIATIVE

The teaching note for The “mi adidas” mass customization initiative can be obtained from IMD. IMD does not permit reprinting of their teaching notes. The teaching note can be obtained by contacting IMD at

Tel: +41-21-618-0217

Fax: +41-21-618-0707

IMD - International Institute for Management Development

Chemin de Bellerive 23, P.O. Box 915, CH-1001 Lausanne

Switzerland www.imd.ch

The IMD Case number for “mi adidas” is IMD-6-0249

C-12

PLASTECH, INC.1

Teaching Note

Synopsis and Purpose

The Plastech case opens with Lincoln Smith reviewing the financial statements for fiscal years 1992 and 1993. The company has seen a dramatic fall in profits, barely breaking even in the current year. Lincoln must determine the cause of the decline and propose a plan of action.

The case requires the student to analyze financial and operating statistics to determine what caused the decline in profits. The detective work should show direct labor productivity falls dramatically, which was caused by the addition of a third shift, which was required to accommodate the additional setup time, which was caused by the greater number of customers and products, which was caused by the change in marketing strategy. Students should conclude that the new marketing strategy resulted in a mismatch between the product mix and the production process; the production process was not flexible enough to successfully implement the marketing strategy. This chain of asking “why? why? why?” is the challenge for the students. They will find that, based on financial measures alone, the cause cannot be identified; they must use their understanding of operations management and operating measures to understand the root cause of the effect on financial statements.

Discussion Questions

1. Complete the income statement analysis that Lincoln Smith began by projecting the growth in each expense category and computing the variance from actual. Which factors contributed to the decline in profits?

2. What did Lincoln Smith find when he computed the direct and indirect labor productivity figures?

3. Identify and describe the root cause of the decline in operating profits. Recommend a plan of action for PlasTech.

Analysis

Question 1.

The FY93 Pro Forma income statement with a variance analysis included is shown below.

1 Adapted from Leschke, John, “PlasTech, Inc - The Importance of Matching Production Capabilities and Market Requirements”, Production and Inventory Management Journal, Third Quarter, 1995 and Darden Teaching Note UVA-OM-0794TN, 1995.

C-13

Fy 93 Pro Forma 2 Variance Analysis 3 Price $.85Production (lbs) 5,069,173 %Sales $4,308,797 100.0 (25,346) PriceCost of Good Sold: Non-consigned mtl $258,276 $1,519 Matl’s Direct labor 638,289 (153,711) Direct L Indirect labor 223,643 1,316 Indirect Utilities 765,476 2,503 Utilities Depreciation 295,000 2,180,684 50.6 0 (148,373) Dep’r Gross Profit 2,128,113 49.4 (173,719)Selling Expenses $991,023 (208,343) Selling EGeneral & Admin 777,586 1,768,609 99,127 (109,216) Admin E Operating profit $359,505 8.3 (282,935) Taxes 82,318 37,321 Tax ENet Profit After Taxes $277,187 6.4 (245,704) Net Prof

The table shows a breakdown of the decline in profits into the component variances. The data indicate that PlasTech might have realized a net profit of $277,187 or 6.4% of sales had the expenses held to 1992 proportions. Clearly, the economies of scale were not achieved for the increase in volume. The variance analysis shows that the drop in average price per pound accounted for only a fraction of the fall in operating profit. The $.005 per pound price differential multiplied by the total number of pounds produced in fiscal 1993 indicates that the sale price variance accounts for only $25,346 of the decline. Since non-consigned material prices and wages did not increase, material and indirect labor costs increased in rough proportion to sales. Less than expected utilities accounted for several thousand dollars in additional profits. Decreases in proportion of sales going to depreciation and general and administrative expenses contributed positively to operating profit. The two significant negative variances were associated with selling expenses and direct labor. The disproportionate rise in selling expenses can be explained by the greater effort needed to increase sales volume and expand the customer base. The disproportionate increase in direct labor is less easily explained.

Question 2.

Why was a 50% increase in direct labor needed to support a 20% increase in sales? Since there was no change in wages it would appear, on the surface, that direct labor productivity went down. However, the source of this decline is not found in the workers being less efficient on either productive time or setup time; both productivity measured in terms of production rate and productivity measured in time per setup improved in 1993. Rather, it is how management utilized the operator’s time that reduced direct labor productivity.

2 Calculations are based on expenses increasing in proportion to production (lbs) from FY92 to FY933 The difference between FY93 Proforma and FY93 actual

C-14

The new marketing strategy, which increased the number of customers and total sales, also resulted in more setups. Therefore, even though time per setup went down, the total time spent on setup went up. In order to provide the additional labor required for these setups (and to cover the added labor required for production), management added a third shift. For this 50% increase in labor costs, management did not receive a corresponding 50% increase in sales, hence costs grew faster than sales. This fact and its effect can be traced to the bottom line.

Question 3.

The problems encountered at PlasTech could have been anticipated with an understanding of a fundamental operations management concept -- the product-process matrix. PlasTech’s change in marketing strategy moved the company significantly to the left on the product dimension, top of the matrix, toward the lower volume and higher variety side of the matrix. The improvements in setup time made by manufacturing moved the company slightly up on the process side of the matrix. As a result, PlasTech moved away from a position of strategic balance between its marketing requirements and manufacturing capabilities. The effects of such a mismatch, as shown in the case, were lower overall productivity and lower profits.

What should PlasTech do now? There are two basic courses of action available. (1) PlasTech can choose to move to the right on the product variety/volume dimension by working with its customers to reduce the variety of products, consolidating orders into larger production runs, or setting policies to discourage small orders (e.g., quantity discounts or charging a standard setup fee). However, the success of this strategy depends on how much power PlasTech has to influence the market or to limit the type of customers selected. (2) PlasTech can choose to mover further up the process flexibility dimension by continuing to reduce setup times and increase production rates. Simply adding new production capacity will not help because it adds to fixed costs while not making the firm more efficient in its current environment.

In this case it is not clear that any one function is at fault; manufacturing has done a better job on a number of dimensions, and marketing has been successful in maintaining revenue levels. However, their independent successes led to a difficult period for the company. Thus top management is ultimately responsible for ensuring marketing and manufacturing remain strategically coordinated.

Teaching Strategy

In teaching this case I start with the variance analysis. I usually ask the student’s to explain why profits have gone down. Just in case the students do not get the point that selling expenses and direct labor have gone up dramatically, I have an overhead copy of the above table on hand to explain the variances encountered.

You can then continue to ask the why? why? why? questions. This should lead the students eventually to a point that the lots have gotten smaller and setup time has increased. A further why should lead them to the product-process matrix, assuming that you have already covered Chapter 4 in the text. At this point you can have the students explain what has happened in terms of movement off the diagonal.

C-15

Basically, the answer to the question about a plan of action is to change either marketing, operations or both to move back on the diagonal. If this has the effect of reducing costs without reducing sales, profits will improve next year. Also look for suggested actions that increase revenues without increasing costs.

This is a very good case illustrating that students must get beyond the financial numbers to understand what marketing and operations must do to solve the problem. The product-process matrix is nicely illustrated by this case.

It usually takes about one hour to teach this case.

C-16

SOUTHWEST AIRLINES: SINGIN’ THE (JET) BLUES

Southwest Airlines is a Darden case study. Darden does not permit reprinting of their teaching notes. The teaching note, if available, can be obtained by contacting http://www.darden.virginia.edu/. Then click on Darden publishing.

Case number: UVA-OM-1150

C-17

The Field Service Division of DMI4

Teaching Note

Synopsis and Purpose

This case was written about the field service division of one of the larger Fortune 100 company. The case provides a good basis for discussion for several related issues including:

1. Business process re-engineering for a service process.

2. Use of technology (cellular phones, artificial intelligence, information technology, etc.).

3. Service quality related to "gaps" between expectations and delivery.

4. Service operations strategy

• Service guarantees.• Service recovery systems.• Market segmentation.

5. Management of systems with queues.

6. Learning organizations -- and how field service can be used to help the organization keep its "ear to the ground" and use this information to give feedback to R&D.

Suggested Readings

Hammer, Michael, "Reengineering Work: Don't Automate, Obliterate," Harvard Business Review, July-August, 1990.

"The Gold Mine of Data in Customer Service," Business Week, March 21, 1994, pp. 113-114.

4    ? 1999 Arthur V. Hill. This case note was written by Professor Arthur V. Hill, Carlson School of Management, University of Minnesota. The company is real, but many names and details have been disguised to maintain anonymity for the company. Reprinted with Permission.

C-18

Answers to the discussion questions follow.

1. Draw the process flow chart for a service call. Where are the queues and delays in this system and what could be done to eliminate them?

There are two somewhat different processes:

a. Customer "phone in" process.

b. The Tech "phone in" process.

These can be drawn as a single process or as two separate processes.

There are MANY queues and delays in this system. Most students miss most of these.

The process is as follows:

Machine failsCustomer finds failureCustomer calls NSC to report failureQueue for call takers to answer the phoneCall taker answers the phoneCall taker takes model and serial numberCall taker confirms other information (phone number address, etc.)Call taker asks some questions to try to fix the machine over the phoneIf fixed, then done.If not fixed then make an ETA (expected time of arrival) usually about 4 hoursSay good-bye.

Send the service call electronically to regional dispatching centerQueue for the printerPrintedWait for dispatcher to pick up and distribute Dispatcher puts on the dispatch board under the tech's nameQueue for tech to call inTech calls inQueue for dispatcher's phone lineDispatcher answers the phoneTech "clears" old service callsDispatcher gives the tech a new service call

C-19

Tech calls the customer and makes a new ETA (expected time of arrival) Tech tried to repair the machine over the phoneTech begins travel to the customer siteTech arrives and begins repairTech diagnoses the problemIf parts are available, then tech fixesIf no parts, then tech "clears the calls" and orders parts from the dispatcherTech says good-bye to customer.

Note: When the needed parts are not available, the tech tells the dispatcher about this when the call is cleared. The dispatcher has the parts sent via Federal Express from the Service Parts Warehouse directly to the customer so that they would be available early the next morning. When the parts have been sent, the computer system automatically generates another service call for the same tech.

Be sure that the students identify ALL of the queues and the waits including those that have the people waiting for someone to answer the phone and people waiting for the computer.

I usually do not get too concerned about the students using particular symbols here.

It is particularly interesting from a systems design perspective to have the students identify the information needs for each step. This case could then be tied into a information systems design case.

2. How might the process be re-engineered? Consider some technologies that might be available to help here.

The means of reducing the mean queue time is to either:

a. Reduce the variability of the arrival process or the service process.

b. Increase the capacity (mean service rate) of the server. This is the same as saying we need to reduce the mean service time -- which can sometimes be accomplished by eliminating non-valued activities. We can also change the process technology to a system that is inherently faster.

Lots of technologies might be considered.

- Expert systems -- for diagnosis. (Also could use other advanced information technologies such as "problem/cause" database.)

- Expert systems or operations research systems -- for dispatching.

C-20

- Cellular phones -- for managing expectations.

- Beepers -- for paging techs.

- Remote RF (radio frequency) terminals -- for dispatching, clearing, parts ordering, etc.

- Satellite location systems -- to be able to know where all techs are at any point in time. (Big brother is watching!)

- How about a totally decentralized system with only an answering machine for each tech! (No control ... loss of opportunities to learn.)

3. Should DMI/FSD consolidate the regional dispatch centers into one location?

Yes. Three strong reasons:

a. If you can have the center handle a larger volume of calls the variability during the day will be less.

b. A centralized center will level the load during the day because it would cover 5 time zones.

c. It will also make it easier for the company for training, application of information systems technology, etc.

The five smaller centers could easily have one or more servers idle even when the system has many customers waiting.

It is true that the dispatchers will be farther from the customer and therefore may not know the customer's areas as well -- but given that the areas are already so large, the dispatchers are already far from most of their customers.

You can still have the dispatchers assigned to a particular area of the country even though we have only one national dispatching center.

4. Evaluate DMI's service guarantee. How could this be improved?

The guarantee is very weak and vague. There is no consequence to the company if the guarantee is not satisfied. (The guarantee has no "teeth" to it.) They are clearly not satisfying the guarantee in many cases.

C-21

They should put some teeth into the guarantee and use it for improving both marketing and operations. I would recommend that they define the service guarantee in terms of response time and extend the service guarantee by a full month if they are ever late on any agreement. They should define the guaranteed response time as a function of the service contract price -- in other words they should charge more for a shorter response time.

However, given that they are making only 80% of the service calls within the quoted time, they will probably have to either improve their quotes or improve their performance BEFORE they offer the service guarantee.

5. Why does DMI/FSD need to measure field service performance? How should performance be measured?

Performance should NOT be measured only against the arbitrary target response time. There is no motivation to complete the service once the target has been missed.

It should be measured against customer requirements. But what are the customer's requirements? Maybe they should do some market research to find out.

They should probably implement a carefully constructed customer satisfaction survey.

Remember that not all customers will have the same needs or expectations.

A service guarantee would help here to make the expectations more clearly defined for everybody.

6. What are the strategic issues for the division and the company?

a. Density is a big issue. If they continue to lose market share, their density will continue to decline and response time will continue to degrade and the cycle will continue until they are out of business.

b. Should they take on third party service -- i.e. perform service for other companies? Can they? They seem pretty weak from a management perspective.

c. Should they subcontract the business to a third-party service provider. They are not well managed -- BUT they really need to keep this function in-house in order to be a learning organization! This is one of the big points of the case! Whirlpool in the USA learned this the hard way.

d. Can they afford to have an "account-representative" relationship between techs and customers in order to build customer loyalty?

C-22

7. How could this organization become more of a learning organization?

They MUST learn from their mistakes and from customer interaction and to feed this information back to ...

· Product development -- MTBF, new functions and features

· Manufacturing -- conformance quality

· Marketing -- did we sell the customer the wrong machine?

· Diagnosis and repair -- problem/cause data base.

· Customers -- about 4 visits per year!

The technicians see each customer machine about 4 times per year. The sales person might not see the machine or the customer but once very 4 years!

8. Prepare an action plan to recommend to DMI/FSD management. Be prepared to present this to the rest of the class.

a. Consolidate the dispatchers in Denver.

b. Combine the call taking and dispatching jobs into one in order to reduce the amount of information that is lost.

c. Have "account representative" techs for all major accounts (80% of the business).

d. Design service agreements that have guaranteed response times in them where the customers pay for better response times. Carefully price these so that the company increases revenues. Have some teeth in the guarantee (e.g. extended guarantee by 1 week if we are late).

e. Stop quoting the standard response time when the call is received. Instead promise only that a tech will call back as soon as possible and then have an internal standard that a tech will call the customer back in 15 minutes. Or possibly promise that the tech will call the customer back in 15-30 minutes.

f. Use cellular phones for communications with techs. Have the techs call the customers within 30 minutes of the time that the service call was received so that the customer "feels" in the system. (Davis and Maggard article in Decision Science suggests that in 2-stage service systems, the first stage wait is more important than the second stage wait.)

C-23

g. Use artificial intelligence procedures to allow techs to be "trained" on more machines.

h. Have machines that report their own failures and allow for remote diagnostic checking so that the tech can bring the right parts.

i. Work harder to become a learning organization -- the real answer to this case is to design equipment that doesn't fail. The field service people should be feeding back information to the R&D people ... and work themselves out of a job.

Teaching Strategy

When I teach this case, I try to constantly urge students to think in terms of service and the customer. This is done by beginning with the process flow chart and the associated analysis of queues. The idea of service is also reinforced through the discussion of the service guarantee, performance measurement and the service strategy.

This case can also be used to illustrate capacity decisions in the service industry. Not only must the level of capacity be considered, but location of capacity, as well. adequate capacity is , of course, the key to providing desired service levels.

C-24

CUSTOMER-DRIVEN LEARNING AT RADISSION HOTELS WORLDWIDE

Teaching Notes

Synopsis and Purpose

Radisson hotels is a case dealing with service quality and customer satisfaction. A central question in the case is whether Radisson should offer a service guarantee to it’s customers and if so what type of guarantee? The case presents a situation where the identity of its brand is unclear and confused due to rapid growth and lack of focus – “growth at all costs”. Management now wishes to stress service quality and customer satisfaction not only through the service guarantee, but also through employee training, employee motivation and information systems. The objective of Radisson is to become a “customer-driven learning organization.”

