Operations Management for Competitive Advantage 11e Ch05

Chase−Jacobs−Aquilano: Operations Management for Competitive Advantage, 11th Edition

II. Process Selection and Design

Introduction © The McGraw−Hill Companies, 2005

section two

section



Introduction © The McGraw−Hill Companies, 2005

sec

tion

2PROCESSSELECTIO

NANDDESIG

N

PROCESS SELECTIONAND DESIGN

5. Process Analysis

Technical Note 5: Job Design and Work Measurement

6. Manufacturing Process Selection and Design

Technical Note 6: Facility Layout

7. Service Process Selectionand Design

Technical Note 7: Waiting Line Management

8. Quality Management: Focus on Six Sigma

Technical Note 8: Process Capability and Statistical

Quality Control

9. Operations Consulting andReengineering

2H O W T O B E C O M E A N E F F I C I E N C Y E X P E R T

MAYBE BECOMING AN EFFICIENCY EXPERT IS NOT YOUR

dream, but it is important to learn the funda-

mentals. Have you ever wondered why you

always have to wait in line at one store, but

another one seems to be on top of the crowds?

The key to serving customers well, whether with

products or services, is having a great process.

We use processes to do most things. You proba-

bly have a regular process that you use every

morning. What are the tasks associated with your

process? Do you brush your teeth, take a shower,

dress, make coffee, and read the paper? Have you

ever thought about how the tasks should be

ordered or what is the best way to execute each

task? This section is about designing efficient

processes—all kinds of processes. Companies

also need to develop a quality philosophy and

integrate it into their processes. Actually, quality

and process efficiency are closely related. Have

you ever done something, but then had to do it

again because it was not done properly the first

time? This section considers these subjects in

both manufacturing and service industries.



5. Process Analysis © The McGraw−Hill Companies, 2005

chapterPROCESS ANALYSIS

chap

ter

5PROCESSANALYSIS

154 Process AnalysisAnalyzing a Las Vegas slot machine Process defined

Cycle time defined

Utilization defined

157 Process Flowcharting

158 Types of ProcessesBuffering, blocking, starving, and bottleneck defined

Make-to-order, make-to-stock, and hybrid processes defined

Pacing defined

162 Measuring Process PerformanceProductivity and efficiency defined

Run time, setup time, and operation time defined

Throughput time and throughput rate defined

Process velocity or throughput ratio defined

Value-added time defined

Little’s Law defined

165 Process Analysis ExamplesA bread-making operationA restaurant operationPlanning a transit bus operation

170 Process Throughput Time Reduction

171 Conclusion

177 Case: Analyzing Casino Money-Handling Processes

178 Case: Kristen’s Cookie Company (A)




C U S T O M E R - D R I V E N S E R V I C E F O RM C D O N A L D ’ S5 S E L F - O R D E R I N G K I O S K S A N D

E N V I R O N M E N TIDEO collaborated with McDonald’s on the first generation of a new service

system in their Lone Tree restaurant, south of Denver. The new system allows

McDonald’s customers to place their orders without assistance, providing

improved flexibility, speed, accuracy, and convenience to both McDonald’s

customers as well as its crews. The system consists of touch-screen self-order

kiosks at the front counter and in the children’s PlayPlace area that have been

fully integrated into the McDonald’s physical environment, operational flow,

and brand message.




Customers place their orders using an icon-based system and pay at the kiosk or at the

pick-up counter. After placing their orders, customers pick up their food at the counter by

showing the order number on their printed receipts. In the PlayPlace area, parents can place

and pay for their orders while supervising their children. A McDonald’s crewmember then

delivers the food to their table.

This new model needed to work within the popular and highly efficient system in use

today. The completed design spanned the entire ordering experience and not just the kiosks

themselves. The team updated the restaurant’s graphics, signage, counters, and crew uniforms,

and created nine self-order kiosks with a fully developed icon-based menu system. All design

elements plus the in-store layout of the new service experience were arranged to comple-

ment the traditional experience of ordering at the counter.

The work began with a national survey of all kinds of quick-serve and self-serve experi-

ences and distilled behavioral patterns of McDonald’s customers to guide the design work.

Since its launch and after thousands of transactions, the new service has had a high customer

adoption rate with virtually no lines. –>

154 section 2 PROCESS SELECTION AND DESIGN

Understanding how processes work is essential to ensuring the competitivenessof a company. A process that does not match the needs of the firm will punish the firm everyminute that the firm operates. Take, for example, two fast-food restaurants. If one restaurantcan deliver a quarter-pound hamburger to the customer for $.50 in direct costs and a secondrestaurant costs $.75, no matter what the second restaurant does, it will lose $.25 in profit forevery hamburger it sells compared to the first restaurant. Many factors need to be consideredwhen one sets up the process to make those hamburgers. These factors include the cost of theraw materials, the costs associated with how the hamburger is prepared, and the cost of tak-ing the order and delivering it to the customer.

What is a process? Aprocess is any part of an orga-nization that takes inputs andtransforms them into outputsthat, it is hoped, are of greatervalue to the organization thanthe original inputs. Considersome examples of processes.Honda Motors assembles theAccord in a plant inMarysville, Ohio. The assem-bly plant takes in parts andcomponents that have beenfabricated for the plant. Usinglabor, equipment along anassembly line, and energy,these parts and components are transformed into automobiles. McDonald’s, at each of itsrestaurants, uses inputs such as hamburger meat, lettuce, tomatoes, and potatoes. To theseinputs, trained labor is added in the form of cooks and order takers, and capital equipment isused to transform the inputs into hamburgers, french fries, and other foods.

In both of these examples, the process produces products as output. However, the outputsof many processes are services. In a hospital, for example, specialized equipment and highlytrained doctors, nurses, and technicians are combined with another input, the patient. The

P R O C E S S A N A L Y S I S

Service

Process




patient is transformed through proper treatment and care into a healthy patient. An airline isanother example of a service organization. The airline uses airplanes, ground equipment,flight crews, ground crews, reservation personnel, and fuel to transport customers betweenlocations all over the world.

This chapter describes how to analyze a process. Analyzing a process allows someimportant questions to be answered, such as these: How many customers can the process han-dle per hour? How long will it take to serve a customer? What change is needed in the processto expand capacity? How much does the process cost? A difficult, but important, first step inprocess analysis is to clearly define the purpose of the analysis. Is the purpose to solve a prob-lem? Is it to better understand the impact of a change in how business will be done in the future?

Clearly understanding the purpose of the analysis is critical to setting the level of detail inmodeling the process. The analysis must be kept as simple as possible. The following sectionsof this chapter discuss the details of constructing flowcharts and measures that are appropri-ate for different types of processes. But first, consider a simple example.

A N A L Y Z I N G A L A S V E G A S S L O T M A C H I N EThe slot machine is common in casinos around the world. Let’s use this machine to illustratehow a simple process is analyzed.

Assume that we work for the casino and management is considering a new type of elec-tronic slot machine that is much faster than the current mechanical machine. Management hasasked how much we can expect to make from the new electronic machine over a 24-hourperiod compared to the old mechanical machine.

Begin by analyzing a mechanical slot machine. The slot machine is activated when thecustomer puts one or more coins in the machine and then pulls the arm on the machine (slotmachines are often called “one-armed bandits”). Three wheels spin, and after a time eachwheel stops and displays a particular symbol. The machine pays money when certain combi-nations of symbols simultaneously appear. For those not familiar with how a slot machineworks, we have included a slot machine simulation program on the CD included with thebook. Sorry, but it does not pay real money.

Slot machines are designed to pay back a certain percentage of what they take in. Typicalpaybacks would be 90 to 95 percent of what is taken in; the casino keeps 5 to 10 percent.These payback percentages are a function of the number of different symbols that are on eachwheel. Each symbol is repeated on each wheel a certain number of times. For example, if a

PROCESS ANALYSIS chapter 5 155

SLOT MACHINES IN A CASINO IN RENO,NEVADA, HAVE A MAJOR IMPACT ON THE

CASINO’S PROFITS. AN ELECTRONIC VERSION

VERSUS A MECHANICAL VERSION IMPACTS

CYCLE TIME, WHICH IN TURN AFFECTS

REVENUES.

Service




wheel has 10 symbols, one might be a single bar, one a double bar, and one a lemon; twomight be cherries, three lucky sevens, and two liberty bells. Because the wheels stop on arandom symbol, the probability of lucky sevens coming up on all three wheels is310 × 3

10 × 310 = 0.027 or 2.7 percent of the time. The probability of certain combinations of

symbols coming up, combined with the payout for each combination, sets the average per-centage that the machine is expected to pay out.

Consider a mechanical slot machine that pays out 95 percent of the coins played. With thismachine, assume the average player feeds coins into the machine at a pace of one coin each15 seconds. This 15-second interval is called the cycle time of the process. The cycle time ofa repetitive process is the average time between completions of successive units. In the caseof the slot machine, the unit is a silver dollar. With a 15-second cycle time, our mechanicalslot machine can process $4 (60 seconds/15 seconds) per minute or $240 ($4/minute ×60 minutes) per hour. Because our slot machine has a payout of 95 percent, we would expectthe machine to give the customer 228 (240 × 0.95) of the silver dollars that it took in andkeep $12 for the casino for each hour that it is in operation. If we started with $100, we couldexpect to play for about 8.3 hours ($100/$12 per hour) before we would run out of silver dol-lars. We might be lucky and win the jackpot, or we might be unlucky and lose it all in the firsthour; but on average we would expect to lose the entire $100 in 8.3 hours.

Now consider the new electronic slot machine. It operates in exactly the same manner; theonly difference is that it processes coins in 10 seconds. With a 10-second cycle time, the machineprocesses $6 (60 seconds/10 seconds) per minute or $360 ($6/minute × 60 minutes) per hour.With a 95 percent payout, the machine would give the customer back 342 (360 × 0.95) silverdollars and keep $18 for the casino each hour. This machine would take our $100 in only5.5 hours ($100/$18 per hour).

So how much does the electronic slot machine make for the casino in 24 hours comparedto the mechanical slot machine? One more critical piece of information is needed to answerthis question: How long will the slot machine operate over the 24 hours? The casino feels thatthe machine will be used 12 out of the 24 hours; this 12 out of 24 hours is the expected uti-lization of the machine. Utilization is the ratio of the time that a resource is actually activatedrelative to the time that it is available for use. Adjusting for utilization, the expected revenuefrom the mechanical machine is $144/day ($12/hour × 24 hours × 0.5) compared to revenueof $216/day ($18/hour × 24 hours × 0.5) for the electronic machine. When an analysis is per-formed, it is important to qualify the analysis with the assumptions made. In this comparison,we assumed that the operator puts only one silver dollar in the machine at a time and that theutilization would be the same for the mechanical and electronic slot machines.

The speed of the slot machine can have a major impact on the casino’s revenue. The sin-gle slot machine is only a small part of the casino. To really understand how much revenuethe casino can generate, we need to consider all of the other revenue-generating processessuch as the blackjack and poker tables, keno games, craps, and the other games in the casino.Many times analyzing an enterprise involves evaluating a number of independent activities,like our slot machine. The aggregate performance of each individual activity may be all thatis needed to understand the overall process. On the other hand, often there is significantinteraction between individual activities or processes that must be considered.

Think about our gambling casino. Many casinos offer great deals on food, which is servedright in the casino. What do you think would be the main priority of the food operation man-ager in one of these casinos? Would great-tasting food be important? How important is thecost of the food? Is speed of service important? Good food certainly is important. If the foodis unpleasant, the customer will not even consider eating at the casino. This is bad for thecasino because if the customers leave, they take their money with them. Remember, the casi-no makes money based on how long the customers gamble. The more time spent gambling,the more money the casino makes. What about cost? If the customers think the meals are tooexpensive, they might leave. So it is important to keep the cost of the meals down so that theycan be priced inexpensively. Many casinos even give meals away. How important is it toserve the customer quickly? Think about it this way: Every minute that the customers aresitting in the restaurant, they are not feeding silver dollars into a slot machine. So speed isimportant because it impacts the revenue generated at the games in the casino.


Utilization

Cycle time





Often the activities associated with a process affect one another so that it isimportant to consider the simultaneous performance of a number of activities, all operating atthe same time. A good way to start analyzing a process is with a diagram showing the basicelements of a process—typically tasks, flows, and storage areas. Tasks are shown as rectan-gles, flows as arrows, and the storage of goods or other items as inverted triangles. Sometimesflows through a process can be diverted in multiple directions depending on some condition.Decision points are depicted as a diamond with the different flows running from the pointson the diamond. Exhibit 5.1 displays examples of these symbols. It sometimes is useful toseparate a diagram into different horizontal or vertical bands. This allows the separation oftasks that are part of the process. For example, with the slot machine, the tasks performed bythe customer can be separated from the tasks performed by the slot machine.

In the slot machine example, the level of abstraction considers the slot machine as a sim-ple black box that takes in silver dollars and either keeps them or returns some of them duringeach cycle. Viewing the slot machine as a black box might be fine if the purpose is just to ana-lyze how much the machine is expected to make for the casino each hour. In reality, moreactivities are required to support the slot machine. Inside the slot machine are two buckets ofsilver dollars. One bucket stores coins needed for internal use by the slot machine. When acustomer wins, the payout comes from this payout bucket. The slot machine is designed to

P R O C E S S F L O W C H A R T I N G

e x h i b i t 5 . 1

WinningsbucketMove coin

to winningsbucket

Move cointo payout

bucket

Payoutbucket

Ispayoutbucketfull?

Activatepayout

Paywinnings

Pull armon slot

machineQuit

Winor

lose

Insert silverdollar into

slot machine

Internal slot machineactivities

Player activities

Lose

Win

Yes

Yes

No

No

Playagain

Flowchart Symbols

Tasks or operations Decision points

Storage areas orqueues (waiting lines)

Flows of materialor customers

Process Flowchart Example




automatically keep this payout bucket filled during play. When the payout bucket is full, thesilver dollars are deposited in a second winnings bucket. The winnings bucket must be peri-odically emptied to claim the winnings for the casino. The flowchart in Exhibit 5.1 depictsthe external activities of the player and the internal movement of the coins within themachine.

Probably the most interesting thing about the payout bucket is how big it should be. Theslot machine is programmed so that if the payout bucket empties, the machine stops and lightson the top of the machine flash, thus notifying casino personnel that a lucky customer hasemptied the machine. The payout bucket would be sized to keep this a rare occurrence. Thinkof the payout bucket as a buffer or intermediate storage area for silver dollars that allows theslot machine to operate on its own. The smaller the payout bucket, the more the casino per-sonnel need to attend to the machine, and the more time the machine is idle for lack of silverdollars. On the other hand, with a larger bucket more money is tied up.

The situation with the winnings bucket in the machine is similar. A small winnings bucketwill need to be emptied more often. On the other hand, a large winnings bucket means thatthe casino does not deposit the money into its bank account as quickly. It is easy to see theadvantage of buffering operations with the slot machine. Large buffers allow the process tooperate independently, whereas small buffers require more attention. In the case of the slotmachine, the buffer is composed of the silver dollars. In other situations, where the buffer isother items such as a raw material, these items have a value, so they also represent money.

Consider a slot machine that we expect to deposit $12 into the winnings bucket every hour.If our winnings bucket can hold 1,000 silver dollars, then we expect to need to empty the win-nings bucket every 83.3 hours ($1,000/$12 per hour) if the slot machine is used 100 percentof the time. It is interesting to think about what happens when the winnings bucket fills up. Ifthe slot machine is smart enough to know that the winnings bucket is full, it might be pro-grammed to just stop working with its lights flashing as they do when the payout bucket emp-ties. This would cause downtime on the machine and might upset a customer using themachine because the customer would have to move to another slot machine. If the slot machinewere not programmed to stop working, the silver dollars would fill the cavity where the buck-et is located in the base of the machine. Imagine the mess when the casino worker opens thatoverflowed machine and all those silver dollars come pouring out. How often would you planon emptying the winnings bucket?

An easy way to draw flowcharts is to use the Drawing toolbar available in the MicrosoftOffice programs (i.e., Word, Excel, and PowerPoint). To access this toolbar, go to View –>Toolbars and click on “Drawing”. The Drawing menu should then appear. Use the Auto-Shapes button and then select “Flowchart”. This will display a number of flowchart symbolsto use to create your flowchart. Text can be added by selecting a symbol and then clicking onthe right mouse button. Select “Add text” to insert text in the symbol. The symbols can beconnected by using “Connectors” available from the AutoShapes button. Nice flowcharts canbe made using these tools.


It is useful to categorize processes to describe how a process is designed. Bybeing able to quickly categorize a process, we can show the similarities and differencesbetween processes.

The first way to categorize a process isto determine whether it is a single-stage ora multiple-stage process. If the slotmachine were viewed as a simple blackbox, it would be categorized as a single-stage process. In this case, all of the activi-ties that are involved in the operation of the slot machine would be collapsed and analyzedusing a single cycle time to represent the speed of the slot machine. A multiple-stage process

T Y P E S O F P R O C E S S E S

Multistage process

Stage 1 Stage 2 Stage 3




has multiple groups of activities that are linked through flows. The term stage is used to indi-cate that multiple activities have been pulled together for analysis purposes.

A multiple-stage process may be buffered internally. Buffering refers to a storage areabetween stages where the output of a stage is placed prior to being used in a downstreamstage. Buffering allows the stages to operate independently. If one stage feeds a second stagewith no intermediate buffer, then the assumption is that the two stages are directly linked.When a process is designed this way, the most common problems that can happen are block-ing and starving. Blocking occurs when the activities in the stage must stop because there isno place to deposit the item just completed. Starving occurs when the activities in a stagemust stop because there is no work.

Consider a two-stage process where the first stage has a cycle time of 30 seconds and thesecond a cycle time of 45 seconds. If this process needs to produce 100 units, then for eachunit produced, the first stage would be blocked for 15 seconds.

What would happen if an inventory buffer were placed between the two stages? In thiscase, the first stage would complete the 100 units in 3,000 seconds (30 seconds/unit ×100 units). During these 3,000 seconds, the second stage would complete only 66 units((3,000 − 30) seconds/45 seconds/unit). The 30 seconds are subtracted from the 3,000 sec-onds because the second stage is starved for the first 30 seconds. This would mean that the

inventory would build to 34 units(100 units − 66 units) over that first 3,000seconds. All of the units would be pro-duced in 4,530 seconds. The second stagein this case is called a bottleneck becauseit limits the capacity of the process.

What would happen if the first stagerequired 45 seconds and the second stagehad the 30-second cycle time? In this casethe first stage would be the bottleneck,and each unit would go directly from thefirst stage to the second. The second stagewould be starved for 15 seconds waitingfor each unit to arrive; however, it wouldstill take 4,530 seconds to complete all100 units. All of this assumes that there isno variability in the cycle time. With therelatively low 67 percent utilization on thesecond stage, variability would have littleimpact on the performance of this system,but if the cycle times were closer, someinventory might collect in the buffer.

Often activities, stages, and even entireprocesses are operated in parallel. Forexample, operating two identical activi-ties in parallel would theoretically doublecapacity. Or perhaps two different sets ofactivities can be done at the same time onthe unit being produced. In analyzing asystem with parallel activities or stages, itis important to understand the context. Inthe case where parallel processes repre-sent alternatives, for example, a diamondshould show that flows divert and whatpercentage of the flow moves in eachdirection. Sometimes two or moreprocesses terminate in a common inven-tory buffer. This normally indicates that


Buffering

BlockingStarving

Multistage process with buffer

Stage 1 Stage 2Buffer

Alternative paths

50%

50%

Different products produced

Simultaneous activities

Bottleneck




the two processes make identical items that are going into this inventory. Separate invento-ries should be used in the diagram if the outputs of the parallel processes are different.

Another useful way to characterize a process is whether the process makes to stockor makes to order. To illustrate these concepts, consider the processes used to make ham-burgers at the three major fast-food restaurant chains in the United States: McDonald’s,Burger King, and Wendy’s. In the case of McDonald’s, in 1999 the company converted to anew make-to-order process, but the company has now revised that into a “hybrid” system. Webegin our tour of the approaches used by the top fast-food restaurants by first reviewing thetraditional approach.

Consider a traditional restaurant making hamburgers. Before the era of fast food, ham-burgers were always made to order. In the traditional process, the customer places an orderspecifying the degree of doneness (medium or well done) and requests specific condiments(pickles, cheese, mustard, onions, catsup). Using this specification, the cook takes raw ham-burger meat from inventory (typically this inventory is refrigerated and the patties havealready been made), cooks the hamburger, and warms the bun. The hamburger is then assem-bled and delivered to the customer. The quality of the hamburger is highly dependent on theskill of the cook.

This make-to-order process is activated only in response to an actual order. Inventory(both work-in-process and finished goods) is kept to a minimum. Theoretically, one wouldexpect that response time would be slow because all the activities need to be completedbefore the product is delivered to the customer. Services by their very nature often use make-to-order processes.

McDonald’s revolutionized the hamburger-making process by developing a high-volumeapproach. A diagram of McDonald’s traditional process is shown in Exhibit 5.2A. Untilrecently, hamburgers were grilled in batches. Standard hamburgers (for example, the“Big Mac” consists of two beef patties, sauce, lettuce, cheese, pickles, and onion on a sesameseed bun) were then prepared and stored in a holding bin for immediate delivery to the cus-tomer. A person that judged current demand and placed orders to keep inventory in the bin atan appropriate level controlled the whole process. This is a highly efficient make-to-stockprocess that produces standard products that can be delivered quickly to the customer. Thisquick process appeals to families with small children, for whom speed of delivery isimportant.

In general, a make-to-stock process ends with finished goods inventory; customer ordersare then served from this inventory. A make-to-stock process can be controlled based on theactual or anticipated amount of finished goods inventory. A target stocking level, for exam-ple, might be set, and the process would be periodically activated to maintain that targetstocking level. Make-to-stock processes are also used when demand is seasonal. In this case,inventory can be built during the slow season and used during the peak season, thus allowingthe process to run at a constant rate throughout the year.

The unique feature of the Burger King process, shown in Exhibit 5.2B, is a highlyspecialized conveyor–broiler. Raw hamburger patties are placed on a moving conveyor thatruns through a flaming broiler. In exactly 90 seconds, the patties are cooked on both sideswith a unique broiler taste. Due to the fixed time for a patty to move through theconveyor–broiler, the thickness of the patties must be the same for all the hamburger prod-ucts. The buns are also warmed on a conveyor. This system results in a unique, highlyconsistent product. The cooked patties are stored in a warmed storage container. During peri-ods of high demand, some standard hamburgers are prepared and inventoried for immediatedelivery. Custom hamburgers with unique combinations of condiments are prepared to order.This hybrid process provides flexibility to respond to customer preferences through theassemble-to-order backend process—thus, the Burger King “have it your way” slogan. Ingeneral, hybrid processes combine the features of both make-to-order and make-to-stock.Here two types of process are parallel alternatives at the end of the Burger King process. Inthe most common hybrid form, a generic product is made and stocked at some point in theprocess. These generic units are then finished in a final process based on actual orders.

Continuing with our tour, Wendy’s uses a make-to-order process (as shown inExhibit 5.2C) that is in full view of the customer. Hamburger patties are cooked on a grill.


Make-to-order

Make-to-stock

Hybrid

Service




During high-volume times, the cook tries to get a little ahead and anticipates the arrival ofcustomers. Patties that are on the grill too long are used in the chili soup. On arrival of a customerorder, a patty is taken from the grill and the hamburger is assembled to the exact specificationsof the customer. Because the process starts with the cooking of the patty, it is a little slower. Thecustomer can see what is going on, and the perception is of a high-quality custom product.

Finally, the new McDonald’s process introduced in 1999 (Exhibit 5.2D) is a hybridprocess. Cooked hamburger patties are inventoried in a special storage device that maintains


e x h i b i t 5 . 2

A. McDonald’s—Old Process

B. Burger King

Customor

standard?

Standard

Custom

D. McDonald’s—New Process

C. Wendy’s

Cook Assemble Deliver

Customerplaces order

CookCustomer

places order Deliver

Assemble

Assemble





Rawmaterial

Finishedgoods

Rawmaterial WIP

Finishedgoods

Rawmaterial

Chili

Rawmaterial WIP

Making Hamburgers at McDonald’s, Burger King, and Wendy’s




the moistness of the cooked patties for up to 30 minutes. The process makes extensive use ofthe latest cooking technologies. Hamburger patties are cooked in less than 45 seconds. Bunsare toasted in only eleven seconds. Individual items on each customer order are transmittedimmediately to the area where the hamburgers are assembled using a specially designed com-puter system. The assembly process that includes toasting the buns is designed to respond toa customer order in only 15 seconds. By combining the latest technology and clever processengineering, McDonald’s has developed a very quick response process. The product is fresh,delivered quickly, and made to the exact specifications of the customer.

Each of the processes used by these companies has its strengths and weaknesses.McDonald’s is the high-volume leader, catering to families with young children. Burger Kinghas its unique taste. Wendy’s appeals to those who want their hamburgers prepared the old-fashioned way. Each company focuses advertising and promotional efforts toward attractingthe segment of the market their process characteristics best support.

One final method for categorizing a process is by whether it is paced or nonpaced. Recallthat Burger King uses the conveyor–broiler to cook hamburgers in exactly 90 seconds.Pacing refers to the fixed timing of the movement of items through the process. In a serialprocess, the movement of items through each activity (or stage) is often paced in somemechanical way in order to coordinate the line. An assembly line may, for example, moveevery 45 seconds. Another mechanism used is a clock that counts down the amount of timeleft in each cycle. When the clock reaches zero, the parts are manually moved to the nextactivity. Dividing the time available to produce a certain product by customer demand forthe product calculates the required cycle time for a process. For example, if an automo-bile manufacturer needs to produce 1,000 automobiles during a shift where the assemblyline operates 420 minutes, the cycle time is 25.2 seconds (420 minutes/1,000 automobiles ×60 seconds/minute = 25.2 seconds/automobile).


There is much variation in the way performance metrics are calculated in prac-tice. This section defines metrics in a manner consistent with the most common use in practice.It is vital, though, to understand exactly how a metric coming from a particular company orindustry is calculated prior to making any decisions. It would be easier if metrics werecalculated more consistently, but this just is not the case. So if a manager says that his utiliza-tion is 90 percent or her efficiency is 115 percent, a standard follow-up question is “How didyou calculate that?” Metrics often are calculated in the context of a particular process. Metricsused in cases that you are studying may be defined slightly differently from what is givenhere. It is important to understand, within the context of the case, how a term is being used.

Comparing the metrics of one company to another, often referred to as benchmarking, isan important activity. Metrics tell a firm if progress is being made toward improvement.Similar to the value of financial measures to accountants, process performance metrics givethe operations manager a gauge on how productively a process currently is operating and howproductivity is changing over time. Often operations managers need to improve the perfor-mance of a process or project the impact of a proposed change. The metrics described in thissection are important for answering these questions. To help in understanding these calcula-tions, Exhibit 5.3 shows how these metrics relate to one another.

Possibly the most common process metric is utilization. As discussed earlier in the chapter,utilization is the ratio of the time that a resource is actually being used relative to the time thatit is available for use. Utilization is always measured in reference to some resource—forexample, the utilization of direct labor or the utilization of a machine resource. The distinc-tion between productivity and utilization is important. Productivity is the ratio of output toinput. Total factor productivity is usually measured in monetary units, dollars for example, bytaking the dollar value of the output (such as goods and services sold) and dividing by the costof all the inputs (that is, material, labor, and capital investment). Alternatively, partial factorproductivity is measured based on an individual input, labor being the most common. Partial

M E A S U R I N G P R O C E S S P E R F O R M A N C E

Pacing

Productivity




factor productivity answers thequestion of how much outputwe can get from a given levelof input; for example, howmany computers are made peremployee working in the com-puter manufacturing plant?(See Chapter 2 for additionalinformation about produc-tivity.) Utilization measuresthe actual activation of theresource. For example, what isthe percentage of time that anexpensive machine is actuallyoperating?

Efficiency is a ratio of theactual output of a process rela-tive to some standard. Forexample, consider a machinedesigned to package cereal ata rate of 30 boxes per minute.If during a shift the operatorsactually produce at a rate of36 boxes per minute, then theefficiency of the machine is120 percent (36/30). An alter-native way that the term efficiency is used is to measure the loss or gain in a process. Forexample, if 1,000 units of energy are put into a process designed to convert that energy tosome alternative form, and the process produces only 800 units of energy in the new form,then the process is 80 percent efficient.