The purpose of this case is to address service quality improvement, service guarantees and the role of employees and information system in achieving improved quality and customer satisfaction. Since students may not be familiar with these concepts the case provides a context for discussion and learning about improvement in a service environment.

Discussion Questions (taken directly from the case)

1. How should Radisson define and implement their service guarantee?

2. What role should information technology play in accelerating the drive to improve service quality?

3. How should Radisson measure and improve customer satisfaction and employee satisfaction?

4. How should Radisson drive commitment to service quality through their franchise organization?

5. How should Radisson align the goals of the hotel management team, hotel workers, owners, corporate management and corporate staff with their new brand strategy?

Analysis

1. How should Radisson define and implement their service guarantee?

One issue in the case is whether they should have a one-step or two-step service guarantee? A one-step guarantee is to simply pay the customer when there is a service failure. A two-step service guarantee is “we will make it right or we pay”. In the two-step guarantee there is an opportunity to correct the service failure before paying.

C-25

Discussion of this question is facilitated by recognizing that there are many types of service failures such as:

1. Lack of towels in the room.2. The room is not ready on time.3. The room is too hot or too cold.4. The guest didn’t sleep well because there was a noisy all-night party next door.

The first failure might not require payment at all, the guest may simply want the missing towels delivered on a timely basis. The second failure may be more serious, because it could disrupt the guest’s plans and schedule. In this case one guest may want to be paid, another guest might want a different room temporarily, and a third guest may not care and can come back later to check in. So, the severity of the service failure and the guest’s expectations both enter into whether the hotel should have a one-step or two-step guarantee.

Implementation of the service guarantee will require extensive employee training. The training can be done by either Radisson personnel or outside consultants. In either case training will be needed to explain the service guarantee to all employees. In the hotel business turnover of employees is very high, approaching 100% per year in some cases. This will require continuous training of new employees.

Franchisees and building investors own most of the Radisson hotels. Radisson manages these hotels for a franchise fee. Therefore, Radisson may need to provide some type of risk coverage to get the hotel management to accept the service guarantee. For example, Radisson could pay for the service guarantee for the first year and the hotel management could pick up the costs in the second and subsequent years.

2. What role should information technology play in accelerating the drive to improve service quality?

Information technology plays a very critical role. Quality measurements such as customer satisfaction, customer complaints, and customer return visits should be taken. These measurements should be compared across hotels in the Radisson chain and over time. The information system should provide periodic reports to management of the various quality measures that are deemed important.

Since they don’t own most of the hotels, Radisson can be viewed as a big computer system and a brand. The information system takes care of reservations and provides call center management. The value that the Radisson Corporation provides to franchisees is advertising and brand awareness together with ease of making reservations and calls via a central system.

The information system at Radisson should be designed to provide a “learning loop” starting with understanding customer needs and wants, and finally monitoring behavior and adjusting as needed. The system should also maintain a “guest profile” based on every touch point in the system (marketing, dinning, guest room, key account management, etc.). The customer touch points can be equated to customer preferences and “moments of truth”. The HARMONY Product Suite and Customer KARE systems will measure local service delivery parameters and provide information to hotel management.

C-26

3. How should Radisson measure and improve customer satisfaction and employee satisfaction?

Radisson will need to get a good response rate on customer surveys to insure that they don’t have non-respondent bias in the results. While many surveys report only a 10 to 20% response rate, this is too low for reliable results.

Using guest comment cards alone is very unreliable since the response rate is often very low and the sample can be extremely biased; only those who are dissatisfied complete cards. Therefore, it is best if Radisson hires an outside firm such as Gallop to collect data on customer satisfaction.

In addition to random surveys of satisfaction, Radisson should collect all complaints from customers even though the complaints may not be representative of the population of guests. Complaints from customers should be handled promptly and courteously. Quick service recovery is a key to satisfying customers. After handling the complaint to the customer’s satisfaction, Radisson should seek to fix the system to prevent future failures of this type.

It is ironic that a rising level of customer complaints is not as bad as it seems. Radisson can’t learn from the customer if the customer doesn’t complain about the service. So, if management wishes to reduce complaints they will also reduce learning. At the same time, Radisson doesn’t want the same complaint for the same reason over and over again. The objective should be to reduce the same type of repetitive complaints, not necessarily the total number of complaints.

Employee satisfaction can be improved in a number of ways. Removing obstacles that prevent employees from serving the customer does this. Obstacles include being too busy, computer systems that don’t work, poor supervision, unclear job descriptions, etc.

Positive feedback and reinforcement for a job well done can also increase employee satisfaction. There are a variety of perks and ways of recognizing employees.

There is an opportunity in this case to emphasize the service-profit chain. It starts with employee satisfaction leading to customer satisfaction/loyalty and ultimately to higher profits.

The case indicates that Radisson was concentrating on franchisees as their customers, not the hotel guests. A focus on the end customer and employees is needed to improve customer satisfaction in the hotels.

C-27

4. How should Radisson drive commitment to service quality through their franchise organization?

They should use both the carrot and the stick approach. The carrots include:

Provide franchisees with comparisons of other hotels and best practices workshops.

Give financial incentives for superior service.

Provide management with means for improvement (e.g. measurements, consultants, programs for improvement, etc.)

Show the connection between service quality and profit. Many franchisees may only see the costs of improvement and not the benefits.

The “stick” could be removal of the franchise if quality standards are not met after some probationary period.

5. How should Radisson align the goals of the hotel management team, hotel workers, owners, corporate management and corporate staff with their new brand strategy?

Alignment of goals is a top management responsibility that cannot be delegated. If goals are not aligned, top management should “look in the mirror”.

Communication – communication – communication. There is no substitute.

Use training programs and meetings to insure that goals are clearly understood.

Try to reduce employee turnover. It may take a certain amount of time to absorb goals and fully understand them. Employee turnover destroys teamwork and goal understanding.

Benchmark the Ritz Carlton Hotels, a two-time Baldrige winner. There are a number of videos available on the Ritz Carlton from the Baldrige Award office.

Teaching Strategy

This case can be taught in the same order as the discussion questions. It is helpful to emphasize the service quality and the service guarantee in this case, since it is one of the central ideas.

I like to go through the discussion questions one at a time and then tell the students what happened in the actual situation. Here is what happened in this case5.

5 For more details see, Hill, Arthur, et.al., “Service guarantees and strategic service quality performance metrics at Radisson hotels worldwide,” Journal of Strategic Performance Measurement, Dec 1998, pp. 27-31.

C-28

Radisson implemented the two-step service guarantee. The guarantee read simply, “Our goal at Radisson is 100% guest satisfaction. If you are not satisfied with something, please let us know, we’ll make it right or you won’t pay.” Management felt based on focus group feedback that the customers weren’t comfortable with simply getting their money back in a one-step guarantee. In many cases what the customer wanted was a quick resolution of the problem.

They chose internal trainers to teach all of the employees the new “100% guest satisfaction” program with service guarantee. They spent over $1million the first year on training.

Radisson also agreed to pay the entire cost of invocations the first year and after that franchises would pay for service failure. The first year cost was $10 million. But hotel utilization and customer satisfaction went up, employee turnover went down, and profitability improved so Radisson management felt it was money well spent.

The service guarantee was more important for the employees than the customers. This is a very interesting point. Marketing often thinks that a service guarantee will bring in more business which it did in this case. But, it also defined the mission of the employees as 100% customer satisfaction and gave management a metric for service failure (the cost of invocations).

There was a discussion among management as to whether the hotels should improve quality first before offering the service guarantee. Often improving quality first is a good course of action as the costs of invocation can be too high. Radisson decided to implement the service guarantee as a way of “forcing” better quality. They didn’t want to wait to improve quality first.

Radisson hired the Gallop Corporation to do their surveys of customer satisfaction. Hiring an outside firm was the only way to get reliable data for comparison among hotels and over time.

By implementing a service guarantee it was clear to everyone that Radisson was selling to the guests, not to the franchisees. Service to the ultimate customer was emphasized.

A service guarantee can be a flash in the pan. Management must be careful to maintain continuity of the service guarantee once it has been adopted. This can be done by continual training of new employees, reporting on results achieved and benchmarking demonstrations to hotels that are failing in using the service guarantee.

Lessons

I like to close this case with lessons that I put on the board.

Radisson is becoming a learning organization by using knowledge of the customer to improve customer service and quality.

The case demonstrates how to structure and implement a service guarantee

The case shows that strategy must guide actions. At Radisson the strategy was 100% customer satisfaction and the service guarantee was used to accomplish the strategy.

The service profit chain is illustrated by this case. In order to increase customer satisfaction it is important to improve employee satisfaction. The two are linked.

C-29

Federal Express6

Teaching Note

Synopsis

The Federal Express (FedEx) case looks closely at the philosophy behind the most successful overnight courier in North America and the quality initiatives that naturally complement this ideology. As a result of its quality efforts, Federal Express won the Malcolm Baldrige National Quality Award (MBNQA) in 1990. The case shows that the theme of FedEx’s quality system targeted both customer satisfaction and FedEx’s employees. The case examines each of the elements of the FedEx quality system and the technologies put in place in support of the customer and employee satisfaction goals. Furthermore, the case details the measurement systems used to track the success of these systems. The case does not pose any obvious issue questions. It looks at the FedEx quality system at the time of the awarding of the Malcolm Baldrige Award, the concludes with a hint at future growth for FedEx, challenging the reader to anticipate how the quality system will have to evolve to keep FedEx on top while continuing to meet the growing needs of both its customers and employees. The primary teaching emphasis is understanding the design feature of the management system and evaluating how the design element interacts and supports the FedEx philosophy.

Use of the Case and Teaching Objectives

This case is well suited for either a required course in operations and management or an elective course on quality management, service operations, or managing organizational improvement. It is particularly useful in introducing students to the management challenges that might face service organizations.

The case

• gives an illustration of a service company that has developed a powerful set of supporting systems and technologies to support its overall corporate philosophy;

• engages students in a critical evaluation of the overall management system of a service organization;

• provides an example of a company that requires continuous improvement of not only its service, but also its people;

6 This teaching note was prepared by Laurence 0. Mueller under the supervision of Professor Robert D. Landel,Henry E. McWane Professor of Business Administration. Copyright 1993 by the University of Virginia DardenSchool Foundation, Charlottesville, VA. All rights reserved. Rev. 5/95. Reprinted with Permission (UVA-OM0721 TN).

C-30

• encourages students to compare FedEx practices to those of other service companies, emphasizing contextual differences and their influence on system evolution;

• invites students to consider the future issues that will affect the evolution of FedEx’s management system.

Teaching Approach

If Federal Express is the first case to be taught as part of a segment on total quality management (TQM), the following list enumerates several class-discussion areas:

1. Present/discuss the Malcolm Baldrige Award, including the concept behind it, its judging criteria, its goal, and the value underlying its sharing requirement. Other quality benchmarks and awards, such as ISO 9000 and the Deming Prize, might also be mentioned and their differences pointed out.

2. Discuss the importance of companies having a quality philosophy as part of their TQM system. Use Federal Express’s quality philosophy as an example, studying how the company encourages and supports “100% customer service,” continuous improvement, “Q = P” (quality productivity), and a “people first” perspective.

3. More generally, explain and consider quality-management systems, the particular elements of total quality management, and the content of world-class quality systems, using the slides of case Exhibits 1-8. These exhibits may be used as the basis for an introductory/overview lecture.

If the Federal Express case succeeds several other quality-management cases, then the following points might be more appropriate discussion topics.

1. Compare FedEx’s TQM system to that of other quality-management systems that the students might have just studied. Focus closely on the comparison of corporate culture and its role, quality philosophies, and stage of development. Cases that are particularly good comparisons to Federal Express include “Florida Power & Light” and “Paul Revere Insurance” (Harvard cases), as well as “USPS” and “Telenet” (Darden cases).

2. Examine the details of the Malcolm Baldrige Award criteria and how they apply to Federal Express’s TQM system.

C-31

Questions for Discussion (These questions are somewhat different than the questions in the text, but cover the same general topics)

The following list of assignment questions may be helpful in guiding student preparation and structuring class discussion. Below each question are listed potential responses that could be raised in class discussion.

1. Describe FedEx’s corporate philosophy and examine how FedEx’s systems fit in the context of its corporate philosophy.

• The core of Federal Express’s corporate philosophy is that quality = people. If Federal Express takes care of its employees, then its employees will take care of the customers. Federal express works to achieve 100 percent satisfaction.

• The Quality Improvement Process and Quality Action Teams (QAT) were developed to allow and encourage employees to ferret out problems and create their own solutions.

• Service Quality Indicators (SQl) give employees ten measurements of service quality as experienced by the customer. The SQl system shows a weighted, composite score of failures to meet promises. The importance is that this information is readily available to all employees, allowing everyone on a daily basis to see the direct effects of any changes they make.

• Customer/Supplier Alignment gives the employees of one department the ability to improve how they are being served by another department through coordinated communication about the most important services. This process creates an internal customer focus in addition to the external focus.

• Survey/Feedback/Action (SFA) and the guaranteed fair treatment procedure (GFTP) are ways for employees to bring concerns to management and be guaranteed a quick response. The SFA process keeps management on top of employee concerns and needs, thus focusing management on taking care of employee so that they, in turn, will care for the customers. The leadership-score link to managerial bonuses is an internal customer measurement that relates quality to people.

• Finally, LEAP, or leadership evaluation awareness process, provides employees a way to move up within the company through their own efforts, with support of management. This is a reward system that allows employees to see how their efforts in customer service can propel them up the leadership path.

C-32

2. How has FedEx used information technology to provide 100% customer service?

• COSMOS (Customer, Operations, Service, Master On-line System) is a sophisticated tracking system that allows employees to pinpoint a package’s current location, estimate its time of delivery, indicate any possible delays, and redirect the package enroute, if necessary.

• DADS (Digitally Assisted Dispatch System) is an on-screen communication system that link courier vans to COSMOS, listing pickups and deliveries and reordering them into the most efficient schedule when additions to or removals from the schedule occur.

• Powership 2 is a direct benefit to the external customer because it links them directly into COSMOS, allowing them to call for pickups directly, track packages, and confirm deliveries.

3. What are the key factors that have contributed to the success of FedEx’s quality-management program?

• Because employees are included in decision making from the start, they “buy in” to the quality program from the start. New company, new culture.

• Empowering the employees to form their own QAT and make their own solutions reinforces the concept that management supports employees so that the employees can better serve customers.

• The constant and up-to-date communication to all employees through FedExTV provides immediate performance measurements each day as well as competition reports, all of which drive employees to continue to improve quality.

• Guaranteeing no lay off, except under extraordinary circumstances, provides stability that fosters the quality program.

4. What challenges does FedEx face in the future? How will its management system need to evolve to meet these challenges?

• Slowing domestic growth.

• Meeting the international market and selling the FedEx way of business to businesses overseas.

• Increased flexibility to meet foreign demands.

• Increased domestic opportunities as FedEx increases inventory management systems for additional customers.

C-33

• Cross-training employees to prepare them for the constant influx of technology that is likely to reduce the number of handlers needed.

Conclusion

Key learning points from the case are as follows:

“The road to the MBNQA winner’s circle has arrows pointed one way.” This statement is a response to the idea that an economic downturn may force FedEx to curtail some if its quality programs: FedEx recognizes that investments in quality are valuable to its associates now and to its associates and customers in the long haul.

“FedEx views its job as selling service.” The idea that FedEx is selling-not just providing-service shows its commitment to being the best, taking the best to the customer, and meeting the customer’s needs, rather than waiting for the customer to come FedEx to provide the service.

“Our people hold in their hands our customers’ perception of quality.” This statementshows the value FedEx places on its employees; they are its number-one resource. In addition, the quote clearly expresses that it is not the physical apparatus nor the computer-technology systems that convey the perception of quality to the customer, but rather the employees’ capable interactions with customers.