Run time is the time required to produce a batch of parts. This is calculated by multiply-ing the time required to produce each unit by the batch size. The setup time is the time


Efficiency

e x h i b i t 5 . 3

Setuptime

Batchsize

Time/unit

Runtime

Operationtime

Throughputtime

Queuetime

Throughput time � Average time for aunit to move throughthe system

Cycle time � Average time betweencompletion of units

Operation time � Setup time � Run time

Velocity �Throughput time

Value-added time

Throughput rate �1

Cycle time

Efficiency �Actual output

Standard output

Productivity �Output

Input

Utilization �Time activated

Time available

Throughputrate

InputsEfficiency Productivity

Timeavailable

UtilizationTime

activated

Cycletime

VelocityStandards

Process Performance Metrics

Run timeSetup time




required to prepare a machine to make a particular item. Machines that have significant setuptime will typically run parts in batches. The operation time is the sum of the setup time andrun time for a batch of parts that are run on a machine. Consider the cereal-boxing machine(on page 163) that is designed to produce at a rate of 30 boxes per minute. The run time foreach box is 2 seconds. To switch the machine from 16-ounce boxes to 12-ounce boxesrequires a setup time of 30 minutes. The operation time to make a batch of 10,000 12-ounceboxes is 21,800 seconds (30 minutes’ setup × 60 seconds/minute + 2 seconds/box ×10,000 boxes) or 363.33 minutes.

In practice, often setup time is not included in the utilization of the process. In essence,setup time is categorized like the downtime caused by repair or some other disruption to theprocess. This assumption can vary from company to company, so it is important when com-paring the utilization of a machine or other resource to understand exactly how the companycategorizes setup time.

The cycle time (also defined earlier in this chapter) is the elapsed time between startingand completing a job.1 Another related term is throughput time. Throughput time includesthe time that the unit spends actually being worked on together with the time spent waiting ina queue. As a simple example, consider a paced assembly line that has six stations and runswith a cycle time of 30 seconds. If the stations are located one right after another and every30 seconds parts move from one station to the next, then the throughput time is three minutes(30 seconds × 6 stations/60 seconds per minute). The throughput rate is the output rate thatthe process is expected to produce over a period of time. The throughput rate of the assemblyline is 120 units per hour (60 minutes/hour × 60 seconds/minute ÷ 30 seconds/unit). In thiscase, the throughput rate is the mathematical inverse of the cycle time.

Often units are not worked on 100 per-cent of the time as they move through aprocess. Because there often is some vari-ability in the cycle time of a process, buffersare incorporated in the process to allow indi-vidual activities to operate independently,at least to some extent. In the six-stationassembly line just described, consider theimpact of having 10 additional buffer posi-tions along the line. Assume that two ofthese positions are between the first and sec-

ond workstation, two are between stations 2 and 3, and so forth. If these positions are alwaysoccupied, then the throughput time would be eight minutes (assuming a total of 16 positionsalong the assembly line and an average cycle time of 30 seconds).

Process velocity (also known as throughput ratio) is the ratio of the total throughput timeto the value-added time. Value-added time is the time in which useful work is actually beingdone on the unit. Assuming that all of the activities that are included in the process are value-added activities, value-added time should be the sum of the activity operation times in theprocess. The process velocity (or throughput ratio) for our assembly line with the 10 addi-tional buffer positions, assuming the positions are used 100 percent of the time, is 2.66(8 minutes/3 minutes).

Little’s Law2 states a mathematical relationship between throughput rate, throughput time,and the amount of work-in-process inventory. Little’s Law estimates the time that an item willspend in work-in-process inventory, which can be useful for calculating the total throughputtime for a process. Using the terminology defined in this section, Little’s Law is defined asfollows:

Throughput time = Work-in-process

Throughput rate

Notice how this law holds for the example of the assembly line without the buffer inventory.If the assembly line has six stations with one unit of work-in-process at each station, and thethroughput rate is two units per minute (60 seconds/30 seconds per unit), then the throughput


2Units1

Unit

2Units1

Unit

2Units1

Unit

2Units1

Unit

2Units1

Unit1

Unit

Process velocity(throughput ratio)

Value-added time

Little’s Law

Throughput rate

InteractiveOperations

Management

Throughput time

Operation time




time is three minutes (6 units/2 units per minute). In general, this equation is useful when twoof the three quantities are known. For example, if the throughput time and throughput rateare known, the work-in-process can be calculated. This formula holds for any process that isoperating at a steady rate.

By steady rate we mean that work is entering and exiting the system at the same rate overthe time period being analyzed. Our assembly line has 120 units entering and 120 units exit-ing the process each hour. If, for example, 150 units were entering the system each hour butonly 120 units were exiting, then the system would not be operating at a steady rate since 30additional units would be accumulating in the system each hour. These 30 units add to work-in-process, which would cause the throughput time to increase each hour. The actual increasein throughput time would be 15 minutes per hour (30 units�120 units per hour � 0.25 hour).Another example of how Little’s Law can be applied is in the following section relative todiagnosing the performance of a process.


In this section the concepts described thus far in the chapter are illustrated withthree examples. These examples are typical of the types of analysis that are performed inmanufacturing, services, and logistics businesses. Keep in mind that the analysis used in eachexample can be applied to many different contexts. Be creative in applying something thatyou have seen in another context to the problem at hand. The first example analyzes a bread-making process. Following this, a restaurant operation is evaluated. Finally, a typical logis-tics operation is appraised.

A B R E A D - M A K I N G O P E R A T I O N 3

EXAMPLE 5.1: Bread MakingFor the manager of a bakery, a first priority is to understand the products that are made and theprocess steps required. Exhibit 5.4A is a simplified diagram of the bread-making process. Thereare two steps required to prepare the bread. The first is preparing the dough and baking the loaves,here referred to as bread making. The second is packaging the loaves. Due to the size of the mixers

P R O C E S S A N A L Y S I S E X A M P L E S

e x h i b i t 5 . 4

A. Bread making on one line

B. Bread making on two parallel lines

Rawmaterial Bread making

Cycle time:1 hour/100 loaves

WIP FinishedGoods

Pack

Cycle time: hour/100 loaves

Rawmaterial

Bread making


Bread making


WIP FinishedGoods

Pack

Cycle time: hour/100 loaves34

34

Bread-Making Processes




in the bakery, bread is made in batches of 100 loaves. Bread making completes a batch of 100 loavesevery hour, which is the cycle time for the activity. Packaging needs only 0.75 hour to place the100 loaves in bags.

From this we see that bread making is the bottleneck in the process. A bottleneck is the activity in aprocess that limits the overall capacity of the process. So if we assume that the bread-making and pack-aging activities both operate the same amount of time each day, then the bakery has a capacity of100 loaves per hour. Notice that over the course of the day the packaging operation will be idle forquarter-hour periods in which the next batch of bread is still being made but packaging has already com-pleted bagging the previous batch. One would expect that the packaging operation would be utilizedonly 75 percent of the time under this scenario.

Suppose that instead of having only one bread-making operation we now have two, as shown inExhibit 5.4B. The cycle time for each individual bread-making operation is still one hour per100 loaves. The cycle time for the two bread-making lines operating together is half an hour. Becausethe packaging operation takes 0.75 hour to bag 100 loaves, the packaging operation now is the bottle-neck. If both bread making and packaging were operated the same number of hours each day, it wouldbe necessary to limit how much bread was made because we do not have the capacity to package it.However, if we operated the packaging operation for three eight-hour shifts and bread making for twoshifts each day, then the daily capacity of each would be identical at 3,200 loaves a day (this assumesthat the packaging operation starts up one hour after the bread-making operation). Doing this requiresbuilding up a shift’s worth of inventory each day as work-in-process. Packaging would bag this duringthe third shift. So what is the throughput time of our bakery?

SOLUTIONIn the original operation with just the single bread-making process, this is easy to calculate becauseinventory would not build between the bread-making and packaging processes. In this case the through-put time would be 1.75 hours. In the case where we operate the packaging operation for three shifts, theaverage wait in work-in-process inventory needs to be considered. If both bread-making operations startat the same time, then at the end of the first hour the first 100 loaves move immediately into packagingwhile the second 100 loaves wait. The waiting time for each 100-loaf batch increases until the bakingis done at the end of the second shift.

This is a case where Little’s Law can estimate the time that the bread is sitting in work-in-process.To apply Little’s Law we need to estimate the average work-in-process between bread making and pack-aging. During the first two shifts inventory builds from 0 to 1,200 loaves. We can estimate the averagework-in-process over this 16-hour period to be 600 loaves (half the maximum). Over the last eight-hourshift inventory drops from the 1,200-loaf maximum down to 0. Again the average work-in-process is600 loaves. Given this, the overall average over the 24-hour period is simply 600 loaves of bread.The packing process limits the cycle time for the process to 0.75 hour per 100 loaves (assume that theloaves are packaged in a batch), and this is equivalent to a throughput rate of 133.3 loaves/hour(100/0.75 = 133.3). Little’s Law calculates that the average time that loaves are in work-in-process is4.5 hours (600 loaves/133.3 loaves/hour).

The total throughput time is the time that the loaves are in work-in-process plus the operations timefor the bread-making and packaging processes. The total throughput time then is 6.25 hours (1 hour forbread making + 4.5 hours in inventory + 0.75 hour packaging). •

A R E S T A U R A N T O P E R A T I O N

EXAMPLE 5.2: A RestaurantOur bakery operates in what is referred to as steady state, meaning that the operation is started up andruns at a steady rate during the entire time that it is in operation. The output of this steady state processis adjusted by setting the amount of time that the operation is run. In the case of the bakery, we assumedthat bread making worked for three shifts and packaging for two shifts.


Service




A restaurant cannot run in this manner. The restaurant must respond to varying customer demandthroughout the day. During some peak times, it may be impossible to serve all customers immediately,and some customers may have to wait to be seated. The restaurant, because of this varying demand, is anon–steady state process. Keep in mind that many of the menu items in a restaurant can be pre-prepared.The pre-prepared items, salads and desserts for example, help speed the processes that must be per-formed when customers are at the restaurant being served.

Consider the restaurant in the casino that we discussed earlier. Because it is important that customersbe served quickly, the managers have set up a buffet arrangement where customers serve themselves.The buffet is continually replenished to keep items fresh. To further speed service, a fixed amount ischarged for the meal, no matter what the customer eats. Assume that we have designed our buffet socustomers take an average of 30 minutes to get their food and eat. Further, assume that they typicallyeat in groups (or customer parties) of two or three to a table. The restaurant has 40 tables. Each tablecan accommodate four people. What is the maximum capacity of this restaurant?

SOLUTIONIt is easy to see that the restaurant can accommodate 160 people seated at tables at a time. Actually, inthis situation it might be more convenient to measure the capacity in terms of customer parties becausethis is how the capacity will be used. If the average customer party is 2.5 individuals, then the averageseat utilization is 62.5 percent (2.5 seats/party ÷ 4 seats/table) when the restaurant is operating at capac-ity. The cycle time for the restaurant, when operating at capacity, is 0.75 minute (30 minutes/table ÷40 tables). So on average a table would become available every 45 seconds. The restaurant could han-dle 80 customer parties per hour (60 minutes ÷ 0.75 minute/party).

The problem with this restaurant is that everyone wants to eat at the same time. Management hascollected data and expects the following profile for customer parties arriving during lunch, which runsfrom 11:30 A.M. until 1:30 P.M. Customers are seated only until 1:00 P.M.

PARTIES

TIME ARRIVING

11:30–11:45 15

11:45–12:00 35

12:00–12:15 30

12:15–12:30 15

12:30–12:45 10

12:45–1:00 5

Total parties 110

Because the restaurant operates for two hours for lunch and the capacity is 80 customer parties perhour, it does not appear that the restaurant has a problem. In reality, though, there is a problem due tothe uneven flow of customers into the restaurant. A simple way to analyze the situation is to calculatehow we expect the system to look in terms of number of customers being served and number waitingin line at the end of each 15-minute interval. Think of this as taking a snapshot of the restaurant every15 minutes.

The key to understanding the analysis is to look at the cumulative numbers. The difference betweencumulative arrivals and cumulative departures gives the number of customer parties in the restaurant(those seated at tables and those waiting). Because there are only 40 tables, when the cumulative dif-ference through a time interval is greater than 40, a waiting line forms. When all 40 tables are busy, thesystem is operating at capacity; and, from the previous calculation, we know the cycle time for the entirerestaurant is 45 seconds per customer party at this time (this means that on average a table empties every45 seconds or 20 tables empty during each 15-minute interval). The last party will need to wait for allof the earlier parties to get a table, so the expected waiting time is the number of parties in line multi-plied by the cycle time.





PARTIES PARTIES PARTIES EITHER CUSTOMER

ARRIVING DEPARTING AT TABLE OR PARTIES EXPECTED

DURING DURING WAITING TO BE TABLES USED WAITING WAITING TIME

TIME PERIOD PERIOD SERVED (AT END (AT END OF (AT END OF (AT END OF

PERIOD (CUMULATIVE) (CUMULATIVE) OF PERIOD) PERIOD) PERIOD) PERIOD)

11:30–11:45 15 0 15 15

11:45–12:00 35(50) 0 50 40 10 7.5 minutes

12:00–12:15 30(80) 15 65 40 25 18.75 minutes

12:15–12:30 15(95) 20(35) 60 40 20 15 minutes

12:30–12:45 10(105) 20(55) 50 40 10 7.5 minutes

12:45–1:00 5(110) 20(75) 35 35

1:00–1:30 0(110) 35(110)

The analysis shows that by 12 noon, 10 customer parties are waiting in line. This line builds to25 parties by 12:15. The waiting line shortens to only 10 parties by 12:45.

So what can we do to solve our waiting line problem? One idea might be to shorten the cycle timefor a single table, but customers are unlikely to be rushed through their lunch in less than 30 minutes.Another idea would be to add tables. If the restaurant could add 25 tables, then a wait would not beexpected. Of course, this would eat into the space used for slot machines, so this alternative might notbe attractive to casino management. A final idea might be to double up parties at the tables, thus gettinga higher seat utilization. Doubling up might be the easiest thing to try. If 25 out of the 40 tables weredoubled up, our problem would be solved. •P L A N N I N G A T R A N S I T B U S O P E R A T I O N

EXAMPLE 5.3: Transit Bus OperationThe final example involves a logistics system. The term logistics refers to the movement of things suchas materials, people, or finished goods. Our example involves a bus route that would be typical of oneused on campus or in a metropolitan area. A similar analysis could be used for analyzing plane routes,truck routes, or ships. Similar to the restaurant, a bus transit route does not operate in steady state. Thereare definite peaks in demand during the day and evening. A good approach to take, the same as was donewith the restaurant, is to analyze distinct periods of time that represent the different types of demandpatterns placed on the service. These distinct analyses can be referred to as scenarios. Depending on thesituation, it might be reasonable to develop either a single solution that covers all the relevant scenariosor a set of solutions for the different scenarios.

A great bus route is the Balabus, or “tourist bus,” in Paris. This route loops past all the major attrac-tions in Paris. Some of the sights along the route include Notre-Dame, the Louvre, Concorde, Champs-Elysées, the Arc de Triomphe, the Eiffel Tower, and others.

70

60

50

40

30

20

10

011:45 12:00 12:15 12:30

Time

Customers in the Restaurant

Tables available in the restaurant

Parties waitingfor tables

12:45 1:00 1:15 1:30


Service

Global




Consider the problem of planning the number of buses needed to service this route. A number of fac-tors need to be considered. Let’s assume that a single bus takes exactly two hours to traverse the routeduring peak traffic. The bus company has designed delays in the route so that even though traffic is busythe bus can keep on schedule. The route has 60 stops, although the bus stops only when passengers onthe bus request a stop or when the driver sees customers waiting to board at a stop. Each bus has seat-ing capacity of about 50 passengers, and another 30 passengers can stand. This route is busy much ofthe day because visitors to the city tend to start visiting the sites early and continue until dark. Finally,the transit authority wants to give good service and have enough capacity to handle peak customerloads. The following is an analysis of the situation.

SOLUTIONA key measure of service is how long a customer must wait prior to the arrival of a bus. Considerinitially the case of only a single bus serving the route. If a person at a random time comes to a bus stop,we know that the maximum time that the customer needs to wait is two hours. Here we assume that thebus is able to cover the route in exactly two hours. If there is significant variability in this cycle time,the waiting time goes up. We discuss the impact of variability in Technical Note 6. This would be thecase when the unlucky customer just missed the bus. If the bus was halfway through the route (relativeto where the customer is waiting), then the customer needs to wait one hour. Continuing with this logic,we can estimate the average wait time for the customer to be one hour. In general, we can say that theaverage wait time would be half the cycle time of the process. If two buses are used, the cycle time isone hour and the average wait is 30 minutes. If we want the average wait to be two minutes, then therequired cycle time is four minutes, and 30 buses are needed (120 minutes ÷ 4 minutes/bus =30 buses).

The next issue relates to the capacity of the system. If we have 30 buses on the route and each busseats 50 passengers with another 30 standing, we know that we can accommodate 1,500 seated or 2,400passengers in total at one point in time.

Assume that the following table is an estimate of the number of passengers that travel the route dur-ing a typical tourist season day. The table shows calculations of the amount of bus capacity requiredduring each hour. If a customer rides the bus for 45 minutes, then one seat is needed for 45 minutes, or0.75 hour, to handle that passenger. Of course, 60 minutes, or a full hour’s worth, of capacity is availablefor each seat that we have. At maximum utilization including standing, each bus can handle 80 passenger-hours’ worth of load. Dividing the expected passenger load during the hour by the maximum load for asingle bus calculates the minimum number of buses needed. Similarly, dividing the expected passengerload by the number of seats on each bus calculates the number of buses needed so that all passengerscan be seated.

LOAD MINIMUM BUSES NEEDED FOR

NUMBER OF AVERAGE TIME (PASSENGER NUMBER OF ALL PASSENGERS

TIME CUSTOMERS ON BUS HOURS) BUSES NEEDED TO BE SEATED

8:00–9:00 A.M. 2,000 45 minutes 1,500 18.75 30

9:00–10:00 A.M. 4,000 30 minutes 2,000 25 40

10:00–11:00 A.M. 6,000 30 minutes 3,000 37.5 60

11:00 A.M.–12:00 NOON 5,000 30 minutes 2,500 31.25 50

12:00–1:00 P.M. 4,000 30 minutes 2,000 25 40

1:00–2:00 P.M. 3,500 30 minutes 1,750 21.875 35

2:00–3:00 P.M. 3,000 45 minutes 2,250 28.125 45

3:00–4:00 P.M. 3,000 45 minutes 2,250 28.125 45

4:00–5:00 P.M. 3,000 45 minutes 2,250 28.125 45

5:00–6:00 P.M. 4,000 45 minutes 3,000 37.5 60

6:00–7:00 P.M. 3,000 45 minutes 2,250 28.125 45

7:00–8:00 P.M. 1,500 45 minutes 1,125 14.0625 22.5

TOTALS 42,000 25,875





From the analysis, if the Paris transit authority uses only 30 buses throughout the day, many peoplewill need to stand. Further, during the morning rush between 10 and 11 A.M. and the evening rushbetween 5 and 6 P.M., not all of the customers can be accommodated. It would seem reasonable that atleast 40 buses should be used between 9 A.M. and 7 P.M. Even with this number of buses, one wouldexpect passengers to be standing most of the time.

If the transit authority decided to use 40 buses between the extended hours of 8 A.M. through 8 P.M.,what would be the average utilization of the buses in terms of seats occupied? Over this 12-hourperiod, 24,000 seat-hours of capacity would be available (40 buses × 12 hours × 50 seats/bus). Thetable indicates that 25,875 seat-hours are needed. The utilization would be 107.8 percent(25,875/24,000 × 100). What this means is that on average 7.8 percent of the customers must stand.Of course, this average value significantly understates the severe capacity problem that occurs duringthe peak times of the day. •

Consider in the preceding example how useful this type of analysis is to the Paris transitauthority. Data can be collected for each day of the week and the analysis performed.Interesting questions concerning the design of the route or the capacity of the buses can be eval-uated. Consider, for example, what would happen if the route were split into two parts. What iflarger buses that could carry 120 passengers were put into service? The analysis can be extend-ed to include the cost of providing the service by considering the wages paid the operators, thecost to maintain and operate the vehicles, and depreciation of the buses. As seen from the aboveexample, designing a transit system involves a trade-off between the convenience of the ser-vice, or how frequently buses arrive at each stop, and the capacity utilization of the buses.


Critical processes are subject to the well-known rule that time is money. Forexample, the longer a customer waits, the more likely it is that the customer will switch to adifferent vendor. The longer material is kept in inventory, the higher the investment cost.Unfortunately, critical processes often depend on specific limited resources, resulting in bot-tlenecks. Throughput time can sometimes be reduced without purchasing additional equip-ment. The following are some suggestions for reducing the throughput time of a process thatdo not require the purchase of new equipment. Often a combination of ideas is appropriate.4

1. Perform activities in parallel. Most of the steps in an operations process are per-formed in sequence. A serial approach results in the throughput time for the entireprocess being the sum of the individual steps plus transport and waiting time betweensteps. Using a parallel approach can reduce throughput time by as much as 80 percentand produces a better result.

A classic example is product development, where the current trend is toward con-current engineering. Instead of forming a concept, making drawings, creating a billof materials, and mapping processes, all activities are performed in parallel by inte-grated teams. Development time is reduced dramatically, and the needs of all thoseinvolved are addressed during the development process.

2. Change the sequence of activities. Documents and products are often transportedback and forth between machines, departments, buildings, and so forth. For instance, adocument might be transferred between two offices a number of times for inspection andsigning. If the sequence of some of these activities can be altered, it may be possible toperform much of the document’s processing when it comes to a building the first time.

3. Reduce interruptions. Many processes are performed with relatively large timeintervals between each activity. For example, purchase orders may be issued onlyevery other day. Individuals preparing reports that result in purchase orders should beaware of deadlines to avoid missing them, because improved timing in these processescan save many days of throughput time.

P R O C E S S T H R O U G H P U T T I M E R E D U C T I O N




To illustrate these ideas, consider an electronics manufacturer that has been receivingcustomer complaints about a long order lead time of 29 days. As assessment of the order-processing system revealed 12 instances where managers had to approve employees’ work. Itwas determined that the first 10 approvals were not needed. This saved an average of sevento eight days in the order processing.

Many subsystems—each performing the same or similar tasks—had interfered with theprocess. The logical step was to eliminate redundancy, and a detailed flowchart of the processwas created. At close inspection, 16 steps proved very similar to one another. Changing thesequence of activities and creating one companywide order document removed 13 of these steps.

Over four months, the order system was totally redesigned to allow information to beentered once and become available to the entire organization. Due to this adjustment, activi-ties could be handled in a parallel manner. After a value-added analysis (focused on elimi-nating the non–value-adding activities), the manufacturer was able to reduce the customerorder lead time from 29 days to 9 days, save cost and employee time per order, and increasecustomer satisfaction.


THE SPEED WITH WHICH A COMPANY CAN

DESIGN AND DEVELOP NEW PRODUCTS IS

A CRITICAL ELEMENT IN ITS ABILITY TO

INTRODUCE NEW PRODUCTS INTO THE

MARKETPLACE. RAPID PROTOTYPING

MACHINES CAN QUICKLY PRODUCE THREE-DIMENSIONAL PROTOTYPES ALLOWING

DESIGN, ENGINEERING, AND PRODUCTION

PEOPLE TO GIVE THEIR INPUT AND TEST THE

DESIGN EARLY IN THE DEVELOPMENT CYCLE.THESE MODELS RESULT IN HIGHER-QUALITY

PRODUCTS AND LOWER DEVELOPMENT

COSTS.

Process analysis is a basic skill needed to understand how a business operates.Great insight is obtained by drawing a simple flowchart showing the flow of materials orinformation through an enterprise. The diagram should include all the operating elements andshow how they fit together. Be sure to indicate where material is stored or where orders arequeued. Often 90 percent or more of the time that is required to serve a customer is spent justwaiting. Hence, merely eliminating the waiting time can dramatically improve the perfor-mance of the process.

Remember this fundamental concept when analyzing a process: What goes into theprocess must come out of the process. A process taken as a whole is like the funnel shown inExhibit 5.5. The outlet of the funnel restricts the amount that can flow through. In a real busi-ness process, certain resources limit output. If liquid is poured into the funnel at a rate greaterthan it can exit, the level in the funnel will continue to grow. As the level of liquid in the

C O N C L U S I O N




funnel grows, the time that it takes the liquid to flow through the funnel increases. If too muchliquid is poured into the funnel, it just spills over the top and never flows through.

The same is true of a real process. If too many jobs are pumped into the process, the timethat it takes to complete a job will increase because the waiting time will increase. At somepoint, customers will go somewhere else and the business will be lost. When a process isoperating at capacity, the only way to take on more work without increasing the waiting timeis to add more capacity. This requires finding what activity is limiting the output of theprocess and increasing the capacity of that activity. In essence, the tube leading out of the fun-nel needs to be made larger.


e x h i b i t 5 . 5 What Goes into a Process Must Come Out of the Process. Input Rate Must Be Less than orEqual to the Output Rate; otherwise, the System Will Overflow.

New work enters the process

Work waiting to be completed

Throughputtime

Process required to complete work

Completed work

K E Y T E R M SProcess Any set of activities performed by an organization thattakes inputs and transforms them into outputs ideally of greatervalue to the organization than the original inputs.

Cycle time The average time between completions of successiveunits in a process (this is the definition used in this book). The termis sometimes used to mean the elapsed time between starting andcompleting a job.

Utilization The ratio of the time that a resource is actually activatedrelative to the time that it is available for use.

Buffering A storage area between stages where the output of a stageis placed prior to being used in a downstream stage. Bufferingallows the stages to operate independently.

Blocking The activities in the stage must stop because there is noplace to deposit the item just completed.

Starving The activities in a stage must stop because there is nowork.

Bottleneck A resource that limits the capacity or maximum outputof the process.

Make-to-order A process that is activated only in response to anactual order.

Make-to-stock A process that produces standard products that arestored in finished goods inventory. The product is delivered quick-ly to the customer from the finished goods inventory.

Hybrid Combines the features of both make-to-order and make-to-stock. Typically, a generic product is made and stocked at somepoint in the process. These generic units are customized in a finalprocess to meet actual orders.

Pacing Movement of items through a process is coordinated througha timing mechanism. Most processes are not paced, but assemblylines usually are paced.

Productivity The ratio of output to input. Taking the dollar value ofthe output and dividing by the dollar value of the inputs usually




S O L V E D P R O B L E MDaffy Dave’s Sub Shop makes custom submarine sandwiches to order. They are analyzing theprocesses at their shop. The general flow of the process is shown below. There is a separate personworking at each of the steps in the process.

Daffy Dave wants to figure out the following for a typical 8-hour work day.a. What is the current maximum output of the process?b. If we add another person where would we add him or her and what is the benefit?c. Is there a benefit if we can shift 1 minute from Bun and Meat to Order Taking? Assume we do

not make the change in part b above.d. Is there a benefit if we shift 1 minute of work from Condiments to Bagging? Assume we do not

make the changes in parts b and c above.

Solutiona. Maximum output is 120 subs per day.

OPERATION OUTPUT

Take Orders (60 min. per hour/1 min. per order) * 8 hours = 480 subs per day

Bun and Meat (60 min. per hour/3 min. per order) * 8 hours = 160 subs per day

Toppings/Condiments (60 min. per hour/4 min. per order) * 8 hours = 120 subs per day

Bag the Order (60 min. per hour/2 min. per order) * 8 hours = 240 subs per day

Output per day is determined by the slowest station; therefore we can only produce 120 per daybecause that is the limit of the Toppings/Condiments station.

b. Dave should add the person to the slowest station (Condiments/Toppings) since it is thebottleneck.

OPERATION OUTPUT

Take Orders 480 subs per day

Bun and Meat 160 subs per day

Toppings/Condiments 120 * 2 = 240 subs per day

Bag the Order 240 subs per day

Take theOrder

1 minute/order 3 minutes/order 4 minutes/order 2 minutes/order

Slice the Bunand Add theMeat/Cheese

Add theToppings andCondiments

Bag the Order


measures total factor productivity. Alternatively, partial factor pro-ductivity is measured based on an individual input and often is notcalculated using dollar values (an example would be units/person).

Efficiency A ratio of the actual output of a process relative to somestandard.

Run time The time required to produce a batch of parts.

Setup time The time required to prepare a machine to make a par-ticular item.

Operation time The sum of the setup time and run time for a batchof parts that are run on a machine.

Throughput time The average time that it takes a unit to movethrough an entire process. Usually the term lead time is used to refer

to the total time that it takes a customer to receive an order (includestime to process the order, throughput time, and delivery time).

Throughput rate The output rate that the process is expected to pro-duce over a period of time.

Process velocity or throughput ratio The ratio of the total through-put time to the value-added time.

Value-added time The time in which useful work is actually beingdone on the unit.

Little’s Law States a mathematical relationship between throughputrate, throughput time, and the amount of work-in-process inventory.Throughput time is equal to work-in-process divided by thethroughput rate.




The impact is not a very big one. Even though the Toppings/Condiments station now can do240 subs per day, the Bun and Meat station can only do 160 so that is the maximum output.

c. Order Taking station will go from 1 minute to 2 minutes and Bun and Meat goes from 3 minutesto 2 minutes.

OPERATION OUTPUT





There is no benefit to this change. Dave can still only make 120 subs per day since we can onlyproduce 120 per day because that is the limit of the Toppings/Condiments station.

d. Toppings/Condiments station will go from 4 minutes to 3 minutes and Bagging goes from2 minutes to 3 minutes.