“Quality action teams organize when employees see a need to change the way they do their jobs.” This declaration indicates that the quality teams are self-forming and that empowerment is encouraged and supported. Moreover, it an example of continuous improvement because the impetus for change is encouraged to come from the associates themselves.

Focus, analyze, develop, execute (FADE). The problem-solving framework all associates are trained to use is QAT. This is another tool that reinforces the continuous-improvement process.

Service Quality Indicators (SQI). The measurement system by which customer satisfaction is measured on a broad-based level” points reflect the degree of FedEx’s performance failure.

Internal Customer/Supplier Alignment. A process by which different departments work together, recognizing each other as customers and making internal customer satisfaction a key to the smoothness of internal operations.

The FedEx family. The corporate atmosphere fostered by avoiding layoffs, the GFTPdrawing leadership from within through the LEAP program, and the interactive SFA covenant between management and employees.

COSMOS and DADS. Information technology available within the company as a resource to associates in serving their customers.

C-34

Powership 2. An information-technology system designed particularly for customers so that they may interact with FedEx technology and “keep up” with their own parcels.

“Change is a matter of survival.” FedEx recognizes it must adjust constantly to the environment, anticipating both its customers and employees needs in order to stay on top.

C-35

BAYFIELD MUD COMPANY

Teaching Note

Synopsis and Purpose

Bayfield Mud Company is a manufacturer of mud treating agents used for oil and natural gas drilling operations. The company has recently shipped several box cars of 50 pound bags of mud treating agent which were light in weight. The customer, who received the shipments, has threatened to find another supplier if the situation is not corrected.

Data is presented in the case from several production shifts. This data can be used to determine the probable cause for bags which do not weigh the proper amounts.

The purpose of this case is to provide a problem in statistical quality control. Students must apply control chart principles to find the cause of the problem and they are asked to recommend corrective action.

Discussion Questions

1. What is your analysis of the bag weights problem?

2. What procedures would you recommend to maintain proper quality control?

Analysis

Data is provided on the Excel Spreadsheet supplied with the text. This spreadsheet contains the data only, no formulas or analysis is provided for the student. The analysis can be done by hand, but it is very tedious.

The first step in the analysis is to construct control charts for the three shifts. These charts can be made for all of the data or for each of the shifts separately. Since the case indicates that the evening and night shifts may be having more problems with bag weights, because of training and inexperience, we will construct control charts separately for each of the three shifts.

We have constructed both average and range charts from the data. The first average chart is developed by using the day shift averages given in the case. Three day shifts consisting of 24 data points are given. These 24 points are averaged to obtain a grand average of 49.80. Likewise, the evening and night shift averages are averaged to obtain grand averages of 48.94 and 48.65 as shown in Exhibit 1. The ranges for each of the samples is also computed. These ranges are averaged by shift to arrive at average ranges of 3.45, 4.38, and 3.36 for the day shift, evening shift, and night shift respectively. We have now computed the center lines for each of the control charts.

C-36

The upper and lower control limits for the average and range charts are computed in the usual fashion. Using a sample size of n = 24, the average and range chart control limits are shown on the bottom of Exhibit 1. We can now determine how many points from the sample data are out of control. The numbers are as follows:

Number of Points out of ControlDay Shift Eve Shift Night Shift

Average Chart 10 21 17Range Chart 1 2 5

Now, we have some interesting data to examine. On the day shift, the range chart appears to be in pretty good control (only one point is out of control in 24 points, which could happen by chance alone). The average for the first shift is badly out or control (10 points out of 24 are out of control). We must therefore conclude that the process is not stable for producing average bag weight on the first shift.

The situation is even worse on the evening shift. The range chart now has two points out of control, which is a little more questionable. The average chart is totally out of control with 21 out of 24 points out of control. On the night shift the range appears to be out of control, as well as the average.

We also note from the data that the averages are well below 50 pounds per bag on the evening shift and the night shift. Therefore, it looks like the bags are indeed being under filled. The first order of business, however, is to bring the process under control. Then the average can be set to the proper level of 50 pounds per bag.

On the day and evening shift, we can probably accept the range as being under reasonable control and look for assignable causes which would cause the average to be out of control. The case indicates that the bag weighing mechanism may not be checked often enough. The mechanism may also be unreliable in weighing the bags. This mechanism may have to be repaired or replaced. In addition, we could look more closely at training, to be sure all personal understand the weighing procedures and are following them. Whatever, the causes of variation, they must be uncovered and removed. Then the process will be brought under control. We can be confident of this, since the case indicates that the process has been stable in the past.

On the night shift the problem is a little more serious. In this case we must bring the range under control first. We need to seek causes of excessive variation. These causes may be the same as those in the day and evening shifts, or there may be additional causes of variation. Once the range is brought under control, then the average on the night shift should also be brought into statistical control.

Once the causes of variation are uncovered and corrected, control charts can be used on a routine basis for maintenance of statistical control. New data would have to be collected to establish the average and range to use for purposes of control. It would also be best to establish a single average and range chart that could be used for all three shifts. There is no reason that each shift should have its own control chart for control purposes, since the same equipment and procedures are being used.

C-37

The average control chart should have an average of 50, since we want the bags to have 50 pounds. It could also have upper and lower control limits based on the process standard deviation which is 1.2 pounds as given in the case. For samples of 24 bags, the upper and lower control limits for the average control chart are as follows:

Average Chart

Upper Control Limit = 50 + 3 (1.2/ ) = 50.37

Lower Control Limit = 50 - 3 (1.2/ ) = 49.27

Similar limits could be computed for the range chart as well. If these charts were used, then process deviations could be detected and the short weight problem could be prevented in the future. It is also possible to base the charts on average sample data instead of process averages.

Teaching Strategy

When teaching this case I start by asking a student to conduct a statistical analysis of the data. Other students are then asked to comment on the analysis until several variations and assumptions have been discussed.

When the data has been thoroughly analyzed, I then turn to a discussion of how the situation can be corrected. Some students will miss the point that the process must first be brought under control before statistical control charts can be used. A great deal of discussion can also be generated about the use of control charts.

It usually takes about 45 minutes to teach this case.

C-38

EXHIBIT 1

Average RangeWeight

Time Shift (pounds) Smallest Largest Difference

6 A.M. Day I 49.6 48.7 50.7 2 7 50.2 49.1 51.2 2.1 8 50.6 49.6 51.4 1.8 9 50.8 50.2 51.8 1.6

10 49.9 49.2 52.3 3.1 11 50.3 48.6 51.7 3.1

12 NOON 48.6 46.2 50.4 4.2 1 P.M. 49.0 46.4 50.0 3.6

2 Evening I 49.0 46.0 50.6 4.6 3 49.8 48.2 50.8 2.6 4 50.3 49.2 52.7 3.5 5 51.4 50.0 55.3 5.3 6 51.6 49.2 54.7 5.5 7 51.8 50.0 55.6 5.6 8 51.0 48.6 53.2 4.6 9 50.5 49.4 52.4 3

10 Night I 49.2 46.1 50.7 4.6 11 49.0 46.3 50.8 4.5

12 MIDNIGHT 48.4 45.4 50.2 4.8 1 A.M. 47.6 44.3 49.7 5.4

2 47.4 44.1 49.6 5.5 3 48.2 45.2 49.0 3.8 4 48.0 45.5 49.1 3.6 5 48.4 47.1 49.6 2.5 6 Day II 48.6 47.4 52.0 4.6 7 50.0 49.2 52.2 3 8 49.8 49.0 52.4 3.4 9 50.3 49.4 51.7 2.3

10 50.2 49.6 51.8 2.2 11 50.0 49.0 52.3 3.3

12 NOON 50.0 48.8 52.4 3.6 1 P.M. 50.1 49.4 53.6 4.2

2 Evening II 49.7 48.6 51.0 2.4 3 48.4 47.2 51.7 4.5 4 47.2 45.3 50.9 5.6 5 46.8 44.1 49.0 4.9 6 46.8 41.0 51.2 10.2 7 50.0 46.2 51.7 5.5 8 47.4 44.0 48.7 4.7 9 47.0 44.2 48.9 4.7

10 Night II 47.2 46.6 50.2 3.6 11 48.6 47.0 50.0 3

12 MIDNIGHT 49.8 48.2 50.4 2.2 1 A.M. 49.6 48.4 51.7 3.3

2 50.0 49.0 52.2 3.2

C-39

Exhibit 1 continued

3 50.0 49.2 50.0 0.8 4 47.2 46.3 50.5 4.2 5 47.0 44.1 49.7 5.6 6 Day III 48.4 45.0 49.0 4 7 48.8 44.8 49.7 4.9 8 49.6 48.0 51.8 3.8 9 50.0 48.1 52.7 4.6

10 51.0 48.1 55.2 7.1 11 50.4 49.5 54.1 4.6

12 NOON 50.0 48.7 50.9 2.2 1 P.M. 48.9 47.6 51.2 3.6

2 Evening III 49.8 48.4 51.0 2.6 3 49.8 48.8 50.8 2 4 50.0 49.1 50.6 1.5 5 47.8 45.2 51.2 6 6 46.4 44.0 49.7 5.7 7 46.4 44.4 50.0 5.6 8 47.2 46.6 48.9 2.3 9 48.4 47.2 49.5 2.3

10 Night III 49.2 48.1 50.7 2.6 11 48.4 47.0 50.8 3.8

12 MIDNIGHT 47.2 46.4 49.2 2.8 1 A.M. 47.4 46.8 49.0 2.2

2 48.8 47.2 51.4 4.2 3 49.6 49.0 50.6 1.6 4 51.0 50.5 51.5 1 5 50.5 50.0 51.9 1.9

Day Shift

Eve Shift

Night Shift

Grand Average 49.80 48.94 48.65 Average Range 3.45 4.38 3.36

Average ChartUpper Control Limit 50.34 49.63 49.18 Center Line 49.80 48.94 48.65 Lower Control Limit 49.25 48.25 48.13

Range ChartUpper Control Limit 5.35 6.79 5.21 Center Line 3.45 4.38 3.36 Lower Control Limit 1.56 1.98 1.52

C-40

Number of Points out of ControlAverage Chart 10 21 17Range Chart 1 2 5

C-41

SIX SIGMA AT 3M, INC.

Teaching Note

Synopsis and Purpose

The purpose of this case is to expose students to the issues involved in implementing Six Sigma in 3M, a large multinational corporation. The case tells us that the new 3M CEO and Board of Directors have already decided to implement Six Sigma. It is just a question of how the program will be rolled out, what challenges are being faced, and what benefits will be achieved.

The case describes some of the key features of the 3M Six Sigma program, since it is assumed that most students are not familiar with Six Sigma. The discussion questions then focus on understanding the potential benefits and costs to 3M, the relationship of Six Sigma to corporate strategy, and its impact on the human resources and information systems of the company.

Students should come away with not only an understanding of what Six Sigma is, but an awareness of the pitfalls and the implementation problems that might be encountered. At the end of the teaching note, we also contrast Six Sigma with the former TQM program in 3M. This provides an interesting discussion about the key differences.

Discussion Questions

1. What will be the benefits of the Six Sigma Program and how will they be tracked and reported?

Benefits will be increased earnings growth, improved quality, common language for suppliers to customers, accelerated product commercialization, and leadership development for employees.

Finance people from each division will be assigned to validate projected financial results and will be assigned to track these results when the programs are implemented. All other non-financial benefits (customer satisfaction, customer loyalty, employee satisfaction, employee turnover, innovation, etc.) will also be tracked by the Six Sigma organization.

C-42

2. What are the costs and risks of this program?

Costs are the outside consultant training fees for top management and master blackbelts. This was $25,000 to $40,000/person trained. Another cost was reallocation of 2-4% of full time resources in each functional area to become full-time blackbelts. This takes resources away from existing activities. Total costs for the program at 3M were confidential but there were a large number of top management folks trained at the $40,000/head rate.

Risks are spending large amounts of money and achieving marginal results from the improvements. For example, savings might be documented on a project basis, but won’t show up on the bottom line due to overall company loss of sales or increasing costs that are not subject to Six Sigma projects.

3. What kinds of change management skills will the organization need to implement Six Sigma?

Each manager will become a project sponsor and will be given specialized training to help manage the change process. This will include division level operating committee members and above. Eventually, all 3M managers will be trained as Six Sigma Green belts. The Green belt training includes sessions on change management as part of the course.

4. How should various functional areas in the organization be included in the Six Sigma initiative and what role should senior and middle management play in the change initiative?

All functions will be included in the Six Sigma initiative. It’s important that all areas of the company improve and get more efficient and better at meeting the needs of internal and external customers to support the corporate goals. Legal, for example may run a project to reduce the cycle time for filing patents or to reduce the variability in this process. HR may undertake a process improvement related to the recruiting and selection of new employees.

Senior management should be held accountable to support the Six Sigma initiative. Executive level blackbelts will be assigned the role of working with executives to develop Six Sigma programs to support corporate Super Y’s. Division VP’s and above will each have “dashboards” to monitor their critical “Y’s”. Targets will be given from the top down and senior management must effectively manage the change process to meet these new goals.

C-43

5. What role should Six Sigma play in corporate strategy?

Six Sigma helps the corporation define its critical business “Y’s from a high level strategic view. Then these corporate “Y’s” are fed down to the division, then to a project cluster, then to individual projects that help all add up to achieve the corporate “Y” goals. 3M will still need to have a corporate strategy that will help in complete in its various markets. Six Sigma can help improve the process of strategy formulation, but does not dictate the strategy itself. Management must choose the strategy and then use Six Sigma to implement it.

6. What kinds of information systems are required to support the Six Sigma initiative?

They need to develop several new databases to track projects. They also need to develop several new systems to measure the critical “Y’s” and produce the division level dashboards. Six Sigma is not only changing the information system of 3M, but is used in IS projects to improve their performance.

7. What are the human resource implications for deploying Six Sigma e.g. employee selection, organizational structure (Green Belt, Black Belt, Master Black Belt, Champion).

This will be a fast track for developing leadership skills and selecting the best leaders for the future. As the company begins to grow faster, more and better opportunities will exist for the people that prefer to work for a faster more aggressive company. It is important to note, that Six Sigma not only provides process improvement, but a common language for improvement across all functions and a path for the high potential employees to be directly involved as improvement specialists (Black Belts). Thus Six Sigma has a huge impact on the human resources of the company and their deployment. It is a leadership development program.

8. What are the most effective incentives for people involved with Six Sigma projects (Black Belts, team members, etc.)?

There is high visibility for Six Sigma projects and it provides opportunities for upper management to view the skills of black belts and team members. If the BB’s and team members perform well, they will receive bigger raises and faster promotions.

C-44

What has happened at 3M, an update.

Where did 3M start?

They started by appointing an Executive Director level V.P. to lead the Six Sigma program for the entire company. This person reported directly to the CEO. Then they brought in a consulting firm to train the top-level executives and first waves of master blackbelts and blackbelts. From there they defined the corporate super “Y’s” of cash, growth, and productivity. These “Y’s” were then funneled down through the divisions and functional groups to develop projects.

While key 3M employees were initially trained by consultants, 3M quickly put together its own internal training program to train the thousands of people in 3M. This training program was conducted by 3M personnel and customized to 3M.

What are the timetables for transforming the 3M organization?

The goal was to have 50% of salaried people on a global basis trained by end of 2002 with 1000 projects up and running. In 2003 they wanted 100% of salaried people trained and 2000 projects running.

How was Six Sigma different from 3M’s existing quality programs

TQM was used by many groups in 3M. TQM has most, or all, of the tools that Six Sigma has but it differs in how the corporation ran the programs. TQM was implemented by staff quality groups, targeting specific projects around the company. These folks were not always the most aggressive high achievers in the company.

Six Sigma was set up by the CEO and driven by him using the new organization made up the best people. These high achievers were selected from all functional areas and were pulled out of their existing jobs, given intense training, and then supported by top management. These master blackbelts and blackbelts work in each functional area along side of their fellow employees to solve the most critical strategic issues for the corporation as they relate to each division.

Six Sigma deploys full-time people (2 to 4%) in 3M to work on projects. Process improvement becomes an essential activity of each division in 3M, not an optional part-time effort.

3M is also tracking the savings and net income improvement from Six Sigma. This is very critical in that the skeptics will focus on the financial results or lack thereof. Significant demonstrated improvements in net income monitored by the finance organization help to demonstrate the worth of the Six Sigma approach to everyone.

C-45

Teaching Strategy

In teaching this case, I start with a review of the discussion questions. After going through these questions I find there are still many skeptics in the class about whether Six Sigma would work in their organization and whether it is the way for everyone to proceed. I point out that Six Sigma does not work well for all organizations. For example, if an organization is not capable of being ISO 9000 certified, it should probably not engage in Six Sigma. Such an organization should first get its processes defined and standardized, and people trained in the processes before attempting to improve them. Other situations where Six Sigma should not be attempted is when the organization is likely to be sold to another company, when there is imminent turmoil in their markets or the possibility of bankruptcy. Six Sigma is for the stable company that already has the basics of quality management in place and wants to move to a higher level. Finally, the CEO and senior management must support Six Sigma and have enough time to implement the changes. If there is lack of commitment or instability in top management, Six Sigma is likely to fail.

It is probably also important to note that the important part of Six Sigma is not the tools. Many organizations have trained their people in the tools and still not achieved good results. A Six Sigma program must be strategic in nature, led by the CEO, use dedicated BB’s and track the financial results to be successful. Without these organizational changes, results will be meager to negative.

This is a very enjoyable case to teach, since students have probably heard a lot about Six Sigma from the popular business press, but are not sure what it is or whether is should be used. Thus, they can learn a great deal from a company that is going about implementing Six Sigma in an aggressive manner.

C-46

FORD MOTOR COMPANY, SUPPLY CHAIN STRATEGY

The teaching note for Ford Motor Company, Supply Chain Strategy can be obtained from Harvard Business School Publishing (HBSP). Harvard Business School does not permit reprinting of their teaching notes. The teaching note can be obtained by calling 1-800-545-7685 or you can send an e-mail request to custserv@hbsp.harvard.edu. They will send you the teaching note by mail. If you are a registered faculty member with the HBSP web site (www.hbsp.harvard.edu), you can receive immediate online access to the teaching note by using your user name and password.

Teaching Note number: 601172

C-47

MERRIWELL BAG COMPANY

Teaching Note

Synopsis and Purpose

Merriwell Bag Company manufactures and distributes stock bags to many small chain stores scattered over a wide geographical area. Presently, due to growth in the business, forecasting demand has become more difficult. As a result, the company would like a forecasting system developed. Monthly data from the past five years is provided in the case.

This case presents an opportunity for the student to design a forecasting system. The case also asks the student to describe how the system can be used and aided by managerial judgment.

Discussion Questions

1. Develop and justify a forecasting method that fulfills the company's specifications.

2. Forecast aggregate demand by month for 2005.

3. In addition to forecasting demand of larger customers and aggregate demand, how might the accuracy of the forecast be improved?

4. What role should Ed Merriwell's "feel" of the market play in establishing new sales forecasts?

Analysis

Because of the seasonal nature of the demand facing Merriwell Bag Company, an appropriate forecasting tool is the classical decomposition method (discussed in the supplement to Chapter 11). The data in the case is provided on the CD-ROM that accompanies the textbook. Only the data is provided on an Excel template. The user must enter the formulas and analysis.

Sixty months of data are provided on the template, see exhibit 1. The first step in classical decomposition is to develop a 12-month moving average which is done in the 3rd column on the worksheet. Then a 2-month moving average is developed in the 4th column which is centered on the original data. The 4th column contains data which is deseasonalized, since 12 months has been used as a base in the moving average. At this point the upward trend in the moving average in column 4 is apparent.

C-48

In column 5 seasonal ratios are computed by dividing the sales data for each month by the moving average in column 4. The data indicates that high seasonal demand occurs before Christmas each year in September, October and November. Low seasonal demand occurs in the Jan, Feb and March time frame. In column 6 average seasonal ratios are computed. These ratios are obtained by averaging the seasonal ratios from the same month in successive years. For example, the July seasonal ratio is obtained by averaging the July 2000, July 01, July 02, and July 03 seasonal ratios.

When the resulting twelve seasonal ratios are added the total is 11.8977. The sum of these ratios should be 12 in accordance with the 12-month seasonal period, because the seasonal ratio is the percentage that a particular month is above or below the average. In order to obtain a sum of 12, the seasonal ratios are normalized in column 7. This is done by dividing each ratio by the sum 11.8977 and multiplying by 12.

A regression analysis is now run to fit a straight line through the moving average data in column 4. The purpose of this regression is to forecast the average level into 2005 on a trend basis. The seasonal ratios will then be applied to this trend to arrive at a forecast. In Excel a regression function is provided. In this case we have data from period 7 through period 54. The formulas and procedure for calculating the regression equation are given in the text. As a result of these calculations the following equation is obtained.

Y = 5997 + 70.24 t

Where Y is the moving average and t is the time period.

To obtain the forecast of interest we calculate Y from the above equation for the twelve months of 2005, which is t = 61 through t = 72. These Y values are multiplied by the monthly average seasonal ratios to arrive at the forecast for each month shown in Exhibit 1. Note, that the total of this forecast is 129,435 bales of bags for the year 2005.

In evaluating the forecast one of the questions that comes to mind is the validity of the linear trend assumption. Note, that total demand in 2004 (113,000) was actually a little less than the total demand for 2003 (115,000). Ed Merriwell should determine if there is some reason for this leveling out of demand or should the historical trend be assumed to resume. If demand flattens out at 115,000 bales, our forecast for 2005 could be too high by about 15,000 bales for the year (129,435 - 115,000).

The seasonal ratios appear to be pretty stable from year to year. While some monthly variation can be expected in seasonal ratios, it will probably not be as serious as the trend assumption discussed above, because the seasonal error from month to month will tend to average out over the course of 2005.

Whereas the above forecasting technique should be useful, Ed Merriwell's "feel" of the market should not be discarded. Any analytical method should be augmented by personal judgment. This judgment would prove very useful in considering mostly non-quantifiable factors that might affect demand (state of the economy, consumer attitudes, activity of competitors, etc.) These effects can be quantitatively introduced into the forecast by adjusting the future trend and possibly the individual seasonal ratios.

C-49

This case can also be analyzed by using exponential smoothing with seasonal adjustments and trend. A spreadsheet could be written using the Winter's formulas from the supplement to Chapter 11. The trend component could be derived from past data or based on judgment as to how fast demand will grow in the future.

Teaching Strategy

This case can be taught by first developing one or more forecasting techniques. Different forecasts can be put on the PowerPoint overhead and compared to each other. After all of the techniques have been described one can then turn to a discussion of the intangible factors.

This case illustrates a realistic forecasting problem. The student must deal with model choice issues and problems of model use. It usually takes about one hour to teach the case.

C-50

EXHIBIT 1

Y(t) 12 2 SEASONAL AVERAGE NORMALIZE SALES PERIOD PERIOD RATIOS SEASONAL SEASONAL M.A. M.A. RATIOS RATIOS JAN 00 2000 FEB 00 3000 MAR 00 3000 APR 00 3000 MAY 00 4000 JUN 00 6000 6500.0 JUL 00 7000 6583.3 6541.7 1.0701 0.8411 0.8484 AUG 00 6000 6666.7 6625.0 0.9057 1.1645 1.1745 SEP 00 10000 6666.7 6666.7 1.5000 1.6539 1.6681 OCT 00 12000 6833.3 6750.0 1.7778 1.7299 1.7447 NOV 00 14000 6916.7 6875.0 2.0364 2.1307 2.1490 DEC 00 8000 7083.3 7000.0 1.1429 1.1889 1.1991 JAN 01 3000 6750.0 6916.7 0.4337 0.4583 0.4623 FEB 01 4000 6916.7 6833.3 0.5854 0.4805 0.4846 MAR 01 3000 7083.3 7000.0 0.4286 0.4618 0.4657 APR 01 5000 7083.3 7083.3 0.7059 0.3781 0.3814 MAY 01 5000 7250.0 7166.7 0.6977 0.6130 0.6183 JUN 01 8000 7416.7 7333.3 1.0909 0.7969 0.8038 JUL 01 3000 7333.3 7375.0 0.4068 11.89772 SUM AUG 01 8000 7416.7 7375.0 1.0847 SEP 01 12000 7583.3 7500.0 1.6000 OCT 01 12000 7416.7 7500.0 1.6000 NOV 01 16000 7333.3 7375.0 2.1695 DEC 01 10000 7166.7 7250.0 1.3793 JAN 02 2000 7500.0 7333.3 0.2727 FEB 02 5000 7666.7 7583.3 0.6593 MAR 02 5000 7916.7 7791.7 0.6417 APR 02 3000 8166.7 8041.7 0.3731 MAY 02 4000 8333.3 8250.0 0.4848 JUN 02 6000 8166.7 8250.0 0.7273 JUL 02 7000 8416.7 8291.7 0.8442 AUG 02 10000 8333.3 8375.0 1.1940 SEP 02 15000 8250.0 8291.7 1.8090 OCT 02 15000 8166.7 8208.3 1.8274 NOV 02 18000 8250.0 8208.3 2.1929 DEC 02 8000 8333.3 8291.7 0.9648 JAN 03 5000 8583.3 8458.3 0.5911 FEB 03 4000 8916.7 8750.0 0.4571 MAR 03 4000 9000.0 8958.3 0.4465

C-51

EXHIBIT 1 (continued) APR 03 2000 9083.3 9041.7 0.2212 MAY 03 5000 9250.0 9166.7 0.5455 JUN 03 7000 9583.3 9416.7 0.7434 JUL 03 10000 9583.3 9583.3 1.0435 AUG 03 14000 9416.7 9500.0 1.4737 SEP 03 16000 9333.3 9375.0 1.7067 OCT 03 16000 9333.3 9333.3 1.7143 NOV 03 20000 9500.0 9416.7 2.1239 DEC 03 12000 9416.7 9458.3 1.2687 JAN 04 5000 9250.0 9333.3 0.5357 FEB 04 2000 8916.7 9083.3 0.2202 MAR 04 3000 9250.0 9083.3 0.3303 APR 04 2000 9583.3 9416.7 0.2124 MAY 04 7000 9750.0 9666.7 0.7241 JUN 04 6000 9416.7 9583.3 0.6261 JUL 04 8000 9818.2 SUM AUG 04 10000 390708.3 SEP 04 20000 OCT 04 20000 NOV 04 22000 DEC 04 8000 REGRESSION COEFFICIENTS B = 70.24578 A = 5997.260

NOTE: NORMALIZED SEASONAL RATIOS ARE CONSTRUCTED SO THAT THE TOTAL OF THE SEASONAL RATIOS IS 12.

FORECAST BASED ON DECOMPOSITION t Y SEASONAL RATIOS FORECAST

JAN05 61 10282 0.4623 4753 FEB05 62 10352 0.4846 5017 MAR05 63 10423 0.4657 4854 APR05 64 10493 0.3814 4002 MAY05 65 10563 0.6183 6531 JUN05 66 10633 0.8038 8547 JUL05 67 10704 0.8484 9081 AUG05 68 10774 1.1745 12654 SEP05 69 10844 1.6681 18090 OCT05 70 10914 1.7447 19043 NOV05 71 10985 2.1490 23606 DEC05 72 11055 1.1991 13257 129435

C-52

LAWN KING, INC.

Teaching Note

Synopsis and Purpose

Lawn King is a manufacturer of lawn mowers facing a highly seasonal demand for its products. At the present time the demand forecast for the coming year has just been increased. This is causing management to evaluate the accuracy of the forecast, and to construct several different production strategies (level, chase, second shift) for meeting demand.

The purpose of this case is to illustrate the issues typically encountered in aggregate planning. The student is asked to make a demand forecast, to construct alternative production strategies and to recommend a particular strategy. A substantial amount of "pencil pushing" and "computer pushing" is required in this case to develop and evaluate the various strategies. The case illustrates the tradeoffs involved in aggregate production planning.

Discussion Questions

1. Develop a forecast to use as a basis for aggregate production planning.

2. Develop an aggregate production plan by month for fiscal 2002. Consider the use of several different production strategies. Which strategy do you recommend? Hint: Use of Excel will greatly save time in making these plans.

Analysis

The first step in analysis of this case is to evaluate the demand forecast. This can be done by calculating the actual increase in total demand over the past year. The increase from FY00 to FY01 was

84,600 / 69,500 = 1.217

The projected increase from FY01 to FY02 is

110,000 / 84,600 = 1.30

Thus a larger increase is being projected than was experienced last year.

We also observe that forecasts in the past have been very accurate (e.g. FY00 actual compared to forecast and FY01 actual compared to forecast). But, the forecasts by model type have not been nearly as accurate as total demand forecasts. Furthermore, the case states that demand is highly influenced by the economy and the weather. In view of this, past forecasts have been remarkably accurate.

C-53

For purposes of analysis we will accept the new forecast of 110,000 units. Although the projected demand increase is larger than last year's actual increase, the forecast still appears reasonable. It may be, however, that marketing is attempting to drive production through a higher forecast to avoid stockouts. Therefore, we may wish to evaluate a somewhat lower forecast, as well as the one given in the case.

It is best to evaluate the various production strategies in terms of aggregate demand. Evaluating these strategies by model type results in a tremendous amount of detailed calculation.

To construct an aggregate plan we need to forecast aggregate demand by month. This can be done by assuming the same monthly pattern as last year. From exhibit 4 in the case, the percentage of annual sales by month can be calculated. These percentages are then multiplied by the total forecast (110,000) to arrive at monthly demand forecasts. (See Exhibit 1 of the teaching note.)

Next, we must decide on the inventory level needed at the end of the year and the stockout policy desired. The current inventory is 16,460 units. On an annual basis this inventory level represents a turnover of

110,000 / 16,460 = 6.7

While a turnover of 6.7 might be considered good, the inventory level should ideally be compared to the demand at the end of the year. Since the demand is seasonal, our goal should be to have 1 or 2 months of inventory at year-end as a safety stock. More inventory is not necessary, since all models are still in production and we can respond to changing demand conditions. One month of inventory would amount to 1216 units (the projected demand for September). Two months of inventory would be 3698 units, the demand for September and October. By this criterion, a great deal of excess inventory exists. Therefore, we will assume an 8/31/02 goal of 3700 units (2 months supply) of inventory for the remainder of this analysis.

Adjusting for the inventory change, we have a production requirement of 97,240 units.

Forecast 110,000Beginning Inventory -16,460Ending Inventory +3,700 Production Required 97,240

There are many alternative strategies to consider. For the sake of simplicity we shall consider four strategies.

1. Level production2. Level production with overtime3. Chase demand4. Two shifts

C-54

The level production strategy is shown in Exhibit 1 attached. A level of 100 workers is used for September through February. This level is then phased down to 85 and then to 49 workers at the end of the year in order to reach an ending inventory of about 3700 units. It is not possible to use a completely level strategy in this case without significant stockouts or a larger ending inventory than desired. Thus an arbitrary initial level of 100 workers is selected with a reduction in work force later in the year. Other levels could also be selected.

The second strategy, shown in Exhibit 2, is a level strategy with overtime. In this case we choose a level of 85 regular workers through May, and then phased down to 52 workers at the end of the year to achieve an ending inventory level of 3700 units. Overtime is used in the months of December through May to meet the peak demands. Other profiles of level production and overtime could be selected.

Note, the ending inventory in all strategies should be the same in order to insure a comparable basis of costing. Students often overlook this point and, as a result, arrive at very different cost estimates.

The third strategy is to chase demand as shown in Exhibit 3. The chase strategy matches demand in Sept through Feb. Then a maximum of 200 workers in used in March and April while inventory is worked off and the level of workers is phased down to chase demand and end the year with 3700 units.

The fourth strategy is a two-shift strategy, shown in Exhibit 4. This strategy starts with a level of 60 workers in Sept through Dec (first shift) and then doubles the level of workers to 120 (second shift) from Jan through May. The second shift is phased down to arrive at the same ending inventory as the rest of the strategies.