OPERATION OUTPUT





There is a benefit to this change. Dave can now make 160 subs per day. This will provide thesame benefit as hiring another worker. However, if Dave wants to increase output further he willhave to hire some additional staff.

R E V I E W A N D D I S C U S S I O N Q U E S T I O N S1 Compare McDonald’s old and new processes for making hamburgers. How valid is McDonald’s

claim that the new process will produce fresher hamburgers for the customer? ComparingMcDonald’s new process to the processes used by Burger King and Wendy’s, which processwould appear to produce the freshest hamburgers?

2 State in your own words what Little’s Law means. Describe an example that you have observedwhere Little’s Law applies.

3 Explain how having more work-in-process inventory can improve the efficiency of a process.How can this be bad?

4 Recently some operations management experts have begun insisting that simply minimizingprocess velocity, which actually means minimizing the time that it takes to process somethingthrough the system, is the single most important measure for improving a process. Can you thinkof a situation in which this might not be true?

P R O B L E M S 5

1 An enterprising student has set up an internship clearinghouse for business students. Each stu-dent that uses the service fills out a form and lists up to 10 companies that he or she would liketo have contacted. The clearinghouse has a choice of two methods to use for processing the forms.The traditional method requires about 20 minutes to review the form and arrange the informa-tion in the proper order for processing. Once this setup is done, it takes only two minutes percompany requested to complete the processing. The other alternative uses an optical scan/retrievesystem, which takes only a minute to prepare but requires five minutes per company for complet-ing the processing. If it costs about the same amount per minute for processing with either of thetwo methods, when should each be used?





2 Rockness Recycling refurbishes rundown business students. The process uses a moving belt,which carries each student through the five steps of the process in sequence. The five steps areas follows:

TIME REQUIRED

STEP DESCRIPTION PER STUDENT

1 Unpack and place on belt 1.0 minute

2 Strip off bad habits 1.5 minutes

3 Scrub and clean mind 0.8 minute

4 Insert modern methods 1.0 minute

5 Polish and pack 1.2 minutes

One faculty member is assigned to each of these steps. Faculty members work a 40-hour weekand rotate jobs each week. Mr. Rockness has been working on a contract from General Eclectic,which requires delivery of 2,000 refurbished students per week. A representative of the humanresources department has just called complaining that the company hasn’t been receiving theagreed-upon number of students. A check of finished goods inventory by Mr. Rockness revealsthat there is no stock left. What is going on?

3 The bathtub theory of operations management is being promoted as the next breakthrough forglobal competitiveness. The factory is a bathtub with 50 gallons of capacity. The drain is the out-let to the market and can output three gallons per hour when wide open. The faucet is the rawmaterial input and can let material in at a rate of four gallons per hour. Now, to test your com-prehension of the intricacies of operations (assume the bathtub is empty to begin with):a. Draw a diagram of the factory and determine the maximum rate at which the market can be

served if all valves are set to maximum. What happens to the system over time?b. Suppose that instead of a faucet, a five-gallon container is used for filling the bathtub (assume

a full container is next to the tub to begin with); it takes two hours to refill the container andreturn it to the bathtub. What happens to the system over time?

4 A local market research firm has just won a contract for several thousand small projects involv-ing data gathering and statistical analysis. In the past the firm has assigned each project to a sin-gle member of its highly trained professional staff. This person would both gather and analyzethe data. Using this approach an experienced person can complete an average of 10 such projectsin an eight-hour day.

The firm’s management is thinking of assigning two people to each project in order to allowthem to specialize and become more efficient. The process would require the data gatherer to fillout a matrix on the computer, check it, and transmit it to the statistical analysis program for theanalyst to complete. Data can be gathered on one project while the analysis is being completedon another, but the analysis must be complete before the statistical analysis program can acceptthe new data. After some practice, the new process can be completed with a standard time of20 minutes for the data gathering and 30 minutes for the analysis.a. What is the production (output per hour) for each alternative? What is the productivity (out-

put per labor hour)?b. How long would it take to complete 1,000 projects with each alternative? What would be the

labor content (total number of labor hours) for 1,000 projects for each alternative?5 A processor makes two components, A and B, which are then packaged together as the final

product (each product sold contains one A and one B). The processor can do only one compo-nent at a time: either it can make As or it can make Bs. There is a setup time when switchingfrom A to B.

Current plans are to make 100 units of component A, then 100 units of component B, then100 units of component A, then 100 units of component B, and so forth, where the setup and runtimes for each component are given below.

COMPONENT SETUP/CHANGEOVER TIME RUN TIME/UNIT

A 5 minutes 0.2 minute

B 10 minutes 0.1 minute





Assume the packaging of the two components is totally automated and takes only two secondsper unit of the final product. This packaging time is small enough that you can ignore it. What isthe average hourly output, in terms of the number of units of packaged product (which includesone component A and one component B)?

6 The following represents a process used to assemble a chair with an upholstered seat. Stations A,B, and C make the seat; stations J, K, and L assemble the chair frame; station X is where the twosubassemblies are brought together; and some final tasks are completed in stations Y and Z. Oneworker is assigned to each of the stations. Generally there is no inventory kept anywhere in thesystem, although there is room for one unit between each of the stations that might be used fora brief amount of time.

Given the following amount of work in seconds required at each station:

A 38 J 32 X 22

B 34 K 30 Y 18

C 35 L 34 Z 20

a. What is the possible daily output of this “process” if 8 hours of processing time is availableeach day?

b. Given your output rate in part a, what is the efficiency of the process?c. What is the throughput time of the process?

7 Wally’s Widget Warehouse takes orders from 7 A.M. to 7 P.M. The manager wants to analyze theprocess and has provided the process flow diagram shown below. There are three steps requiredto ship a customer order. The first step is to take the order from a customer. The second step isto pick the order for the customer and then they have to pack the order ready for shipping. Wallypromises that every order placed today gets shipped tomorrow. That means that the picking andpacking operation must finish all orders before they go home.

Wally wants to figure out the following.a. What is the current maximum output of the process?b. How long will the picking and packing operations have to work if we have a day with the

maximum orders?c. What is the maximum number of orders waiting to be picked?d. What is the maximum number of orders waiting to be packed?e. If we double the packing capacity (from 60 to 120 orders per hour), what impact does this

have on your answers in parts b, c, and d ?8 The National State Bank is trying to make sure that they have enough tellers to handle the Friday

afternoon rush of workers wanting to cash their paychecks. They are only concerned with the lasthour of the day from 4:00 to 5:00 P.M. It takes 5 minutes per customer to be processed by thetellers. The average customer arrivals are shown in the table below.

CustomerOrders

PickedOrders

Wait forShipperOrder

Taker

100 Customers/hour

80 Customers/hour

60 Customers/hour

PickOrders

PackOrders

Customers

B

C

A

K

L

J

Y ZX





TIME CUSTOMERS ARRIVING

4:00–4:05 2

4:05–4:10 5

4:10–4:15 6

4:15–4:20 8

4:20–4:25 10

4:25–4:30 12

4:30–4:35 16

4:35–4:40 12

4:40–4:45 10

4:45–4:50 6

4:50–4:55 4

4:55–5:00 2

5:00–5:05 0

Total 93

The bank currently has 8 teller stations and all are staffed during the Friday afternoon rush hour.a. What is the current maximum output at the bank during rush hour?b. Can the bank process all the customers by 5:00 P.M.?c. What is the maximum waiting time for customers and what time period does it occur in?

A D V A N C E D P R O B L E M9 Remember Mr. Rockness in Problem 2? He now retrains college professors. It is a much more

challenging task but still involves five steps. He has worked hard to balance the line; however,there is a lot of variability. Each stage in the process now handles between one and six facultymembers per hour depending on how bad the case is. If there is some inventory available forevery position (do not worry about the start-up), what is the expected output per hour? (Assumethat each stage is independent and that it is equally likely that one, two, three, four, five, or sixfaculty members get processed each hour at each stage.)6


C A S E : A N A L Y Z I N G C A S I N O M O N E Y - H A N D L I N G P R O C E S S E S

Retrieving money from a slot machine is referred toas the drop process. The drop process begins with a security offi-cer and the slot drop team leader obtaining the slot cabinet keysfrom the casino cashier’s cage. Getting the keys takes about15 minutes. The slot drop team consists of employees from thehard count coin room, security, and accounting. The slot dropleader, under the observation of a security officer and a personfrom accounting, actually removes the drop bucket from the slotmachine cabinet. When the drop bucket is pulled from the slot cab-inet, a tag with the proper slot machine number is placed on top ofthe coins to identify where that bucket came from when the weighprocess begins. Retrieving the drop bucket takes about 10 minutesper slot machine. Once a cart is filled with buckets from 20 differ-ent slot machines, the drop team leader and security and account-ing people deliver the buckets to the hard count room. The bucketsare securely locked in the hard count room to await the start of thehard count process. Delivering and securing the buckets takesabout 30 minutes per cart.

The hard count process is performed at a designated time knownto gaming regulatory authorities. The hard count team first tests theweigh scale, which takes 10 minutes. The scale determines the dollarvalue, by denomination, for set weights of 10 and 25 pounds. These

results are compared to calibration results, calculated when the scalewas last serviced, to determine if a significant variance exists. If onedoes exist, the hard count supervisor must contact the contractorresponsible for maintaining the scale and the controller’s office. If nosignificant variance is found, the weigh process can continue.

Following the scale check, each drop bucket is emptied into theweigh scale holding hopper. Using information from the identifica-tion tag, the specific slot machine number from which the bucketoriginated is entered into the weigh scale computer. The weigh scalecomputer is programmed to convert the weight of coins, by denom-ination, into specific dollar values, which are recorded in the weighjournal along with the slot machine number. This weighing andrecording process takes seven minutes per bucket. Once the scalehas weighed the contents of the drop bucket, the coins automaticallydrop onto a conveyor belt, which transports them to wrappingmachines. As the coins are wrapped, the rolls of coins drop ontoanother conveyor belt, which takes them to a canning station.Twenty-five silver dollars are wrapped in each roll at a rate of10 rolls per minute.

At the canning station, the coin rolls are placed in metal or plas-tic cans that hold specific dollar amounts based on coin denomina-tion. The cans are stacked to facilitate counting the wrapped coins.

Excel:BottleneckSimulation





C A S E : K R I S T E N ’ S C O O K I E C O M P A N Y ( A )

You and your roommate are preparing to startKristen’s Cookie Company in your on-campus apartment. The com-pany will provide fresh cookies to starving students late at night.You need to evaluate the preliminary design for the company’s pro-duction process to figure out many variables, including what pricesto charge, whether you will be able to make a profit, and how manyorders to accept.

BUSINESS CONCEPTYour idea is to bake fresh cookies to order, using any combinationof ingredients that the buyer wants. The cookies will be ready forpickup at your apartment within an hour.

Several factors will set you apart from competing products suchas store-bought cookies. First, your cookies will be completelyfresh. You will not bake any cookies before receiving the order;therefore, the buyer will be getting cookies that are literally hot outof the oven.

Second, like Steve’s Ice Cream,7 you will have a variety ofingredients available to add to the basic dough, including chocolatechips, M&M’s, chopped Heath bars, coconut, walnuts, and raisins.Buyers will telephone in their orders and specify which of theseingredients they want in their cookies. You guarantee completelyfresh cookies. In short, you will have the freshest, most exotic cook-ies anywhere, available right on campus.

THE PRODUCTION PROCESSBaking cookies is simple: mix all the ingredients in a food proces-sor; spoon out the cookie dough onto a tray; put the cookies into theoven; bake them; take the tray of cookies out of the oven; let thecookies cool; and, finally, take the cookies off the tray and carefullypack them in a box. You and your roommate already own all thenecessary capital equipment: one food processor, cookie trays, andspoons. Your apartment has a small oven that will hold one tray ata time. Your landlord pays for all the electricity. The variable costs,therefore, are merely the cost of the ingredients (estimated to be$.60/dozen), the cost of the box in which the cookies are packed

($.10 per box; each box holds a dozen cookies), and your time (whatvalue do you place on your time?).

A detailed examination of the production process, which speci-fies how long each of the steps will take, follows. The first step is totake an order, which your roommate has figured out how to doquickly and with 100 percent accuracy. (Actually, you and yourroommate devised a method using the campus electronic mail sys-tem to accept orders and to inform customers when their orders willbe ready for pickup. Because this runs automatically on your per-sonal computer, it does not take any of your time.) Therefore, thisstep will be ignored in further analysis.

You and your roommate have timed the necessary physical oper-ations. The first physical production step is to wash out the mixingbowl from the previous batch, add all of the ingredients, and mix themin your food processor. The mixing bowls hold ingredients for up to3 dozen cookies. You then dish up the cookies, one dozen at a time,onto a cookie tray. These activities take six minutes for the washingand mixing steps, regardless of how many cookies are being made inthe batch. That is, to mix enough dough and ingredients for two dozencookies takes the same six minutes as one dozen cookies. However,dishing up the cookies onto the tray takes two minutes per tray.

The next step, performed by your roommate, is to put the cook-ies in the oven and set the thermostat and timer, which takes aboutone minute. The cookies bake for the next nine minutes. So totalbaking time is 10 minutes, during the first minute of which yourroommate is busy setting the oven. Because the oven holds only onetray, a second dozen takes an additional 10 minutes to bake.

Your roommate also performs the last steps of the process byfirst removing the cookies from the oven and putting them aside tocool for 5 minutes, then carefully packing them in a box and accept-ing payment. Removing the cookies from the oven takes only a neg-ligible amount of time, but it must be done promptly. It takes twominutes to pack each dozen and about one minute to accept pay-ment for the order.

That is the process for producing cookies by the dozen inKristen’s Cookie Company. As experienced bakers know, a few

Silver dollar cans hold $1,000, or 40 rolls, and take five minutes tofill and stack. When the weigh process is completed, the weigh scalecomputer runs a summary report totaling the weight by denomina-tion. These totals are recorded on the weigh/wrap verificationreport, which takes five minutes to produce.

When the wrap portion of the count is completed and all of therolled coins have been canned and stacked, they are manuallycounted by denomination. These totals are also recorded on theweigh/wrap verification report. The variance in both dollar amountsand percentages, for each denomination, is calculated. Variancesthat exceed plus or minus 2 percent or are $1,000 or greater(whichever is less) must be investigated by the hard count supervi-sor, who writes an explanatory report. If no significant variancesexist, all members of the hard count team sign the weigh/wrap ver-ification report. To complete the hard count process, the casinocashier’s cage is then notified that the slot drop is ready to be trans-ferred into cage accountability. Manually counting and verifying thecounts take on average two minutes per can.

In a process separate from the hard count, a cage cashier per-forms an independent count and verification, by denomination, ofthe wrap. If everything balances, the main bank cashier signs theweigh/wrap verification report, accepting the slot drop into cageaccountability. It is at this point that the actual slot gross gamingrevenue is recognized.