In order to evaluate these four strategies we will need to make various assumptions about costs and resources. The first assumption is the nominal production rate of a worker in a month. Using the data from Exhibit 4 in the case, an average daily production of 373 units is computed as the following weighted average:

Since there are 260 production days in a year (52 weeks x 5 days per week), the average monthly production is 8,082 units:

260(373)-------- = 8,082 units per month. 12

C-55

The initial variable work force level is 85 workers (excludes 10 maintenance and 5 office workers). The production per direct (or variable) worker is therefore:

8,082----- = 95 units per worker per month 85

The hiring cost per worker is $800 and the layoff cost is $1500 per worker as given in the case. To calculate inventory carrying cost, we need to know the cost of producing a unit. Using labor and material costs for FY92, we arrive at a unit cost of

$10,600,000------------ = $125 per unit84,450 units

This unit cost is multiplied by 2.5% a month carrying cost (30% a year) to arrive at

$3.125 per unit/month

This unit carrying cost is multiplied by the total inventory carried to arrive at the inventory carrying cost for each month.

The direct labor cost per hour is obtained by taking the direct labor costs from the Profit and Loss Statement (Exhibit 1 in the case) and dividing by the number of direct workers (85) times 2000 hours per year as follows:

$2,595,000---------- = $15.30 per hour 85(2000)

The overtime rate is 150% of direct labor and thus the overtime rate is 150% times $15.30 = $23.00 per hour.

In order to calculate the costs of each strategy we use the spreadsheet for this case on the CD-ROM supplied with the textbook (file named LAWNKING). The above cost numbers are input into the spreadsheet for all of the strategies. Then each strategy is evaluated, one at a time, as shown in Exhibits 1 to 4.

The result of these cost evaluations is as follows:

Strategy 1: Level $3,414,411Strategy 2: Level with Overtime $3,360,109Strategy 3: Chase $3,425,407Strategy 4: Two Shift $3,290,880

As noted above, the Two-Shift Strategy has the lowest cost, by about $70,000 per year, over the Level with Overtime Strategy. The two-shift strategy is the cheapest, because overtime is more expensive and it is relatively expensive to hire and layoff workers. In a sense the two-shift strategy does the best job of fitting the demand profile by using regular workers.

C-56

The two-shift strategy not only offers cost advantages, but more flexibility and less inventory risk than the level strategy. The two-shift strategy is therefore preferred to the level strategy, provided employees can be found to work a second shift for only part of the year.

While the use of level strategy with overtime is more attractive than a pure level strategy, the flexibility to meet further demand increases is gone once overtime is built into the aggregate plan. Especially in view of the fact that a large amount of overtime is needed in strategy 2.

The two-shift strategy is also preferred to the chase strategy because it not only costs less but requires less personnel turmoil. Hiring and layoff is only done once for the two-shift strategy rather than frequently throughout the year. The chase strategy also implies that the production line can be easily speeded up and slowed down, while the two shift strategy provides for a constant line speed.

I think the above arguments provide a compelling case for the two-shift strategy. Lively arguments can be constructed, however, because the company can "squeak by" without using a second shift. Also, the two-shift strategy might not be entirely obvious to some students as an option that should be evaluated.

Teaching Strategy

This case can be taught using the same order of discussion as the above analysis: forecast, alternative strategies, costing, recommendation. The case will take one hour or more to teach depending on how much analysis is put on the board and how much discussion is encouraged.

This case provides practice in formulating and evaluating aggregate production strategies. Management is facing a dilemma because the sales manager has pushed the forecast a little beyond a one-shift operation, unless large amounts of overtime or hiring and layoff are used. The demand here is also very seasonal so flexibility in addition to cost is an important issue in production planning. There is also the question of how much inventory to carry at the end of the year.

When assigning the case, the students should be warned not to engage in excessive number crunching. Students can easily become bogged down in developing one schedule after another, while not reaching a sound conclusion.

C-57

EXHIBIT 1 (LEVEL STRATEGY)

FILENAME: LAWNKING NAME: KEY*

SECTION: *BASIC INPUT DATA DATE: 03-Apr =MONTHLY PROD. RATE (UNITS/WORKER/MONTH): 95 UNITSBEGINNING INVENTORY (UNITS): 16460 UNITSBEGINNING NUMBER OF WORKERS:

85 WORKERS

HIRING COST PER WORKER: $800 LAYOFF COST PER WORKER: $1,500 INVENTORY HOLDING COST ($ PER UNIT PER MONTH): $3.13 INVENTORY SHORTAGE COST (COST/UNIT SHORT) : $0.00 REGULAR HOURLY WAGE RATE: $15.30 OVERTIME HOURLY WAGE RATE: $23.00 HOURS/MONTH: 160

MONTH SEPT. OCT. NOV. DEC. JAN.- - - - -

Sales Forecast: 1,216 2,482 4,677 5,970 6,950 Units Produced: 9500 9500 9500 9500 9500

Ending Inventory: 24744 31762 36585 40115 42665 Number of Workers: 100 100 100 100 100

Overtime Percent: 0% 0% 0% 0% 0%Total Equivalentnumber of workers: 100 100 100 100 100 (# workers + O.T.)

Reg. Labor Costs: $244,800 $244,800 $244,800 $244,800 $244,800 O.T. Labor Costs: $0 $0 $0 $0 $0

Change in # workersfrom last month: 15 0 0 0 0

Hiring Cost: $12,000 $0 $0 $0 $0 Layoff Cost: $0 $0 $0 $0 $0

End-Inv. Hold. Cost: $77,325 $99,256 $114,328 $125,359 $133,328 End-Inv. Short. Cost: $0 $0 $0 $0 $0

- - - - -MONTHLY TOTAL COST: $334,125 $344,056 $359,128 $370,159 $378,128

C-58

EXHIBIT 1 (continued)

FEB. MAR. APR. MAY JUNE JULY AUG.- - - - - - -

10,877 23,185 24,642 15,743 6,598 4,299 3,352 9500 8075 8075 8075 5700 5700 4655

41288 26178 9611 1943 1045 2446 3749 100 85 85 85 60 60 49 0% 0% 0% 0% 0% 0% 0%

100 85 85 85 60 60 49

$244,800 $208,080 $208,080 $208,080 $146,880 $146,880 $119,952 $0 $0 $0 $0 $0 $0 $0

0 -15 0 0 -25 0 -11 $0 $0 $0 $0 $0 $0 $0 $0 $22,500 $0 $0 $37,500 $0 $16,500

$129,025 $81,806 $30,034 $6,072 $3,266 $7,644 $11,716 $0 $0 $0 $0 $0 $0 $0

- - - - - - -$373,825 $312,386 $238,114 $214,152 $187,646 $154,524 $148,168

STRATEGY DESCRIPTION & ASSUMPTIONS=====================

=========

=========

LEVEL STRATEGYTHIS IS A CLASSICAL LEVEL STRATEGY WITH A REDUCTION IN THE LEVELTOWARD THE END OF THE PLANNING PERIOD IN ORDER TO ARRIVE AT AN ENDING INVENTORY OF 3700 UNITS. THE INITIAL WORK FORCE ISARBITRARILY SELECTED TO BE 100 WORKERS.

$3,414,411 TOTAL COST OF THIS STRATEGY

C-59

EXHIBIT 2 (OVERTIME)

MONTHS: SEPT. OCT. NOV. DEC. JAN.Sales Forecast: 1,216 2,482 4,677 5,970 6,950 Units Produced: 8075 8075 8075 9206 9206

Ending Inventory: 23319 28912 32310 35546 37801 Number of Workers: 85 85 85 85 85

Overtime Percent: 0% 0% 0% 14% 14%Total Equivalentnumber of workers: 85 85 85 96.9 96.9 (# workers + O.T.)

Reg. Labor Costs: $208,080 $208,080 $208,080 $208,080 $208,080 O.T. Labor Costs: $0 $0 $0 $43,792 $43,792

Change in # workersfrom last month: 0 0 0 0 0

Hiring Cost: $0 $0 $0 $0 $0 Layoff Cost: $0 $0 $0 $0 $0

End-Inv. Hold. Cost: $72,872 $90,350 $100,969 $111,080 $118,128 End-Inv. Short. Cost: $0 $0 $0 $0 $0

- - - - -MONTHLY TOTAL COST: $280,952 $298,430 $309,049 $362,952 $370,000

FEB. MAR. APR. MAY JUNE JULY AUG.- - - - - - -

10,877 23,185 24,642 15,743 6,598 4,299 3,352 9206 9206 9206 9206 7125 5700 4940

36130 22150 6714 176 703 2104 3692 85 85 85 85 75 60 52

14% 14% 14% 14% 0% 0% 0%

96.9 96.9 96.9 96.9 75 60 52

$208,080 $208,080 $208,080 $208,080 $183,600 $146,880 $127,296 $43,792 $43,792 $43,792 $43,792 $0 $0 $0

0 0 0 0 -10 -15 -8 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $15,000 $22,500 $12,000

$112,905 $69,219 $20,980 $550 $2,197 $6,575 $11,538 $0 $0 $0 $0 $0 $0 $0

- - - - - - -$364,777 $321,091 $272,852 $252,422 $200,797 $175,955 $150,834

STRATEGY DESCRIPTION & ASSUMPTIONSLEVEL WITH OVERTIMETHIS IS A LEVEL STRATEGY WITH OVERTIME USED TO AVOID STOCKOUTS.SOME WORKERS ARE LAID OFF AT THE END OF THE PLANNING PERIOD TOACHIEVE AN ENDING INVENTORY OF 3700 UNITS.

$3,360,109 TOTAL COST OF THIS STRATEGY

C-60

EXHIBIT 3 (CHASE STRATEGY)SEPT. OCT. NOV. DEC. JAN.

Sales Forecast: 1,216 2,482 4,677 5,970 6,950 Units Produced: 1235 2470 4655 5985 6935

Ending Inventory: 16479 16467 16445 16460 16445 Number of Workers: 13 26 49 63 73

Overtime Percent: 0% 0% 0% 0% 0%Total Equivalentnumber of workers: 13 26 49 63 73 (# workers + O.T.)

Reg. Labor Costs: $31,824 $63,648 $119,952 $154,224 $178,704 O.T. Labor Costs: $0 $0 $0 $0 $0

Change in # workersfrom last month: -72 13 23 14 10

Hiring Cost: $0 $10,400 $18,400 $11,200 $8,000 Layoff Cost: $108,000 $0 $0 $0 $0

End-Inv. Hold. Cost: $51,497 $51,459 $51,391 $51,438 $51,391 End-Inv. Short. Cost: $0 $0 $0 $0 $0

- - - - -MONTHLY TOTAL COST: $191,321 $125,507 $189,743 $216,862 $238,095

FEB. MAR. APR. MAY JUNE JULY AUG.- - - - - - -

10,877 23,185 24,642 15,743 6,598 4,299 3,352 10830 19000 19000 13300 6650 3800 3325 16398 12213 6571 4128 4180 3681 3654

114 200 200 140 70 40 35 0% 0% 0% 0% 0% 0% 0%

114 200 200 140 70 40 35

$279,072 $489,600 $489,600 $342,720 $171,360 $97,920 $85,680 $0 $0 $0 $0 $0 $0 $0

41 86 0 -60 -70 -30 -5 $32,800 $68,800 $0 $0 $0 $0 $0

$0 $0 $0 $90,000 $105,000 $45,000 $7,500 $51,244 $38,166 $20,534 $12,900 $13,063 $11,503 $11,419

$0 $0 $0 $0 $0 $0 $0 - - - - - - -

$363,116 $596,566 $510,134 $445,620 $289,423 $154,423 $104,599

STRATEGY DESCRIPTION & ASSUMPTIONSCHASE STRATEGYTHIS IS A CLASSIC CHASE STRATEGY. MAXIMUM WORK FORCE IS LIMITED ARBITRARILY TO 200 WORKERS WHICH REDUCES THE INVENTORY IN PEAK PERIODS TO ACHIEVE AN ENDING INVENTORY OF 3700 UNITS.

$3,425,407 TOTAL COST OF THIS STRATEGY

C-61

EXHIBIT 4 (TWO-SHIFT STRATEGY)

SEPT. OCT. NOV. DEC. JAN.- - - - -

Sales Forecast: 1,216 2,482 4,677 5,970 6,950 Units Produced: 5700 5700 5700 5700 11400

Ending Inventory: 20944 24162 25185 24915 29365 Number of Workers: 60 60 60 60 120

Overtime Percent: 0% 0% 0% 0% 0%Total Equivalentnumber of workers: 60 60 60 60 120 (# workers + O.T.)

Reg. Labor Costs: $146,880 $146,880 $146,880 $146,880 $293,760 O.T. Labor Costs: $0 $0 $0 $0 $0

Change in # workersfrom last month: -25 0 0 0 60

Hiring Cost: $0 $0 $0 $0 $48,000 Layoff Cost: $37,500 $0 $0 $0 $0

End-Inv. Hold. Cost: $65,450 $75,506 $78,703 $77,859 $91,766 End-Inv. Short. Cost: $0 $0 $0 $0 $0

- - - - -MONTHLY TOTAL COST: $249,830 $222,386 $225,583 $224,739 $433,526

FEB. MAR. APR. MAY JUNE JULY AUG.- - - - - - -

10,877 23,185 24,642 15,743 6,598 4,299 3,352 11400 11400 11400 11400 8550 4750 4180 29888 18103 4861 518 2470 2921 3749

120 120 120 120 90 50 44 0% 0% 0% 0% 0% 0% 0%

120 120 120 120 90 50 44

$293,760 $293,760 $293,760 $293,760 $220,320 $122,400 $107,712 $0 $0 $0 $0 $0 $0 $0

0 0 0 0 -30 -40 -6 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $0 $45,000 $60,000 $9,000

$93,400 $56,572 $15,191 $1,619 $7,719 $9,128 $11,716 $0 $0 $0 $0 $0 $0 $0

- - - - - - -$387,160 $350,332 $308,951 $295,379 $273,039 $191,528 $128,428

STRATEGY DESCRIPTION & ASSUMPTIONSA TWO-SHIFT STRATEGY IS USED TO CLOSELY MATCH DEMAND. THE FIRSTSHIFT PROVIDES A TOTAL OF 120 WORKERS. SOME WORKERS ARE LAID-OFFAT THE END OF THE PLANNING PERIOD TO ARRIVE AT 3700 UNITS ININVENTORY.

$3,290,880 TOTAL COST OF THIS STRATEGY

C-62

WORLD INDUSTRIAL ABRASIVES DIVISION

Teaching Note

Synopsis and Purpose

World Industrial Abrasives is a manufacturer of abrasive materials used to make sandpaper and other abrasive products. The case presents a scheduling problem for machines used to crush abrasive material from large chunks into smaller pieces. The resulting pieces are used on sandpaper and other abrasive products. At the present time the crushing machines are scheduled by hand and the case asks for development of an automated algorithm to assist the scheduler.

The purpose of this case is to provide a vehicle for the design of a scheduling system. Several different approaches to the scheduling system design are considered and a preferred approach selected.

Discussion Questions

1. Develop a detailed flow chart that replicates the manual method currently used by the scheduler.

2. Evaluate the advisability of using linear programming, simulation, heuristic rules, or the present scheduling method to solve this problem.

3. Develop a conceptual model to solve this problem. Specify the inputs, outputs, and computational algorithm you would use.

Analysis

In response to the first question, a flow chart is shown in Exhibit 1. The MRP requirements and the yield table are inputs to the scheduling method. The scheduler then selects an input quantity of mineral and a particular set of crushing and screening operations. Multiplying the input quantity by the yields for this particular set of operations results in the output of crushed minerals. These outputs are compared to the requirements to determine whether the requirements are met or whether the input weight or crushing and screening operations should be changed. This process of calculation is continued until the requirements are met or until no more time is available for calculation. If the time limit is reached, the best approximate solution is selected based on those crushing and screening combinations which have been examined. Altogether, there are 19 x 19 x 4 = 1444 possible combinations of crushing and screening operations.

Linear programming cannot be used to solve this problem since discrete crushing and screening operations must be selected. However, an integer programming formulation can be developed as follows:

C-63

Let Xi = 1 if crushing and screening operations i is selected0 otherwise i = 1, . . ., 1444

Aij = yield of mineral of size j from operation iRj = requirement of mineral of size j (lbs)

W = total input of raw mineral in lbs.

The requirements for mineral can be either exceeded or not met. Thus we have:

Where:

Uj = amount under the requirement

Sj = amount over the requirement

In this formulation Uj is a slack variable and Sj is a surplus variable. If Uj is positive, then additional material will have to be purchased from outside. If S j

is positive, then surplus material will have to be stored and possible recrushed to use it later.

There are three costs to be considered:

1. Recrushing2. Purchasing outside mineral3. Storage

Let dj = unit cost of recrushing

bj = unit cost premium of purchasing outside materials

cj = unit cost of storage

Then we wish to minimize

cost =

Subject to:

j = 1, . . , N

Xi = {0,1}, Uj 0, Sj 0

C-64

This is a nonlinear mixed integer programming problem. We can remove the nonlinearity by assuming a value for W and solving the resulting problem. As a first approximation we can select W = . Then smaller and larger values of W can also be used. Each time we select a different value of W, a new integer programming problem must be solved.

The problem can also be solved by simulation. In this case, the simulation amounts to an automation of the Flow Chart shown in Exhibit 1. A combination of crushing and screening operations is selected along with an input weight, and the process output is calculated. After 1444 iterations, all combinations can be evaluated for a particular input weight and a particular week. Since the calculations are quite simple, these combinations can be quickly evaluated and the results can be ranked by any criteria desired. The total cost can be calculated as shown above or other criteria can also be used.

The scheduling problem can also be solved by heuristics. For example, if more than 50,000 lbs of #10 grit are required, use primary crush #16. If less than 50,000 lbs are required use primary crush #14. Heuristics of this type could be developed by the scheduler based on past experience and rules of thumb which have worked in practice. Heuristics can be used alone to solve the problem or in connection with simulation to reduce the number of options which must be evaluated.

In this problem, simulation offers a great deal of appeal because the options can be easily enumerated and a variety of different objective functions and constraints can be included. The integer programming problem may be difficult to solve with available codes and offers less flexibility in formulation. Heuristics are probably not needed in this case, unless they are used to simplify the simulation model.

Teaching Strategy

This case can be taught by asking a student to describe the present scheduling system. A flowchart of the type shown in Exhibit 1 can be put on the board. Once agreement is reached on how the present system operates, a model can then be constructed.

Depending on the class background you may want to discuss math programming approaches to this problem or limit the discussion to simulation or heuristics approaches. The case provides a good opportunity to contrast and compare several different analytic approaches to the same problem. It also illustrates that the designer should begin with a sound understanding of the present system before designing a new scheduling system.

C-65

EXHIBIT 1

Present Scheduling Method

C-66

Yield TablesYield Tables

Assumed MineralInput and Crushingand Screening Options

MineralProduced

AreReqmtsmet?

Stop

Stop

Pick the best approximate solution

Is timeavailableto reassess

Change Assumptions

MRP Requirements by type for 8 weeks

Yes

No

Yes

No

CONSOLIDATED ELECTRIC

Teaching Notes

Synopsis and Purpose

Consolidated Electric is a wholesale distributor of electrical products primarily to electrical contractors. The company wants to design a system for inventory management of the 20,000 line-items carried in stock. A description is given in the case of the business environment and the current inventory control system in use.

The purpose of this case is to expose students to many of the issues encountered in inventory system design. These issues include: forecasting, replenishment decision rules, error control, multiple item interactions, ABC analysis, and top management control. The case also asks the student to describe how the system they design will help improve inventory management in the company.

Discussion Questions

1. Design an inventory control system for this business.

2. Describe how the system you have designed will help the company meet customer service and cost objectives.

Analysis

One issue that needs to be considered during inventory system design is whether the system will utilize periodic or continuous review. Periodic review should be used in order to take advantage of purchase discounts and shipping economies. With periodic review, different items from the same supplier can be consolidated into a single order/single shipment. Since most product lines are reviewed on a weekly basis, this practice should be continued for control and shipping purposes, unless a bi-weekly cycle becomes an obvious choice.

Items should be consolidated by line for ordering purposes. Thus, an entire line must be reviewed before ordering decisions are made. The case indicates there are about 200 different lines to consider on a weekly basis.

An inventory control system is shown in Exhibit 1. The forecasting model receives actual demand and produces a forecast for each item on a weekly basis. The forecast should extend through the ordering lead time for each item.

The order module accepts the forecast along with management inputs for service level and costs. As a result, orders are calculated for each line and aggregate control information is fed back to management prior to order placement. Orders are then placed, shipments occur from the vendor, and disbursements are made to customers.

C-67

Before designing the forecasting module we should examine the data in Appendix 1 of the case. A frequency distribution of the individual customer demands for seven months is shown in Exhibit 2. The distribution indicates there are two types of customers, those who order a small number of units and those who place very large orders. It would be very expensive to carry safety stock for the few customers who order 50, 60, 65 or 100 units. Further examination of Appendix 1 indicates that the same customer has ordered the 65 units on each of the four occasions and 100 units on one occasion. A second customer has ordered 60 units on one date and 50 units on another. If these demands occur on a predictable or known basis, perhaps they can be scheduled in advance with the two customers and therefore be excluded from inventory uncertainty.

The demand data also indicates that the distribution is highly non-normal. Nevertheless, the demand over lead time may approach a normal distribution, particularly when the large orders are excluded. The lead time in this case is about 10 to 15 days (see Appendix 1).

A forecasting system that will react to different types of demand distributions is needed. Exponential smoothing or other techniques should be used to forecast demand and both small item demand (Poisson) and normal distributions should be included. The forecasting model should also handle seasonal demand and it should provide error tracking of forecasts. When a forecast is not tracking with actual data, an exception notice should be given to the user. The forecasting system should also detect outliers in demand when they occur and notify the user. Finally, the forecasting system should forecast lead-time and MAD or standard deviation, as well as demand.

The ordering module should calculate EOQ and the reorder point for each item. The EOQ formula given in Exhibit 1 of the case can be used, but the reorder point from Exhibit 1 should not be used. That reorder point is not related to service level or to statistical considerations.

The reorder point for normal demand should be:

R = (L + P) + Z

where: R = reorder pointL + P = lead time + review period(L + P) = forecast demand over L + P periodsZ = safety factor = single period standard deviation of demand

This formula assumes a constant lead time. If lead time varies, the formula can be appropriately modified. In this case P = 1 week.

Reorder points should also be incorporated in the module for small demand items. In this case the appropriate Poisson probability tables should be included. Finally, the module should have an option to use any demand distribution supplied by the user.

C-68

Orders for each item in a line should be calculated without considering price discounts or shipment costs. After calculating all orders in the line, the total order size and weight should be accumulated. If these amounts are not sufficient to obtain a discount, an economic analysis should be made of increased order size. Each order should be increased by the percentage needed to reach the discount level or levels. The total cost (of ordering + carrying + purchasing + freight) should then be compared for the EOQ and each discount level using the line as a whole. This is preferable to the line-point procedure suggested in Exhibit 1.

After calculating the orders required for each product line, the system should supply aggregate control information for management prior to the order placement. This information should include, for example, total dollars ordered, total inventory levels projected, earn and turn ratios, and service levels. The aggregate information should allow management to decide whether to release the orders in total or not.

In terms of implementation it is clear that the ABC principle applies. The A items should, perhaps, be put on the system first and monitored more carefully than B or C items. Also simple reorder rules and forecasts might be used for B or C items to reduce computations.

The proposed inventory control system will help the company meet customer service and cost objectives in several ways:

1. The system relates inventory to customer service objectives. Precise service goals can be stated and reorder points set to achieve those goals.

2. The system relates inventory to costs and demand forecasts. It allows management to minimize costs subject to the service goals set while reacting to forecasted demand.

3. The proposed system should save time on the part of the buyers and management. The role of the buyer will also be changed to one of a systems manager and exception handler rather than dealing with routine orders.

4. The system will provide more uniform decision making across product lines and across buyers. The company will be less vulnerable to turnover of buyers.

5. The system should reduce inventory required for the same service level or alternatively it should provide more service for the same inventory level. As a result, profits should be improved.

Teaching Strategy

When teaching this case, I usually ask a student to describe the system he or she has designed. I then ask several other students to describe their systems. This leads to a discussion of good or bad features in each of the proposed systems.

After these features are described, I ask how the system will help improve inventory management at Consolidated Electric. This leads to a discussion of the costs and benefits of an inventory control system. In some cases questions of implementation and computerization are also raised.

C-69

The case should take about one hour to teach. Many of the important points in management of independent demand inventory systems can be taught by means of this case.

EXHIBIT 1

Inventory System

Forecasting Module

C-70

Customer Service Goals and Costs

Aggregate Control Info for Top Management

Order Module

Vendor Inventory

Purchase Orders

DemandOrders

Shipments Sales

EXHIBIT 2

Demand Distribution

0

5

10

15

20

25

1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96

Order Size

Nu

mb

er o

f O

rder

s

Series1

C-71

SOUTHERN TORO DISTRIBUTOR, INC.

Teaching Notes

Synopsis

This case describes a conversation between Joe Melaney, the owner of the Toro distributorship in Galveston, Texas, and his son, Joe Jr. The immediate subject of the conversation is to decide on the spring season order for the entire irrigation line. In the course of the conversation, other issues emerge, ranging from the proper inventory level for specific parts to the future ownership of the distributorship.

Purpose

This case can be used at two levels. At the first level, the case is an exercise in the development of an effective system for managing independent demand inventory, taking into account the particular problems faced by a distributorship. At the second level, the case can be used to demonstrate the importance of inventory management as a policy variable. Since Southern Toro is a distributorship, profitability depends heavily on inventory management.

Discussion Questions

1. What would you recommend that Joe Jr. do, assuming he takes control of Southern Toro?

2. Evaluate the importance of inventory and inventory management of Southern Toro Distributorship, both for irrigation products and spare parts. Should the inventory be cut back?

3. Evaluate the current inventory management system at Southern Toro. What inventory management system would you recommend?

Analysis

Joe Jr.'s course of action if he takes control of Southern Toro certainly depends on a financial analysis of the company. Exhibit TN-1 shows some of the common financial ratios for the fiscal years 2000, 2001 and 2002.

Southern Toro Distributorship has been steadily increasing in net worth over these years but the return on invested capital has been low. Furthermore, the future outlook is potentially disturbing. As Exhibit TN-1 shows, the distributorship is highly leveraged, with a sharp increase in 2002. This will increase the cost of any future financing. Liquidity is decreasing, particularly quick liquidity, suggesting that the company may be forced to seek additional financing unless other action is taken. The company's activity is decreasing also; this is particularly noticeable in inventory turnover. As might be expected, the ROA of the distributorship, never very high, has been steadily declining over the last three years.

C-72

The return may be improved with better management, but will probably never become extremely high. It is for Joe Jr. to decide whether or not the return can be enough to satisfy him. However, based on the above analysis, it appears that the distributorship is likely to encounter difficulties if present trends are allowed to continue. Since the sale of irrigation equipment is heavily dependent on the weather, the company must plan to be flexible to accommodate irregular sales patterns. Liquidity, particularly quick liquidity, must be re-established and maintained to ensure Southern Toro's ability to pay off its short-term obligations without relying on the sales of inventories.

Future plans for Southern Toro must be based on good inventory management. The majority of Southern Toro's assets are in inventory, yet inventory levels do not reflect sales levels. From FY 2000 to 2001, sales declined 10.3% while ending inventory was reduced only 3.2%. Between FY 2001 and 2002, sales increased 20%, but ending inventory increased by 67%. This was an increase of $400,000 in ending inventory. It was necessary to increase notes payable by $371,000 to finance this increase. Inventories need to be reduced to improve the turnover ratio and to increase liquidity once again.

A rough idea of the amount of inventory required to support this business can be calculated by using the relationship:

average inventory = Q/2 + safety stockwhere Q is the order size

In this case the order size (Q) will be four months of sales, since orders are placed three times a year. The average inventory will therefore be 2 months + safety stock. If there is no safety stock, the inventory will turn 12/2 = 6 times per year. Safety stock will merely reduce this turn rate. For example, one month of safety stock results in 4 turns.

Since the current turns are 4.0, it is clear that it will be difficult to reduce inventories while meeting current demand, lead time, and safety stock assumptions -- the most critical constraint being the supply lead time imposed by Toro. The Southern Toro Distributorship must seek to reduce this supply lead time or seek additional sources of funds, preferably from Toro in the form of notes or accounts payable to support larger inventories. Assuming this can be done, we turn to the next problem of designing an inventory control system.

The new inventory computer system does not appear to be successful. The system was installed in October of 2001. In FY 2002, an abnormally dry year, the company "stocked out of most goods" yet had ending inventories of $1,000,000 in June of 2002.

C-73

The software package, as described, is completely unsuited to Southern Toro's situation. The following points should be raised:

(1) Southern Toro has three preset ordering and delivery times each year. A reorder point for each item is therefore inappropriate.

(2) Southern Toro must order at the set times sufficient inventory to last until the next period's delivery. An EOQ is irrelevant.

(3) Demand for irrigation equipment is seasonal. A forecast of usage based on average monthly demand is insufficient.

(4) Large turf orders, from golf courses and other commercial installations, cause irregular demand "peaks". A mechanical forecast of the affected products is inappropriate.

Development of an appropriate inventory system for Southern Toro must begin by examining the specific circumstances. There are three separate ordering cycles each year. Although the periods and lead times are irregular, sufficient inventory must be ordered to cover the order period plus the lead time. Based on Exhibit 6 of the case, the following coverage times can be determined:

Order PeriodOrder Placement Coverage Until Plus Lead Time

Oct. 15 - Oct. 30 May and June 7 - 8 months

Feb. 15 - Feb. 30 August and Sept. 6 - 7 months

June 15 - June 30 December and Jan. 6 - 7 months

For the approaching October order, demand must be forecast for the period until May and June, when February's orders will be received. It may be appropriate to adjust the forecast based on weather predictions for the period. The forecasts for individual items, where larger orders are expected, may be adjusted manually.

An ABC analysis is needed to center attention on crucial inventory items. Exhibit TN-2 shows such an analysis for the irrigation products. The exhibit also shows the current inventory as a percentage of FY 2002 sales. As can be seen, the products are unequally stocked with stocks ranging from 9% to 80% of sales. In general, the A and B items are less heavily stocked than the C items. This suggests that the inventory investment is overly weighted with C items and that the inventory turns, with the resulting risk of stock out, primarily come from A and B items. This is borne out by the case, where concern is shown for possible overstock of a Monitor Controller, a C item, and possible understock of a B item valve.

C-74

Although an exact inventory system cannot be suggested, a new software package should be developed which is carefully tailored for Southern Toro. The software package should be used to automatically manage C items, both products and spare parts. A items, and to some extent B items, should be carefully monitored and the computerized system adjusted based on experience and knowledge.

The new system should be based on the following principles:

1. A forecast for each item should be produced from analysis of past data. One of several different models could be used provided that seasonal factors and trends are included in the model. For A items, and possibly some of the critical B or C items, the forecast should be carefully examined and adjusted, based on marketing information available. The average demand and standard deviation should be forecasted through the lead time plus review period.

2. A service level should be established for A, B, and C items. Based on the service level, forecasted average demand and standard deviation, a target inventory level can then be established for each item.

3. Order sizes should be determined to bring the inventory up to the desired target levels.

4. The order sizes should be converted to dollars and the resulting inventory levels projected on a monthly basis into the future. The total purchasing dollars required and the future projected inventory levels should be examined for business feasibility on a total dollar basis. If these expenditures and inventory levels cannot be financed or are considered excessive, then service levels should be revised or other assumptions modified (lead time) to achieve the total aggregate financial levels desired.

If this system is followed, it may be possible to bring the inventory levels into overall conformance with business goals. At this point it will become apparent how inventory levels can be reduced, liquidity improved, and return on capital invested increased. If the desired service levels are ultimately incompatible with investment and sales goals, then perhaps Joe Jr. should consider looking for other business opportunities.