Q U E S T I O N S1 Draw a diagram of the drop process. How long should it take

to empty 300 silver dollar slot machines?2 Draw a diagram of the hard count process. How long should

this process take to complete for 300 silver dollar slotmachines? Assume that each slot machine has an average of750 silver dollars when it is emptied.

3 The casino is considering the purchase of a second coin-wrapping machine. What impact would this have on the hardcount process? Is this the most desirable machine to purchase?




F O O T N O T E S1 Often the term cycle time is used to mean throughput time. It is important to carefully determine how the term is being used in the

context of the process being studied.

2 J. D. C. Little, “A Proof for the Queuing Formula: L = λW,” Operations Research 9, no. 3 (1961), pp. 383–87. Special thanks toJ. F. Muth, Indiana University.

3 This example is similar to one given by A. E. Gray in “Capacity Analysis: Sample Problems,” Harvard Business School 9-696-058.

4 B. Andersen, “Process Cycle Time Reduction,” Quality Progress, July 1999, p. 120. For some additional guidelines for improvingprocess, also see Chapter 17.

5 The authors are indebted to D. Clay Whybark of the University of North Carolina for contributing Problems 1–4 and Problem 9.

6 The idea for this problem came from an exercise developed by Dr. Eli Goldratt titled “The Great Manufacturing Crapshoot.”

7 Steve’s Ice Cream was started in the Boston area by a young entrepreneur to provide make-to-order ice cream, using mix-ins.


simplifications were made in the actual cookie production process.For example, the first batch of cookies for the night requires pre-heating the oven. However, such complexities will be put aside fornow. Begin your analysis by developing a process flow diagram ofthe cookie-making process.

KEY QUESTIONS TO ANSWER BEFORE YOU LAUNCH THE BUSINESSTo launch the business, you need to set prices and rules for accept-ing orders. Some issues will be resolved only after you get startedand try out different ways of producing the cookies. Before youstart, however, you at least want a preliminary plan, with as muchas possible specified, so that you can do a careful calculation of howmuch time you will have to devote to this business each night, andhow much money you can expect to make. For example, when youconduct a market survey to determine the likely demand, you willwant to specify exactly what your order policies will be. Therefore,answering the following operational questions should help you:

1 How long will it take you to fill a rush order?2 How many orders can you fill in a night, assuming you are

open four hours each night?3 How much of your own and your roommate’s valuable time

will it take to fill each order?4 Because your baking trays can hold exactly one dozen cook-

ies, you will produce and sell cookies by the dozen. Shouldyou give any discount for people who order two dozen cook-ies, three dozen cookies, or more? If so, how much? Will ittake you any longer to fill a two-dozen cookie order than aone-dozen cookie order?

5 How many food processors and baking trays will you need?

6 Are there any changes you can make in your production plansthat will allow you to make better cookies or more cookies inless time or at lower cost? For example, is there a bottleneckoperation in your production process that you can expandcheaply? What is the effect of adding another oven? Howmuch would you be willing to pay to rent an additional oven?

P R O B L E M S F O R F U R T H E RT H O U G H T

1 What happens if you are trying to do this by yourself with-out a roommate?

2 Should you offer special rates for rush orders? Suppose youhave just put a tray of cookies into the oven and someonecalls up with a “crash priority” order for a dozen cookies ofa different flavor. Can you fill the priority order while stillfulfilling the order for the cookies that are already in theoven? If not, how much of a premium should you charge forfilling the rush order?

3 When should you promise delivery? How can you lookquickly at your order board (list of pending orders) and tella caller when his or her order will be ready? How much of asafety margin for timing should you allow?

4 What other factors should you consider at this stage of plan-ning your business?

5 Your product must be made to order because each order ispotentially unique. If you decide to sell standard cookiesinstead, how should you change the production system? Theorder-taking process?

S E L E C T E D B I B L I O G R A P H YAnupindai, R.; S. Chopra; S. D. Deshmukh; J. A. van Mieghem; and

E. Zemel. Managing Business Process Flows. Upper Saddle River,NJ: Prentice Hall, 1999.

Gray, A. E., and J. Leonard. “Process Fundamentals.” Harvard BusinessSchool 9-696-023.

Hopp, J. W., and M. L. Spearman. Factory Physics: Foundations of -Manufacturing Management. New York: Irwin/McGraw-Hill,2000.

Smith, Howard, and Peter Fingar. Business Process Management (BPM):The Third Wave. Tampa, FL: Meghan-Kiffer Press, 2003.

Suri, R. Quick Response Manufacturing: A Companywide Approach toReducing Lead Times. Portland, OR: Productivity Press, 1998.

KRISTEN’S COOKIE COMPANY (A), CASE 9-686-093, WRITTEN BY ROGER BOHN. COPYRIGHT © 1986 BY THE HARVARD BUSINESS SCHOOL PUBLISHING CORPORATION. ALL RIGHTS RESERVED.




© The McGraw−Hill Companies, 2005

TECHNIC

ALNOTE5

tech

nica

l not

e

t e c h n i c a l n o t e f i v eJ O B D E S I G N A N D W O R KM E A S U R E M E N T

1 8 1 Job Design DecisionsJob design defined

1 8 2 Behavioral Considerations in Job DesignDegree of labor specialization Specialization of labor defined

Job enrichment Job enrichment defined

Sociotechnical systems Sociotechnical systems defined

1 8 4 Physical Considerations in Job DesignWork physiology defined

Ergonomics defined

1 8 5 Work MethodsA production processWorker at a fixed workplaceWorker interacting with equipmentWorkers interacting with other workers

1 9 0 Work Measurement and StandardsWork measurement techniques Work measurement defined

Work sampling compared to time study Time study defined

Work sampling defined

Predetermined motion-time data systems defined

Elemental data defined

Normal time defined

Standard time defined

1 9 9 Financial Incentive PlansBasic compensation systemsIndividual and small-group incentive plansOrganizationwide plans

2 0 1 Conclusion

2 0 6 Case: Jeans Therapy—Levi’s Factory Workers Are Assigned to Teams, and Morale Takes a Hit





The operations manager’s job, by definition, dealswith managing the personnel that create a firm’sproducts and services. To say that this is a chal-lenging job in today’s complex environment is anunderstatement. The diversity of the workforce’scultural and educational background, coupledwith frequent organization restructuring, calls fora much higher level of people management skillsthan has been required in even the recent past.

The objective in managing personnel is toobtain the highest productivity possible withoutsacrificing quality, service, or responsiveness.The operations manager uses job design tech-niques to structure the work so that it will meetboth the physical and psychological needs of thehuman worker. Work measurement methods areused to determine the most efficient means ofperforming a given task, as well as to set reason-able standards for performing it. People are moti-vated by many things, only one of which is finan-cial reward. Operations managers can structure such rewards not only to motivate consis-tently high performance but also to reinforce the most important aspects of the job.

JOB DESIGN AND WORK MEASUREMENT technical note 181

Job design may be defined as the function of specifying the work activities ofan individual or group in an organizational setting. Its objective is to develop job structuresthat meet the requirements of the organization and its technology and that satisfy the job-holder’s personal and individual requirements. Exhibit TN5.1 summarizes the decisionsinvolved. These decisions are affected by the following trends:

1. Quality control as part of the worker’s job. Now often referred to as “quality atthe source” (see Chapter 8), quality control is linked with the concept of empowerment.

J O B D E S I G N D E C I S I O N S

e x h i b i t T N 5 . 1

Who What Where When Why How

Mental andphysicalcharacteristicsof theworkforce

Task(s) tobe performed

Geographiclocale oforganization;location ofwork areas

Time of day;time of occur-rence in thework flow

Organizationalrationale forthe job; ob-jectives andmotivation ofthe worker

Method ofperformance and motivation

Ultimatejob

structure

Job Design Decisions

Job design





Empowerment, in turn, refers to workers being given authority to stop a production lineif there is a quality problem, or to give a customer an on-the-spot refund if service wasnot satisfactory.

2. Cross-training of workers to perform multiskilled jobs. As companies downsize,the remaining workforce is expected to do more and different tasks.

3. Employee involvement and team approaches to designing and organizing work.This is a central feature in total quality management (TQM) and continuous improve-ment efforts. In fact, it is safe to say that virtually all TQM programs are team based.

4. “Informating” of ordinary workers through e-mail and the Internet, therebyexpanding the nature of their work and their ability to do it. In this context,informating is more than just automating work—it is revising work’s fundamentalstructure. Northeast Utilities’ computer system, for example, can pinpoint a problemin a service area before the customer service representative answers the phone. Therep uses the computer to troubleshoot serious problems, to weigh probabilities thatother customers in the area have been affected, and to dispatch repair crews beforeother calls are even received.

5. Extensive use of temporary workers. Manpower, a company specializing in pro-viding temporary employees, has over 1.9 million temporary employees worldwide onits payroll.

6. Creation of “alternative workplaces” such as shared offices, telecommuting, andvirtual offices to supplement or replace traditional office settings. These are usedto increase productivity, reduce travel and real estate costs, and aid in recruiting andretaining employees. IBM, AT&T, and American Express are major proponents of theapproach.1

7. Automation of heavy manual work. Examples abound in both services (one-person trash pickup trucks) and manufacturing (robot spray painting on auto lines).These changes are driven by safety regulations as well as economics and personnelreasons.

8. Most important of all, organizational commitment to providing meaningful andrewarding jobs for all employees. Companies featured on Fortune magazine’s“100 Best Companies to Work For” use creative means to keep employees satis-fied, and offer generous severance and compassion when cuts must be made (seewww.fortune.com for the current list of companies).


D E G R E E O F L A B O R S P E C I A L I Z A T I O NSpecialization of labor is the two-edged sword of job design. On one hand, specializationhas made possible high-speed, low-cost production, and from a materialistic standpoint, ithas greatly enhanced our standard of living. On the other hand, extreme specialization (aswe see in mass-production industries) often has serious adverse effects on workers, whichin turn are passed on to management. In essence, the problem is to determine how muchspecialization is enough. At what point do the disadvantages outweigh the advantages?(See Exhibit TN5.2.)

Recent research suggests that the disadvantages dominate the advantages much morecommonly than was thought in the past. However, simply stating that, for purely humanitar-ian reasons, specialization should be avoided is risky. The reason, of course, is that people dif-fer in what they want from their work and what they are willing to put into it. Some workersprefer not to make decisions about their work, some like to daydream on the job, and othersare simply not capable of performing more complex work. To improve the quality of jobs,leading organizations try different approaches to job design. Two popular contemporaryapproaches are job enrichment and sociotechnical systems.

B E H A V I O R A L C O N S I D E R A T I O N S I N J O B D E S I G N

Specialization of labor





J O B E N R I C H M E N TJob enlargement generally entails adjusting a specialized job to make it more interesting to thejob holder. A job is said to be enlarged horizontally if the worker performs a greater numberor variety of tasks, and it is said to be enlarged vertically if the worker is involved in planning,organizing, and inspecting his or her own work. Horizontal job enlargement is intended tocounteract oversimplification and to permit the worker to perform a “whole unit of work.”Vertical enlargement (traditionally termed job enrichment) attempts to broaden workers’ influ-ence in the transformation process by giving them certain managerial powers over their ownactivities. Today, common practice is to apply both horizontal and vertical enlargement to agiven job and refer to the total approach as job enrichment.

The organizational benefits of job enrichment occur in both quality and productivity.Quality in particular improves dramatically because when individuals are responsible fortheir work output, they take ownership of it and simply do a better job. Also, because theyhave a broader understanding of the work process, they are more likely to catch errors andmake corrections than if the job is narrowly focused. Productivity improvements also occurfrom job enrichment, but they are not as predictable or as large as the improvements in qual-ity. The reason is that enriched work invaribly contains a mix of tasks that (for manual labor)causes interruptions in rhythm and different motions when switching from one task to thenext. Such is not the case for specialized jobs.2

S O C I O T E C H N I C A L S Y S T E M SConsistent with the job enrichment philosophy but focusing more on the interaction betweentechnology and the work group is the sociotechnical systems approach. This approachattempts to develop jobs that adjust the needs of the production process technology to the needsof the worker and work group. The term was developed from studies of weaving mills in Indiaand of coal mines in England in the early 1950s. These studies revealed that work groups couldeffectively handle many production problems better than management if they were permittedto make their own decisions on scheduling, work allocation among members, bonus sharing,and so forth. This was particularly true when variations in the production process requiredquick reactions by the group or when one shift’s work overlapped with other shifts’ work.


e x h i b i t T N 5 . 2Advantages and Disadvantages of Specialization of Labor

ADVANTAGES OF SPECIALIZATION

TO MANAGEMENT TO LABOR

1. Rapid training of the workforce 1. Little or no education required to obtain work

2. Ease in recruiting new workers 2. Ease in learning job

3. High output due to simple, repetitive work

4. Low wages due to ease of substitutability of labor

5. Close control over work flow and workloads

DISADVANTAGES OF SPECIALIZATION

TO MANAGEMENT TO LABOR

1. Difficulty in controlling quality because no one 1. Boredom stemming from repetitive nature ofhas responsibility for entire product work

2. Worker dissatisfaction leading to hidden costs 2. Little gratification from work itself becausearising from turnover, absenteeism, tardiness, of small contribution to each itemgrievances, and intentional disruption of 3. Little or no control over the work pace,production process leading to frustration and fatigue (in

3. Reduced likelihood of improving the process assembly-line situations)because of workers’ limited perspective 4. Little opportunity to progress to a better job

4. Limited flexibility to change the production because significant learning is rarely possible process to produce new or improved products on fractionated work

Job enrichment

Sociotechnical systems





Since those pioneering studies, the sociotechnical approach has been applied in manycountries—often under the heading of “autonomous work groups,” “Japanese-style workgroups,” or employee involvement (EI) teams. Most major American manufacturing companiesuse work teams as the basic building block in so-called high employee involvement plants. Theyare now becoming common in service organizations as well. The benefits of teams are similarto those of individual job enrichment: They provide higher quality and greater productivity(they often set higher production goals than general management), do their own support workand equipment maintenance, and have increased chances to make meaningful improvements.3

One major conclusion from these applications is that the individual or work group requiresa logically integrated pattern of work activities that incorporates the following job designprinciples:

1. Task variety. An attempt must be made to provide an optimal variety of tasks with-in each job. Too much variety can be inefficient for training and frustrating for theemployee. Too little can lead to boredom and fatigue. The optimal level is one thatallows the employee to rest from a high level of attention or effort while working onanother task or, conversely, to stretch after periods of routine activity.