C-75

EXHIBIT TN-1

FINANCIAL RATIOS

Ratio 2000 2001 2002

Liquidity:

Current: current assets 4.87 3.54 2.26 current liabilities

Quick: current - inventory 2.14 1.82 .88

current liabilities

Leverage:

Debt to equity .49 .59 .98

Activity:

Inventory turnover: 4.5 4.4 4.0 Cost of Goods Sold average inventory

Day's Receivables: 44 days 62 days 54 days

A/R x 365 sales

Total asset turnover 3.33 2.83 2.8 sales average assets

Profitability:

Return on Assets (ROA) .03 .03 .02net profit total assets

Return on net worth .05 .04 .05net profit net worth

C-76

EXHIBIT TN-2AN ABC ANALYSIS

FY 2002 FY 2002 Item % Sales inProduct Description $ Sales % Sales Class

Inventory

Free Controllers Series 150 - 4 + 8 15 2 C 80

Custom Controllers Series 123 - 8 + 11 12 1 C 67

Monitor Controllers Series 176 - 11 + 23 26 3 C 58

3/4" + 1" Valve Glove/angle in-line 78 8 B 59

1-1/2" + 2" Valve Glove/angle in-line 62 7 B 10

Brass Valve Glove/angel in-line 7 1 C 57

Pop-up Bodies 77 8 B 26

570 Series Nozzles 68 7 B 20

Stream Rotors Series 300 144 15 A 9

Rain Pro Series 320 26 3 C 46

Gear Driven Rotary Series 600 22 2 C 45

Gear Driven Rotary Series 620 39 4 C 54

Gear Driven Rotary Series 640 194 20 A 42

Gear Driven Rotary Series 670 180 19 A 20TOTALS $950 100%

C-77

TOYSPLUS, INC.

Teaching Note

Synopsis and Purpose

ToysPlus is a small company in the toy industry. They control the manufacturing of toys by using an MRP System. Management wants to improve service levels and inventory turns by doing a better job of planning and scheduling production. The case provides data on three toys which are manufactured including the forecasts, costs, bill of materials, and planning factors for these toys.

The purpose of this case is to illustrate the principles of scheduling with an MRP system. A spreadsheet is provided on the CD-ROM which accompanies the text to provide the MRP logic. The student inputs forecasts, costs and a master schedule. The spreadsheet then performs detailed parts explosions. The student must analyze the resulting schedules and re-plan until an acceptable MRP plan is obtained. Some organizational issues are also presented in the case which are encountered in using a formal MRP system.

Discussion Questions

1. Calculate economic order quantities for each of the three types of toys. The EOQ formula is recommended from the supplement to Chapter 15 that considers uniform lot delivery of toys.

2. Prepare a master production schedule for the next 6 weeks using the EOQ's calculated in Question 1 and a work force of 10 employees. What inventory turnover ratio is achieved by this master schedule? How does this turnover compare with past levels and with management's goals?

3. Prepare a parts explosion to support the master schedule. What parts should be ordered each week?

4. What should Andrea Meline do in order to meet the inventory and service goals stated by management?

5. How should Andrea deal with the organization issues presented in this case?

Analysis

The analysis of this case will be described using the Excel Worksheet provided with the text. This analysis can be done by hand, but it is quite tedious.

C-78

The first thing the students should do is to input the forecasts into the worksheet. The forecasts given in the case can be used directly, as shown in Exhibit 1. There is some question whether the forecasts coming from marketing are accurate. Andrea Meline usually adjusts the marketing forecasts, but we have no way of making any adjustments in this particular situation.

C-79

From the forecasts for each of the three toys, the spreadsheet will calculate average weekly demand. These figures are used for EOQ and runout calculations. The next step is to calculate the EOQ's for each of the toys. These are calculated by entering the weekly toy production rates and the unit costs given in the case, see Exhibit 1.

Next, the setup cost must be provided to the spreadsheet which is calculated as follows: Each final assembly setup requires 1 hour of time by 10 direct labor people. The shop rate for direct labor is $6 per hour for wages, 33% for fringes and $6 per hour for overhead or a total of $14 per hour. There is some question whether overhead should be counted or not, but we have included it here, assuming that it is variable overhead. The cost of setting up the final assembly line is therefore $140 (10 hours at $14 per hour). Each final assembly setup also induces orders to be placed for components using a lot-for-lot procedure. As a matter of fact, 4 orders for components will be placed for each final assembly setup, since each toy has four components. At $25 per order, this is an additional $100 for component orders. The total cost of setup is therefore $240.

The carrying cost is given in the case as 25% per year. After entering carrying cost, the program will calculate the EOQ's for the three toys using the uniform lot delivery EOQ given in the supplement to Chapter 15. The program then asks for the beginning inventory level input for use later in the analysis.

The next step is to do master scheduling for each toy for each week. This is done by first finding the smallest runout time which happens to occur for robots (runout time is just the inventory level divided by demand and is interpreted as the length of time the present inventory will last until we run out of stock). Robots are thus scheduled first with an EOQ of 4643 units. Only 2267 units can be scheduled in the first week and stay within the capacity constraint of 350 hours (each robot takes .15 assembly hours). Only 2333 units of robots can be scheduled in week 2 within the capacity constraint of 350 hours and the balance of the EOQ for robots (43 units) is scheduled for the third week. The product with the next lowest runout time, as shown by the spreadsheet, is toy autos. We can produce 3340 of these toy autos in week 3 within the capacity constraint of 350 hours (this includes 10 hours lost to the changeover). Then 3500 autos can be produced in week 4 and the balance of the EOQ for autos is produced in week 5. Next, toy trucks have the lowest runout time, and we schedule the balance of the capacity for toy trucks in week 5. This procedure is continued for week 6 to complete the master schedule.

The program then computes projected inventories and inventory turnover. Note, the turnover is only 13.4 which does not meet management’s objective of 15 turns. This indicates that we should revise the master schedule to reduce capacity and thus inventory. We will do this later, but first we wish to check the feasibility of the master schedule with respect to component parts.

C-80

This is done by entering the safety stock levels which are desired into the spreadsheet. Beginning inventories, parts/unit and lead times are already shown in the spreadsheet. The case indicates that one week of safety stock is carried for each of the components to cover late deliveries from suppliers. The usage for each of the three toys is multiplied by the number of parts per toy in order to arrive at the safety stock figures shown in Exhibit 1.

C-81

Each component part is now exploded in the bill of materials, twelve parts in all. However, the amounts due to be delivered in Exhibit 4 of the case must be entered first in order to get the correct answers. These on-order amounts are 800 units of part number 615 due on 10/3 and 1200 units of part number 621 due on 10/10.

As a result of these inputs, the planned order releases for each of the twelve parts is given in Exhibit 1. Note, that some of the planned orders may be past due. This is indicated by looking at the net requirements and seeing if they are reflected in planned orders. For example, part #523 has a net requirement of 2115 units on Oct 10 which is not translated into a planned order release. Therefore, we will not have enough of part #523 to support the master schedule. In the case of part number 525 all of the net requirements have been translated into planned order releases, so we will have enough of these parts. This process of examination is continued and indicates the following components are short of parts to support the master schedule #730, #732, #734, and #736. This situation must be relieved by revising the master schedule to "live" within the available parts.

The new master schedule is shown in Exhibit 2. This master schedule reflects the available components and is reduced in capacity toward the end of the six-week period in order to achieve 15 turns. This master schedule is constructed by using all of the available parts, including the safety stock, but no more in scheduling each toy. For example, starting with robots which has the lowest runout time, we can schedule 1600 robots based on the availability of components. The constraining component in this case is robot bodies which take two weeks to get. The next lowest runout time is autos, and we can schedule 2450 autos before we run out of parts (900 in week 1 and 1550 in week 2).This constraint of 2450 is based on the 9800 wheels which are used at the rate of 4 per auto and it takes 2 weeks to get more. This process is continued until there is sufficient lead time to schedule an EOQ of each toy.

Note, that capacity is reduced to 315 hours in week 3 and then to 280 hours for the remainder of the schedule. These numbers were selected by trial-and-error in order to increase inventory turns. However, this capacity plan is not entirely satisfactory, because we should have enough capacity at the end of the planning horizon to sustain the schedule. The required capacity in week 6 can be computed as follows: 1500(.1) + 300(.2) + 600(.15) = 300 hours. At the end of the horizon we have only 280 hours of capacity, so perhaps we should cut capacity sooner, but not as deeply.

Review of the parts explosion in Exhibit 2 shows that the plan is feasible and all net requirements are met with planned order releases or safety stock. This plan also meets management's requirement for 15 turns. We have no way of estimating the service level provided without knowing more about the standard deviation of forecast errors.

C-82

So far, we have answered the first 4 case discussion questions. Question 5 asks how Andrea should deal with the organization issues presented in this case. I think that the essence of the answer to this question is that she needs cooperation from all of the departments to develop and implement her production plans. Marketing should be held accountable for the accuracy of its forecasts. Andrea should not adjust the forecasts that marketing gives her without negotiating changes with marketing. Purchasing should buy what is ordered by production. When everyone is working to the same production plan, improved results will be achieved. This is easier said than done, but cooperation among functions is the essence to achieving good results from a formal production planning system.

One way to help achieve the level of cooperation needed is to have all Department Heads and the General Manager meet each week to review and approve the final production plans developed. This meeting will help insure that everyone has agreed to the same set of numbers and assumptions. The company must also reward a cooperative team approach and management must not override the formal system with edicts. Discipline to live by the agreed plan is needed by everyone.

Teaching Strategy

In teaching this case I start by asking a student to give me the master schedule for the next six weeks. After writing this schedule on the board, I ask the student to explain how the numbers were derived. Other students will then propose alternative master schedules and different sets of assumptions. For example, some of the master schedules provided by students may be infeasible with regard to available component parts, and other master schedules may not provide the required 15 inventory turns. These points can be brought out by the instructor as the discussion proceeds.

One of the other pitfalls in this case is the setup cost. Some students may not include the ordering costs for components along with the final assembly line setup costs. Students may also forget to include safety stock in their calculations or they may calculate it incorrectly. There are many numbers in this case and one false input changes the answer.

This case provides an excellent vehicle for explaining the concepts of production planning. The case illustrates rough-cut capacity planning which is done at the master schedule level, by the 350 hour limit. We do not, however, have detailed capacity planning, because no work center data is incorporated in the case. EOQ calculations, the handling of safety stock, restrictions from component parts, and leveling the master schedule can all be illustrated. The disk provides a useful vehicle to relieve the students of the drudgery of calculations. It also guides them through the calculations which are needed. It will take about one hour to teach the case.

C-83

Exhibit 1MASTER SCHEDULE FORECAST INPUT- - -----

SEP26 OCT3 OCT10 OCT17 OCT24 OCT31----- ----- ----- ----- ----- -----

TOY AUTO 1100 1150 1200 1300 1400 1500TOY TRUCK 500 450 400 350 300 300TOY ROBOT 700 650 650 625 625 600

------ ------ ------ ------ ------ ------ TOTAL UNITS 2300 2250 2250 2275 2325 2400

AVERAGE WEEKLY DEMAND (CALCULATED)TOY AUTO 1275 TOY TRUCK 383 TOY ROBOT 642

MASTER SCHEDULEEOQ CALCULATIONS AUTO TRUCK ROBOT- ------------ ----- ----- -----WEEKLY DEMAND RATE 1275 383 642 PRODUCTION RATE PER WEEK 3500 1750 2333UNIT COST 3.2 6.5 4.1

SETUP COST 240CARRYING COST(% Per Year) 25%

ECONOMIC ORDER QUANTITY 7910 2746 4643

BEGINNING INVENTORY LEVEL INPUT 4000 2000 1500

C-84

MASTER SCHEDULE (INPUT PLANNED QUANTITY FOR EACH WEEK)

PLANNED PRODUCTION (WEEK BEGINNING)SEP26 OCT3 OCT10 OCT17 OCT24 OCT31

----- ----- ----- ----- ----- -----TOY AUTO 0 0 3340 3500 1070 350TOY TRUCK 0 0 0 0 1170 1576TOY ROBOT 2267 2333 43 0 0 0

------ ------ ------ ------ ------ ------ TOTAL UNITS 2267 2333 3383 3500 2240 1926TOTAL LABOR HOURS 350 350 350 350 351 350

CALCULATED RUN-OUT TIMES (WEEKS) EOQ

TOY AUTO 9.6 7910 TOY TRUCK 12.4 2746 TOY ROBOT 9.6 4643

PROJECTED INVENTORIES OF FINISHED GOODSSEP26 OCT3 OCT10 OCT17 OCT24 OCT31

----- ----- ----- ----- ----- -----TOY AUTO 2900 1750 3890 6090 5760 4610TOY TRUCK 1500 1050 650 300 1170 2446TOY ROBOT 3067 4750 4143 3518 2893 2293

------ ------ ------ ------ ------ ------ TOTAL UNITS 7467 7550 8683 9908 9823 9349

FINISHED GOODS INVENTORY ANALYSIS:AVERAGE WEEKLY USAGE IN $ $10,929 AVERAGE WEEKLY INVENTORY IN $ $42,557

FINISHED GOODS TURNOVER RATIO 13.4

C-85

COMPONENT PART INPUTS

BEGINNINGG PARTS LEAD SAFETYCOMPENENT PART INVENTORYY \UNIT TIME STOCK

523 CAR BODY 2500 1 3 1275525 AUTO WHEELS 9800 4 2 5100529 AUTO SIDE WINDOWS 4300 2 1 2550531 AUTO WINDSHIELD 2620 1 2 1275615 TRUCK CAB 1200 1 3 383617 TRUCK DUAL WHEELS 9900 8 2 3064619 TRUCK SINGLE WHEELS 2500 2 2 726621 TRUCK TRAILER 4600 1 4 383730 ROBOT BODY 1600 1 2 642732 ROBOT ARMS 3500 2 2 1284734 ROBOT LEGS 4020 2 1 1284736 ROBOT HEAD 2150 1 2 642

C-86

COMPONENT PARTS EXPLOSIONWEEK BEGINNING

SEP 26 OCT3 OCT10 OCT 17 OCT 24 OCT 31

523GROSS REQUIREMENTS 0 0 3340 3500 1070 350 SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 2500 2500 1275 1275 1275 1275NET REQUIREMENTS 0 0 2115 3500 1070 350 PLAN. ORDER RELEASES 3500 1070 350 0 0 0

- - - - - - - -525GROSS REQUIREMENTS 0 0 13360 14000 4280 1400

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 9800 9800 5100 5100 5100 5100NET REQUIREMENTS 0 0 8660 14000 4280 1400 PLAN. ORDER RELEASES 8660 14000 4280 1400 0 0

- - - - - - - -529GROSS REQUIREMENTS 0 0 6680 7000 2140 700

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 4300 4300 2550 2550 2550 2550NET REQUIREMENTS 0 0 4930 7000 2140 700 PLAN. ORDER RELEASES 0 4930 7000 2140 700 0

- - - - - - - -531GROSS REQUIREMENTS 0 0 3340 3500 1070 350

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 2620 2620 1275 1275 1275 1275NET REQUIREMENTS 0 0 1995 3500 1070 350 PLAN. ORDER RELEASES 1995 3500 1070 350 0 0

- - - - - - - -615GROSS REQUIREMENTS 0 0 0 0 1170 1576

SCHEDULED RECEIPTS 0 800 0 0 0 0 PROJ. ON HAND INV. 1200 2000 2000 2000 830 383NET REQUIREMENTS 0 0 0 0 0 1129 PLAN. ORDER RELEASES 0 0 1129 0 0 0

- - - - - - - -617GROSS REQUIREMENTS 0 0 0 0 9360 12608

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 9900 9900 9900 9900 3064 3064NET REQUIREMENTS 0 0 0 0 2524 12608 PLAN. ORDER RELEASES 0 0 2524 12608 0 0

- - - - - - - -619GROSS REQUIREMENTS 0 0 0 0 2340 3152

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 2500 2500 2500 2500 726 726NET REQUIREMENTS 0 0 0 0 566 3152 PLAN. ORDER RELEASES 0 0 566 3152 0 0

- - - - - - - -

C-87

621GROSS REQUIREMENTS 0 0 0 0 1170 1576 SCHEDULED RECEIPTS 0 0 1200 0 0 0 ON HAND INVENTORY 4600 4600 5800 5800 4630 3054NET REQUIREMENTS 0 0 0 0 0 0 PLAN. ORDER RELEASES 0 0 0 0 0 0

- - - - - - - -730GROSS REQUIREMENTS 2267 2333 43 0 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 642 642 642 642 642 642NET REQUIREMENTS 1309 2333 43 0 0 0 PLAN. ORDER RELEASES 43 0 0 0 0 0

- - - - - - - -732GROSS REQUIREMENTS 4534 4666 86 0 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 1284 1284 1284 1284 1284 1284NET REQUIREMENTS 2318 4666 86 0 0 0 PLAN. ORDER RELEASES 86 0 0 0 0 0

- - - - - - - -734GROSS REQUIREMENTS 4534 4666 86 0 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 1284 1284 1284 1284 1284 1284NET REQUIREMENTS 1798 4666 86 0 0 0 PLAN. ORDER RELEASES 4666 86 0 0 0 0

- - - - - - - -736GROSS REQUIREMENTS 2267 2333 43 0 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 642 642 642 642 642 642NET REQUIREMENTS 759 2333 43 0 0 0 PLAN. ORDER RELEASES 43 0 0 0 0 0

- - - - - - - -

C-88

Exhibit 2MASTER SCHEDULE FORECAST INPUT- - -----

SEP26 OCT3 OCT10 OCT17 OCT24 OCT31----- ----- ----- ----- ----- -----

TOY AUTO 1100 1150 1200 1300 1400 1500TOY TRUCK 500 450 400 350 300 300TOY ROBOT 700 650 650 625 625 600

------ ------ ------ ------ ------ ------ TOTAL UNITS 2300 2250 2250 2275 2325 2400

AVERAGE WEEKLY DEMAND (CALCULATED)TOY AUTO 1275 TOY TRUCK 383 TOY ROBOT 642

MASTER SCHEDULEEOQ CALCULATIONS AUTO TRUCK ROBOT- ------------ ----- ----- -----WEEKLY DEMAND RATE 1275 383 642 PRODUCTION RATE PER WEEK 3500 1750 2333UNIT COST 3.2 6.5 4.1

SETUP COST 240CARRYING COST(% Per Year) 25%

ECONOMIC ORDER QUANTITY 7910 2746 4643

BEGINNING INVENTORY LEVEL INPUT 4000 2000 1500

C-89

MASTER SCHEDULE (INPUT PLANNED QUANTITY FOR EACH WEEK)- - - - - --

PLANNED PRODUCTION (WEEK BEGINNING)SEP26 OCT3 OCT10 OCT17 OCT24 OCT31

----- ----- ----- ----- ----- -----TOY AUTO 900 1550 0 0 0 740TOY TRUCK 0 925 1576 245 0 0TOY ROBOT 1600 0 0 1470 1867 1306

------ ------ ------ ------ ------ ------ TOTAL UNITS 2500 2475 1576 1715 1867 2046TOTAL LABOR HOURS 350 350 315 280 280 280

CALCULATED RUN-OUT TIMES (WEEKS) EOQ- - - -TOY AUTO 5.6 7910 TOY TRUCK 12.4 2746 TOY ROBOT 12.1 4643

PROJECTED INVENTORIES OF FINISHED GOODSSEP26 OCT3 OCT10 OCT17 OCT24 OCT31

----- ----- ----- ----- ----- -----TOY AUTO 3800 4200 3000 1700 300 -460TOY TRUCK 1500 1975 3151 3046 2746 2446TOY ROBOT 2400 1750 1100 1945 3187 3893

------ ------ ------ ------ ------ ------ TOTAL UNITS 7700 7925 7251 6691 6233 5879

FINISHED GOODS INVENTORY ANALYSIS:AVERAGE WEEKLY USAGE IN $ $10,929 AVERAGE WEEKLY INVENTORY IN $ $37,101

FINISHED GOODS TURNOVER RATIO 15.3

C-90

COMPONENT PART INPUTS

BEGINNINGG PARTS LEAD SAFETYCOMPENENT PART INVENTORYY \UNIT TIME STOCK- - - - - - -523 CAR BODY 2500 1 3 1275525 AUTO WHEELS 9800 4 2 5100529 AUTO SIDE WINDOWS 4300 2 1 2550531 AUTO WINDSHIELD 2620 1 2 1275615 TRUCK CAB 1200 1 3 383617 TRUCK DUAL WHEELS 9900 8 2 3064619 TRUCK SINGLE WHEELS 2500 2 2 726621 TRUCK TRAILER 4600 1 4 383730 ROBOT BODY 1600 1 2 642732 ROBOT ARMS 3500 2 2 1284734 ROBOT LEGS 4020 2 1 1284736 ROBOT HEAD 2150 1 2 642

COMPONENT PARTS EXPLOSIONWEEK BEGINNING

SEP 26 OCT3 OCT10 OCT 17 OCT 24 OCT 31- - - - - - -

523GROSS REQUIREMENTS 900 1550 0 0 0 740 SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 1600 1275 1275 1275 1275 1275NET REQUIREMENTS 0 1225 0 0 0 740 PLAN. ORDER RELEASES 0 0 740 0 0 0

- - - - - - - -525GROSS REQUIREMENTS 3600 6200 0 0 0 2960

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 6200 5100 5100 5100 5100 5100NET REQUIREMENTS 0 5100 0 0 0 2960 PLAN. ORDER RELEASES 0 0 0 2960 0 0

- - - - - - - -529GROSS REQUIREMENTS 1800 3100 0 0 0 1480

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 2550 2550 2550 2550 2550 2550NET REQUIREMENTS 50 3100 0 0 0 1480 PLAN. ORDER RELEASES 3100 0 0 0 1480 0

- - - - - - - -531GROSS REQUIREMENTS 900 1550 0 0 0 740

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 1720 1275 1275 1275 1275 1275NET REQUIREMENTS 0 1105 0 0 0 740 PLAN. ORDER RELEASES 0 0 0 740 0 0

- - - - - - - -

C-91

615GROSS REQUIREMENTS 0 925 1576 245 0 0 SCHEDULED RECEIPTS 0 800 0 0 0 0 PROJ. ON HAND INV. 1200 1075 383 383 383 383NET REQUIREMENTS 0 0 884 245 0 0 PLAN. ORDER RELEASES 245 0 0 0 0 0

- - - - - - - -617GROSS REQUIREMENTS 0 7400 12608 1960 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 9900 3064 3064 3064 3064 3064NET REQUIREMENTS 0 564 12608 1960 0 0 PLAN. ORDER RELEASES 12608 1960 0 0 0 0

- - - - - - - -619GROSS REQUIREMENTS 0 1850 3152 490 0 0

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 2500 726 726 726 726 726NET REQUIREMENTS 0 76 3152 490 0 0 PLAN. ORDER RELEASES 3152 490 0 0 0 0

- - - - - - - -

621GROSS REQUIREMENTS 0 925 1576 245 0 0 SCHEDULED RECEIPTS 0 0 1200 0 0 0 ON HAND INVENTORY 4600 3675 3299 3054 3054 3054NET REQUIREMENTS 0 0 0 0 0 0 PLAN. ORDER RELEASES 0 0 0 0 0 0

- - - - - - - -730GROSS REQUIREMENTS 1600 0 0 1470 1867 1306

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 642 642 642 642 642 642NET REQUIREMENTS 642 0 0 1470 1867 1306 PLAN. ORDER RELEASES 0 1470 1867 1306 0 0

- - - - - - - -732GROSS REQUIREMENTS 3200 0 0 2940 3734 2612

SCHEDULED RECEIPTS 0 0 0 0 0 0 PROJ. ON HAND INV. 1284 1284 1284 1284 1284 1284NET REQUIREMENTS 984 0 0 2940 3734 2612 PLAN. ORDER RELEASES 0 2940 3734 2612 0 0

- - - - - - - -

C-92

734GROSS REQUIREMENTS 3200 0 0 2940 3734 2612 SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 1284 1284 1284 1284 1284 1284NET REQUIREMENTS 464 0 0 2940 3734 2612 PLAN. ORDER RELEASES 0 0 2940 3734 2612 0

- - - - - - - -736GROSS REQUIREMENTS 1600 0 0 1470 1867 1306

SCHEDULED RECEIPTS 0 0 0 0 0 0 ON HAND INVENTORY 642 642 642 642 642 642NET REQUIREMENTS 92 0 0 1470 1867 1306 PLAN. ORDER RELEASES 0 1470 1867 1306 0 0

- - - - - - - -

C-93

U.S. STROLLER

Teaching Note

Synopsis and Purpose

U.S. Stroller is a manufacturer and distributor of a line of baby strollers. The company makes three types of strollers in a typical batch production system. Recently, profits are down and competition is entering the market that the company dominates. Proposals have been made to go to a JIT production system or to a Cell approach for production. Students are asked to analyze these proposals and come up with a recommendation.

The purpose of this case is to show how a batch system can be dramatically improved by JIT or Cellular concepts. The case clearly describes the changes that must be made in the production system to achieve the benefits desired. This amounts to a complete change in layout of the production floor and a substantial overhaul of the MRP system. The case nicely ties together some of the material from the text on JIT, GT, EOQ and MRP. The case also requires the student to estimate cost and benefits of these types of proposals and to identify negative effects of JIT and GT.

Discussion Questions

1. Evaluate the current situation facing U.S. Stroller.

2. Discuss the pros and cons of the options presented in the case.

3. What will be the impact of these options on the MRP system currently in use?

4. What option do you recommend and why?

Analysis

Question 1. U.S. Stroller must do something quickly. Profits have eroded. They are not losing money yet, but financial results are poor based on percent of sales and return on equity. Also, Japanese competition is coming. Even though U.S. Stroller still commands the dominant share of the market, they are vulnerable on both quality and cost grounds. The Japanese are following a typical Japanese strategy by entering the market with a high quality stroller on the low end of the market. The Japanese can be expected to come after U.S. Stroller directly after gaining a foothold in the market.

U.S. Stroller cannot delay making improvements, until after they have lost market share. They must improve now in order to preempt the competition and to improve profits in the face of probable price erosion. There should be a sense of urgency in making drastic improvements in order to protect their market share. Some changes should be made, assuming U.S. Stroller wants to survive in the long-run.

C-94

Question 2. The Pro's and Con's of each alternative will be analyzed.

Option 1: JIT. This option proposes putting in three final assembly lines and dedicating various pieces of equipment to each of the three stroller models. A JIT pull system will be put in place, setup times will be reduced, and smaller lot sizes will be used.

Pro's

Option 1 reduces inventories by the following amounts

Change Inventory Savings Finished Goods from 80 days to 22 days6 $765,000 to $210,000Work in Process from 4 weeks to 2 weeks $322,000 to $161,000Raw Materials from 12 weeks to 2 weeks $337,000 to $56,000

The Savings in inventory will improve the balance sheet and return on investment. In addition the annual savings will be 25% of the inventory reduction or a total savings of $249,000.

In addition to the above savings, U.S. Stroller will achieve a shorter response time for customer orders. Customers will be able to order with a 2 week lead time instead of a 4 week lead time. While U.S. Stroller has been supplying the customers from its finished goods warehouses, the shorter lead time will provide advantages for promotions, special orders or other customer changes.

Option 1 will also improve quality through faster correction of problems in the plant. This will not only result in less rework and scrap, but will result in a better product being delivered to the customer. If the cost of quality, for example, is 30% of sales at U.S. Stroller and is reduced to 20% of sales, a savings of $456,000 per year will be realized. Thus, the quality savings could be even greater than the inventory savings. Of course, quality can be improved without JIT, but the two approaches reinforce each other.

Con's

Option 1 will cost $200,000 to implement in new assembly equipment, plus a large amount of time and other costs (including training, lost production time during changeover, management time, etc). Nevertheless, these costs could be paid off in one or two years depending on what savings are assumed for quality improvement, and inventory reduction.

The effort in implementing Option 1 should not be underestimated. Top management will have to be involved, everyone will have to be retrained, a lot of changes will be needed in procedures and systems by all departments. If this is not done with a careful and dedicated approach, a con could be failure to achieve the promised benefits of this option or failure of the effort.

6    ? The case says finished goods will be 15 to 30 days which is an average of 22 days.C-95

Another con is the loss of flexibility in option 2. Since equipment and final assembly lines are dedicated, changes in final demand cannot be easily handled. You cannot simply shift over from one product to another. One part of the plant could be put on overtime, while another part is idle, unless you have very high worker cross training. A second shift may even be needed in one part of the plant while other workers are on one shift. This can cause problems with the work force and is a result of the loss of flexibility.

Option 2 Cellular Approach: This option involves setting up a separate cell for each of the three products. Some additional equipment must be purchased because of the indivisibility’s of equipment and people must be dedicated to each of the product lines.

Pro's

The Pro's are the same as option one except things are carried one step further. Assuming all inventories can be reduced to 1 week, this option will reduce finished goods inventory from $210,000 to $67,000; work-in-process inventory from $161,000 to $80,000; and raw materials from $56,000 to $28,000. The annual savings of these reductions is an additional $62,000.

Option 2 will also improve quality even further because of quick visual feedback and cell stoppage when there is a problem. Further savings of as much as $450,000 per year can be realized if the Cost of Quality is reduced from 20% to 10%. The response time (lead time) of this option will be reduced to one week and could be even shorter if suppliers are located close by.

Con's

This option will cost more ($150,000 for additional equipment over option 1) plus other costs of moving equipment, training, etc. Nevertheless, this option would probably be attractive as an investment proposition because of relatively short payback.

The cellular approach results in even less production flexibility, because workers and equipment are totally dedicated. Thus the company must have confidence that demands will not fluctuate too wildly from week to week. The case indicates that weekly demand fluctuates 25% from averages. This amount of demand change could certainly be absorbed by safety stock. But the presence of safety stock introduces more costs into the system.

Question 3: What will be the impact of these options on the MRP system currently in use? The MRP system will have to be dramatically changed. In both options a pull system will be used, so that MRP cannot be used to control the shop floor. The most that MRP could be used for is for "Major Event Planning". In this case orders would be launched for Major Events such as changes in demand or product changes. MRP can be useful as a planning tool, but not for control of the actual production process itself.

Major event planning implies a skeleton (Type I) MRP system. The master schedule for a few weeks in advance (say 2 to 4 weeks) would be exploded through the bill of materials and planning orders would then be given to work center supervisors and suppliers for planning purposes only. The actual order to produce will come from the Kanban Pull system that is dependent on what is happening on the shop floor.

C-96

MRP could be scrapped altogether, especially if U.S. Stroller believes that schedule changes will be minimal. If approximately the same amount is produced from week to week and there are few engineering change orders, there is really no need for MRP. Any changes can be "manually" exploded and advance notices sent to suppliers and work centers, especially for the simple product line at U.S. Stroller.

Question 4.

I recommend that the company proceed with Option 1 and then move to Option 2, if demand and product changes are relatively stable. It appears in this case that Option 1 can be financially justified and that substantial benefits will flow to the company in profits, balance sheet improvements, and in holding market share. After implementing option 1, they will have time to evaluate whether a further move to option 2 is also desirable.

Teaching Strategy

When I teach this case, I follow the questions in the book. This provides a framework for analyzing and discussing the case. I ask a student to start by summarizing the current situation that U.S. Stroller is facing in the market place. Then I lead the discussion toward evaluating the pro's and con's of each of the options. An interesting discussion can be held on the usefulness of MRP in this situation. Some students are surprised to learn that MRP may no longer be needed.

I use this case to illustrate how JIT and GT can be evaluated. Along the way the students also gain a better understanding of how a traditional batch oriented company can make the changes needed to move to JIT or cells. The case requires the students to think in depth about the pro's and con's of these approaches - not only the pro's.

C-97