2. Skill variety. Research suggests that employees derive satisfaction from using anumber of skill levels.

3. Feedback. There should be some means for informing employees quickly when theyhave achieved their targets. Fast feedback aids the learning process. Ideally, employeesshould have some responsibility for setting their own standards of quantity and quality.

4. Task identity. Sets of tasks should be separated from other sets of tasks by someclear boundary. Whenever possible, a group or individual employee should haveresponsibility for a set of tasks that is clearly defined, visible, and meaningful. In thisway, work is seen as important by the group or individual undertaking it, and othersunderstand and respect its significance.

5. Task autonomy. Employees should be able to exercise some control over theirwork. Areas of discretion and decision making should be available to them.


Beyond the behavioral components of job design, another aspect warrants con-sideration: the physical side. Indeed, while motivation and work group structure stronglyinfluence job performance, they may be of secondary importance if the job is too demanding

P H Y S I C A L C O N S I D E R A T I O N S I N J O B D E S I G N

THE ERGONOMICS PROGRAM AT BOEING

HAS PROFESSIONALS AVAILABLE TO MAKE

“HOUSE CALLS” AT BOEING SITES. THEY

ANALYZE RISKS IN THE SHOP OR OFFICE

AND SUGGEST NEW PROCESSES, TOOLS

OR PROTECTIVE EQUIPMENT. THIS WORKER

APPLYING SEALANT TO ENGINE MOUNT

SURFACES IN WICHITA REQUIRED THE

MECHANIC, IN THE PAST, TO REACH OVER

HER HEAD FROM AN AWKWARD SEATED

POSITION. A HEIGHT-ADJUSTABLE BELT

ROTATOR NOW HOLDS AND POSITIONS

THE STRUT.





from a physical (or “human factors”) standpoint. One approach to incorporating the physi-cal costs of moderate to heavy work in job design is work physiology. Pioneered byEastman Kodak in the 1960s, work physiology sets work–rest cycles according to theenergy expended in various parts of the job. For example, if a job entails caloric expendi-ture above five calories per minute (the rough baseline for sustainable work), the requiredrest period must equal or exceed the time spent working. Obviously, the harder the work,the more frequent and longer the rest periods. (Exhibit TN5.3 shows caloric requirementsfor various activities.)

Ergonomics is the term used to describe the study of the physical arrangement of the workspace together with the tools used to perform a task. In applying ergonomics, we strive to fitthe work to the body rather than forcing the body to conform to the work. As logical as thismay sound, it is actually a recent point of view.

e x h i b i t T N 5 . 3Calorie Requirements for Various Activities

TYPICAL ENERGY COST IN REQUIRED MINUTES OF REST

TYPE OF ACTIVITY CALORIES PER MINUTE* FOR EACH MINUTE OF WORK

Sitting at rest 1.7 —

Writing 2.0 —

Typing on a computer 2.0 —

Medium assembly work 2.9 —

Shoe repair 3.0 —

Machining 3.3 —

Ironing 4.4 —

Heavy assembly work 5.1 —

Chopping wood 7.5 1

Digging 8.9 2

Tending furnace 12.0 3

Walking upstairs 12.0 3

*Five calories per minute is generally considered the maximum sustainable level throughout the workday.

Work physiology

Ergonomics

In contemporary industry, responsibility for developing work methods in largefirms is typically assigned either to a staff department designated methods analysis or to anindustrial engineering department. In small firms, this activity is often performed by consult-ing firms that specialize in work methods design.

The principal approach to the study of work methods is the construction of charts, suchas operations charts, worker–machine charts, simo (simultaneous motion) charts, and activ-ity charts, in conjunction with time study or standard time data. The choice of which chart-ing method to use depends on the task’s activity level—that is, whether the focus is on(1) a production process, (2) the worker at a fixed workplace, (3) a worker interactingwith equipment, or (4) a worker interacting with other workers (see Exhibit TN5.4).(These charting techniques were introduced in Chapter 5, where they were used to aid inprocess analysis. Chapter 7 introduces the service blueprint that accounts for customerinteractions.)

A P R O D U C T I O N P R O C E S SThe objective in studying a production process is to identify delays, transport distances,processes, and processing time requirements to simplify the entire operation. The underlying

W O R K M E T H O D S






philosophy is to eliminate any step in the process that does not add value to the product. Theapproach is to flowchart the process and then ask the following questions:

What is done? Must it be done? What would happen if it were not done?

Where is the task done? Must it be done at that location or could it be done somewhereelse?

When is the task done? Is it critical that it be done then or is there flexibility in time andsequence? Could it be combined with some other step in the process?

How is the task done? Why is it done this way? Is there another way?

Who does the task? Can someone else do it? Should the worker be of a higher or lowerskill level?

These thought-provoking questions usually help eliminate much unnecessary work andsimplify the remaining work by combining processing steps and changing the order ofperformance.

The process chart is valuable in studying an overall system, though care must be taken to -follow the same item throughout the process. The subject may be a product being manufac-tured, a service being created, or a person performing a sequence of activities. Exhibit TN5.5shows a process chart (and flow diagram) for a clerical operation. Exhibit TN5.6 shows com-mon notation in process charting. Can you suggest any ways to improve this process? (SeeProblem 2.)

W O R K E R A T A F I X E D W O R K P L A C EMany jobs require the worker to remain at a specified workstation. When the nature of thework is primarily manual (such as sorting, inspecting, making entries, or assembly opera-tions), the focus of work design is on simplifying the work method and making the requiredoperator motions as few and as easy as possible.

There are two basic ways to determine the best method when a methods analyst studiesa single worker performing an essentially manual task. The first is to search among theworkers and find the one who performs the job best. That person’s method is then accept-ed as the standard, and others are trained to perform it in the same way. This was basical-ly F. W. Taylor’s approach, though after determining the best method, he searched for“first-class men” to perform according to the method. (A first-class worker possessed thenatural ability to do much more productive work in a particular task than the average.Workers who were not first class were transferred to other jobs.) The second way is to


e x h i b i t T N 5 . 4 Work Methods Design Aids

ACTIVITY OBJECTIVE OF STUDY STUDY TECHNIQUES

Production process Eliminate or combine steps; Flow diagram, serviceshorten transport distance; blueprint, process chartidentify delays

Worker at fixed Simplify method; minimize Operations charts, simoworkplace motions charts; apply principles

of motion economy

Worker’s interaction Minimize idle time; find number Activity chart,with equipment or combination of machines to worker–machine charts

balance cost of worker andmachine idle time

Worker’s interaction Maximize productivity; minimize Activity charts, gang with other workers interference process charts






e x h i b i t T N 5 . 5

Note: Requisition is written by a supervisor, typed by a secretary, approved by a superintendent, and approved by apurchasing agent. Then a purchase order is prepared by a stenographer.

Superintendent

Secretary

Supervisor’s office

Purchasingagent

Stenographer

RESEARCH LABORATORY

Offices

30

180

1

120

30

.5

480

15

480

.5

240

30

240

.25

120

30

240

2237.25

Present MethodProposed Method

SUBJECT CHARTED Requisition for small tools

Chart begins at supervisor’s desk and ends at

typist’s desk in purchasing department

DEPARTMENT Research laboratory

DATE

CHART BY J.C.H.

CHART NO. R136

SHEET NO. 1 OF 1

PROCESS CHART

Requisitions written by supervisor (one copy)

On supervisor’s desk (awaiting messenger)

By messenger to superintendent’s secretary

On secretary’s desk (awaiting typing)

Requisition typed (original requisition copied)

By secretary to superintendent

On superintendent's desk (awaiting messenger)

Examined and approved

On superintendent’s desk (awaiting approval)

To purchasing department

On purchasing agent's desk (awaiting approval)

Examined and approved

On purchasing agent's desk (awaiting messenger)

To stenographer's desk

On stenographer's desk (awaiting typing of purchase order)

Purchase order typed

On stenographer's desk (awaiting transfer to main office)

Total

PROCESS DESCRIPTION

DIST. IN

FEET

TIMEIN

MINS.CHART

SYMBOLS

65

15

20

5

105 3 4 2 8

Flow Diagram and Process Chart of an Office Procedure—Present Method

observe the performance of a number of workers, analyze in detail each step of their work,and pick out the superior features of each worker’s performance. This results in a com-posite method that combines the best elements of the group studied. Frank Gilbreth, thefather of motion study, used this procedure to determine the “one best way” to perform awork task.

Taylor observed actual performance to find the best method; Frank Gilbreth and hiswife Lillian studied movie film. Through micromotion analysis—observing the filmed work





performance frame by frame—the Gilbreths studied work very closely and defined its basicelements, which were termed therbligs (“Gilbreth” spelled backward, with the t and h trans-posed). As part of his work, Gilbreth constructed wire representations of the path of motion.Their study led to the rules or principles of motion economy, such as “The hands should beginand complete the motions at the same time” and “Work should be arranged to permit naturalrhythm.”

Once the motions for performing the task have been identified, an operations chart maybe made, listing the operations and their sequence of performance. For greater detail, a simo(simultaneous motion) chart may be constructed, listing not only the operations but also thetimes for both left and right hands. This chart may be assembled from the data collected witha stopwatch, from analysis of a film of the operation, or from predetermined motion–time data(discussed later in the technical note). Many aspects of poor design are immediately obvious:a hand being used as a holding device (rather than a jig or fixture), an idle hand, or an excep-tionally long time for positioning.

W O R K E R I N T E R A C T I N G W I T H E Q U I P M E N TWhen a person and equipment operate together to perform a productive process, interestfocuses on the efficient use of the person’s time and equipment time. When the operator’sworking time is less than the equipment run time, a worker–machine chart is a useful devicein analysis. If the operator can operate several pieces of equipment, the problem is to find


e x h i b i t T N 5 . 6 Notation for the Process Chart in Exhibit TN5.5

Operation. Something is actually being done. This may be work on a product, some support activity, or anything that is directly productive in nature.

Transportation. The subject of the study (product, service, or person) moves from one location to another.

Inspection. The subject is observed for quality and correctness.

Delay. The subject of the study must wait before starting the next step in the process.

Storage. The subject is stored, such as finished products in inventory or completed papers in a file. Frequently, a distinction is made between temporary storage and permanent storage by inserting a T or P in the triangle.

PACIFIC NORTHWEST NATIONAL

LABORATORY’S REMOTELY OPERATED PIT

VIPER RADIOACTIVE CLEANUP TECHNOLOGY

RESULTS IN 50–75 PERCENT REDUCTION IN

WORKER EXPOSURE. THE REMOTE CAPABILITY

ALLOWS WORKERS TO PERFORM OPERATIONS

IN MORE THAN 600 EQUIPMENT PITS ON

THE HANFORD SITE NEAR RICHLAND,WASHINGTON.





the most economical combination of operator and equipment, when the combined cost of theidle time of a particular combination of equipment and the idle time for the worker is at aminimum.

Worker–machine charts are always drawn to scale, the scale being time as measured bylength. Exhibit TN5.7 shows a worker–machine chart in a service setting. The question hereis, whose utilization use is most important?

W O R K E R S I N T E R A C T I N G W I T H O T H E RW O R K E R SThe degree of interaction among teams may be as simple as one operator handing a part toanother, or as complex as a cardiovascular surgical team of doctors, nurses, anesthesiologist,operator of an artificial heart machine, X-ray technician, standby blood donors, and patholo-gist (and perhaps a minister to pray a little).

An activity or a gang process chart is useful in plotting each individual’s activities on atime scale similar to that of the worker–machine chart. A gang process chart is usuallyemployed to trace the interaction of a number of workers with machines in a specifiedoperating cycle to find the best combination of workers and machines. An activity chart isless restrictive and may be used to follow the interaction of any group of operators, with orwithout equipment being involved. Such charts are often used to study and define eachoperation in an ongoing repetitive process, and they are extremely valuable in developing a


e x h i b i t T N 5 . 7

Person Machine

Customer Coffee Grinder

Time in Seconds

Ask grocer for 1 pound of coffee (brand and grind)

Wait

Wait

Wait

Receive coffee from grocer, pay grocer, and

receive change

Idle

Timein sec.

Timein sec.

Timein sec.

1.2.

3.

4.

5.

5

15

21

12

17

5

15

21

12

17

5

15

21

12

17

Grind coffee

Idle

Idle

Idle

Summary

Idle time

Working time

Total cycle time

Utilizationpercentage

Customer Clerk Coffee Grinder

48 sec.

22

70

2270

� 31%

21 sec.

49

70

49 sec.

21

70

Customer utilization � Clerk utilization � Machine utilization �4970

� 70% 2170

� 30%

Clerk

Give coffee to customer, wait for customer to pay for coffee, receive money, and

make change

Listen to order

Get coffee and put in machine, set grind, and

start grinder

Idle while machine grinds

Stop grinder, place coffee in package, and close it

The customer, the clerk, and the coffee grinder (machine) are involved in this operation. It required 1 minute 10 seconds for the customer to purchase a pound of coffee in this store. During this time the customer spent 22 seconds, or 31 percent of the time, giving the clerk his order, receiving the ground coffee, and paying the clerk. He was idle the remaining 69 percent of the time. The clerk worked 49 seconds, or 70 percent of the time, and was idle 21 seconds, or 30 percent of the time. The coffee grinder was in operation 21 seconds, or 30 percent of the time, and was idle 70 percent of the time.

10

20

30

40

50

60

70

0

Worker–Machine Chart for a Gourmet Coffee Store






e x h i b i t T N 5 . 8 Activity Chart of Emergency Tracheotomy

NURSE FIRST DOCTOR ORDERLY SECOND DOCTORNURSE

SUPERVISOR SCRUB NURSE

Detects problemNotifies doctor

Makes diagnosis

Notifies nurse supervisor

Notifies second doctor

Notifies orderly

Moves patientto OR

Helps patientto breathe

Moves to OR

Scrubs

Dons gown and gloves

Performstracheotomy

Assures availability of laryngoscope and endotracheal tube

Operates laryngoscope and inserts endotracheal tube

Calls for IPPB machine

Opens ORCalls scrub

nurse

Moves to ORSets up equipment

Gets mobile cart

Moves patient to OR

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

SOURCE: DATA TAKEN FROM H. E. SMALLEY AND J. FREEMAN, HOSPITAL INDUSTRIAL ENGINEERING (NEW YORK: REINHOLD, 1966), P. 409.

The fundamental purpose of work measurement is to set time standards for ajob. Such standards are necessary for four reasons:

1. To schedule work and allocate capacity. All scheduling approaches require someestimate of how much time it takes to do the work being scheduled.

2. To provide an objective basis for motivating the workforce and measuringworkers’ performance. Measured standards are particularly critical where output-based incentive plans are employed.

3. To bid for new contracts and to evaluate performance on existing ones.Questions such as “Can we do it?” and “How are we doing?” presume the existenceof standards.

4. To provide benchmarks for improvement. In addition to internal evaluation,benchmarking teams regularly compare work standards in their company with thoseof similar jobs in other organizations.

W O R K M E A S U R E M E N T A N D S T A N D A R D SWork measurement

standardized procedure for a specific task. Exhibit TN5.8, for example, shows an activitychart for a hospital’s emergency routine in performing a tracheotomy (opening a patient’sthroat surgically to allow the patient to breathe), where detailed activity analysis is criticaland any delay could be fatal.





Work measurement and its resulting work standards have been controversial since Taylor’stime. Much of this criticism has come from unions, which argue that management often setsstandards that cannot be regularly achieved. (To counter this, in some contracts, the industrialengineer who sets the standard must demonstrate that he or she can do the job over a repre-sentative period of time at the rate that was set.) There is also the argument that workers whofind a better way of doing the job get penalized by having a revised rate set. (This is com-monly called rate cutting.)

With the widespread adoption of W. Edwards Deming’s ideas, the subject has receivedrenewed criticism. Deming argued that work standards and quotas inhibit process improve-ment and tend to focus the worker’s efforts on speed rather than quality.

Despite these criticisms, work measurement and standards have proved effective. Muchdepends on sociotechnical aspects of the work. Where the job requires work groups to func-tion as teams and create improvements, worker-set standards often make sense. On the otherhand, where the job really boils down to doing the work quickly, with little need for creativ-ity (such as delivering packages for UPS as the box on page 192 relates), tightly engineered,professionally set standards are appropriate.

W O R K M E A S U R E M E N T T E C H N I Q U E SThere are four basic techniques for measuring work and setting standards. These consist oftwo direct observational methods and two indirect methods: The direct methods are timestudy, which uses a stopwatch to time the work, and work sampling, which entails record-ing random observations of a person or teams at work. The two indirect methods arepredetermined motion-time data systems (PMTS), which sum data from tables of genericmovement times developed in the laboratory to arrive at a time for the job (the most widelyused are proprietary systems—Methods Time Measurement [MTM] and Most WorkMeasurement System [MOST]), and elemental data, which sums times from a database ofsimilar combinations of movements to arrive at job time. The choice of techniques depends onthe level of detail desired and the nature of the work itself. Highly detailed, repetitive work usu-ally calls for time study and predetermined motion-time data analysis. When work is done inconjunction with fixed-processing-time equipment, elemental data are often used to reduce theneed for direct observation. When work is infrequent or entails a long cycle time, work samplingis the tool of choice. (See box “What the Pros Say . . . About Work Measurement Applicationsin Retailing” for an example of how the different techniques are used in a service setting.)

We now turn to a discussion of the technical details of time study and work sampling. Atime study is generally made with a stopwatch, either on the spot or by analyzing a videotapefor the job. The job or task to be studied is separated into measurable parts or elements, andeach element is timed individually.

Some general rules for breaking down the elements are

1. Define each work element to be short in duration but long enough so that it can betimed with a stopwatch and the time can be written down.


“We use the predetermined motion time system, MOST, to setstandards in retail stores, such as the Gap. In addition toMOST, we use work sampling and some automated time studyin the standards development process. MOST is used for themajority of the manual tasks—receiving product, stockingproduct, conditioning the store, checking a person out, etc.

Work sampling is used to determine frequencies, understandcustomer shopping behaviors and for standards validation.Time study is used for process related activities such asresponse time on a credit card or POS machine, and for cus-tomer engagement/sales activities.”

WHAT THE PROS SAY . . . ABOUT WORK MEASUREMENT APPLICATIONS IN RETAILING

SOURCE: JEFF PERETIN OF H. B. MAYNARD (ONE OF THE WORLD’S LEADING CONSULTING FIRMS IN WORK MEASUREMENT).

Time studyWork sampling

Predetermined motion-time data systems (PMTS)

Elemental data





2. If the operator works with equipment that runs separately (meaning the operatorperforms a task and the equipment runs independently), separate the actions of theoperator and of the equipment into different elements.

3. Define any delays by the operator or equipment into separate elements.

After a number of repetitions, the collected times are averaged. (The standard devia-tion may be computed to give a measure of variance in the performance times.) Theaveraged times for each element are added, yielding the performance time for the opera-tor. However, to make this operator’s time usable for all workers, a measure of speed orperformance rating must be included to “normalize” the job. The application of a ratingfactor gives what is called normal time. For example, if an operator performs a task intwo minutes and the time-study analyst estimates her to be performing about 20 percentfaster than normal, the operator’s performance rating would be 1.2, or 120 percent ofnormal. The normal time would be computed as 2 minutes × 1.2, or 2.4 minutes. Inequation form,

Normal time = Observed performance time per unit × Performance rating

In this example, denoting normal time by NT,

NT = 2(1.2) = 2.4 minutes

When an operator is observed for a period of time, the number of units produced during thistime, along with the performance rating, gives

NT = Time worked

Number of units produced× Performance rating


Normal time

WORK MEASUREMENT AT UNITED PARCEL SERVICE

Grabbing a package under his arm, Joseph Polise, a driver forUnited Parcel Service (UPS), bounds from his brown deliverytruck toward an office building. A few paces behind him,Marjorie Cusack, a UPS industrial engineer, clutches a digitaltimer.

Her eyes fixed on him, Cusack counts Polise’s steps and timeshis contact with customers. Scribbling on a clipboard, Cusackrecords every second taken up by stoplights, traffic, detours,doorbells, walkways, stairways, and coffee breaks.

Eighty thousand UPS drivers travel 1.2 billion miles per year anddeliver more than 13 million packages a day. On average, UPSdrivers move in and out of the truck 200 times a day. An unnec-essary step or indirect travel path reduces the effectiveness ofthe driver and impacts service to the customer. One minutesaved each day saves the company $5 million annually. For thisreason, UPS spends millions each year to train its drivers inproper, efficient, and safe work methods.

Approximately 1,118 industrial engineers at UPS ensureefficient and reliable customer service by conducting timestudies on drivers’ routes to provide job method instruction.They have measured even the finest details of the drivers’job, including determining on which finger drivers should con-sistently carry their key rings to avoid losing them.

In addition to developing specific job methods, UPS providesdrivers with custom-built package vehicles with featuresincluding

• Domed seats that allow the driver to slide on and offeasily at each delivery stop.

• A drop floor well located behind the rear wheel hous-ing, making the rear of the vehicle only a short stepfrom the ground for easy entry.

• Bulkhead doors that allow easy access to the packagecompartment and save the driver steps in selectingparcels for delivery.

SOURCE: ABSTRACTED FROM D. MACHALABA, “UP TO SPEED: UNITED PARCEL SERVICE GETS DELIVERIES DONE BY DRIVING ITS WORKERS,” THE WALL STREET JOURNAL,APRIL 22, 1986, P. 1. INFORMATION PROVIDED BY UPS, 1999.





Standard time is derived by adding to normal time allowances for personal needs (suchas washroom and coffee breaks), unavoidable work delays (such as equipment breakdown orlack of materials), and worker fatigue (physical or mental). Two such equations are

Standard time = Normal time + (Allowances × Normal time)

or

[TN5.1] ST = NT (1 + Allowances)

and

[TN5.2] ST = NT

1 − Allowances

Equation (TN5.1) is most often used in practice. If one presumes that allowances shouldbe applied to the total work period, then equation (TN5.2) is the correct one. To illustrate, sup-pose that the normal time to perform a task is one minute and that allowances for personalneeds, delays, and fatigue total 15 percent; then by equation (TN5.1)

ST = 1(1 + 0.15) = 1.15 minutes

In an eight-hour day, a worker would produce 8 × 60/1.15, or 417 units. This implies 417minutes working and 480 − 417 (or 63) minutes for allowances.

With equation (TN5.2),

ST = 1

1 − 0.15= 1.18 minutes

In the same eight-hour day, 8 × 60/1.18 (or 408) units are produced with 408 workingminutes and 72 minutes for allowances. Depending on which equation is used, there is a dif-ference of nine minutes in the daily allowance time.

EXAMPLE TN5.1: Time Study for a Four-Element JobExhibit TN5.9 shows a time study of 10 cycles of a four-element job. For each element, there is a spacefor the watch readings that are recorded in 100ths of a minute. Space also is provided for summarizingthe data and applying a performance rating.


e x h i b i t T N 5 . 9

Time Study Observation Sheet

Identification of Operation ASSEMBLE 24" � 36" CHART BLANKS Date 10/9Began Timing: 9:26Ended Timing: 9:32

Operator 109 Observer

Element Description and Breakpoint

Fold over end(grasp stapler)Staple five times(drop stapler)Bend and insert wire(drop pliers)Dispose of finished chart(touch next sheet)

.07

.07

.16

.23

.22

.45

.09

.54

.07

.61

.14

.75

.25

.00

.09

.09

.05

.14

.14

.28

.22

.50

.10

.60

.07

.67

.15

.82

.25

.07

.08

.15

.09

.24

.16

.40

.23

.63

.09

.72

.06

.78

.16

.94

.23

.17

.11

.28

.05

.33

.14

.47

.21

.68

.12

.80

.08

.88

.17

.05

.26

.31

.08

.39

.08

.47

.14

.61

.25

.86

.17

.03

.06

.09

.15

.24

.24

.48

.08

.56

.68

1.51

2.36

1.01

.07

.15

.24

.10

.90

1.05

1.00

.90

.06

.16

.24

.09

Normal cycle time _______ + Allowance __________________ � Std. time _____________

Approval

0.55 (0.55 � 0.143) or 0.08 0.63 min./pc.

0.55normalminute

forcycle

Cycles Summary

0.001 2 3 4 5 6 7 8 9 10 ∑T PR NT

1

2

3

4

5

6

10

T

Time-Study Observation Sheet

Standard time





SOLUTIONThe value of �T is obtained by averaging the observed data. PR denotes the performance rating and ismultiplied with T to obtain the normal time (NT) for each element. The normal time for the job is thesum of the element normal times. The standard time, calculated according to equation (TN5.1), is givenat the bottom of Exhibit TN5.9. •

How many observations are enough? Time study is really a sampling process; that is, wetake relatively few observations as being representative of many subsequent cycles to be per-formed by the worker. Based on a great deal of analysis and experience, Benjamin Niebel’stable shown in Exhibit TN5.10 indicates that “enough” is a function of cycle length and num-ber of repetitions of the job over a one-year planning period.

A second common technique for measuring a job is called work sampling. As the name sug-gests, work sampling involves observing a portion or sample of the work activity. Then, basedon the findings in this sample, statements can be made about the activity. For example, if wewere to observe a fire department rescue squad at 100 random times during the day and foundit was involved in a rescue mission for 30 of the 100 times (en route, on site, or returning froma call), we would estimate that the rescue squad spends 30 percent of its time directly on res-cue mission calls. (The time it takes to make an observation depends on what is beingobserved. Many times, only a glance is needed to determine the activity, and the majority ofstudies require only several seconds’ observation.)

Observing an activity even 100 times may not, however, provide the accuracy desired inthe estimate. To refine this estimate, three main issues must be decided. (These points are dis-cussed later in this section, along with an example.)

1. What level of statistical confidence is desired in the results?2. How many observations are necessary?3. Precisely when should the observations be made?


e x h i b i t T N 5 . 1 0 Guide to Number of Cycles to Be Observed in a Time Study

MINIMUM NUMBER OF CYCLES OF STUDY (ACTIVITY)WHEN TIME PER CYCLE

IS MORE THAN OVER 10,000 PER YEAR 1,000–10,000 UNDER 1,000

8 hours 2 1 13 3 2 12 4 2 11 5 3 2

48 minutes 6 3 230 8 4 320 10 5 412 12 6 58 15 8 65 20 10 83 25 12 102 30 15 121 40 20 15.7 50 25 20.5 60 30 25.3 80 40 30.2 100 50 40.1 120 60 50

Under .1 140 80 60

Service

SOURCE: B. W. NIEBEL, MOTION AND TIME STUDY, 9TH ED. (BURR RIDGE, IL: RICHARD D. IRWIN, 1993), P. 390. THE MCGRAW-HILL

COMPANIES, INC. USED WITH PERMISSION.





The three primary applications for work sampling are

1. Ratio delay to determine the activity-time percentage for personnel or equipment. Forexample, management may be interested in the amount of time a machine is runningor idle.

2. Performance measurement to develop a performance index for workers. When theamount of work time is related to the quantity of output, a measure of performance isdeveloped. This is useful for periodic performance evaluation.

3. Time standards to obtain the standard time for a task. When work sampling is used forthis purpose, however, the observer must be experienced because he or she mustattach a performance rating to the observations.

The number of observations required in a work-sampling study can be fairly large, rang-ing from several hundred to several thousand, depending on the activity and desired degreeof accuracy. Although the number can be computed from formulas, the easiest way is torefer to a table such as Exhibit TN5.11, which gives the number of observations neededfor a 95 percent confidence level in terms of absolute error. Absolute error is the actualrange of the observations. For example, if a clerk is idle 10 percent of the time and thedesigner of the study is satisfied with a 2.5 percent range (meaning that the true percent-age lies between 7.5 and 12.5 percent), the number of observations required for the worksampling is 576. A 2 percent error (or an interval of 8 to 12 percent) would require 900observations.

Five steps are involved in making a work-sampling study:

1. Identify the specific activity or activities that are the main purpose for the study. Forexample, determine the percentage of time that equipment is working, idle, or underrepair.

2. Estimate the proportion of time of the activity of interest to the total time (e.g., thatthe equipment is working 80 percent of the time). These estimates can be made fromthe analyst’s knowledge, past data, reliable guesses from others, or a pilot work-sampling study.

3. State the desired accuracy in the study results.4. Determine the specific times when each observation is to be made.5. At two or three intervals during the study period, recompute the required sample size

by using the data collected thus far. Adjust the number of observations if appropriate.

The number of observations to be taken in a work-sampling study is usually dividedequally over the study period. Thus, if 500 observations are to be made over a 10-dayperiod, observations are usually scheduled at 500/10, or 50 per day. Each day’s observa-tions are then assigned a specific time by using a random number table.

EXAMPLE TN5.2: Work Sampling Applied to NursingThere has been a long-standing argument that a large amount of nurses’ hospital time is spent onnonnursing activities. This, the argument goes, creates an apparent shortage of well-trained nursing per-sonnel, wastes talent, hinders efficiency, and increases hospital costs because nurses’ wages are thehighest single cost in the operation of a hospital. Further, pressure is growing for hospitals and hospitaladministrators to contain costs. With that in mind, let us use work sampling to test the hypothesis thata large portion of nurses’ time is spent on nonnursing duties.

SOLUTIONAssume at the outset that we have made a list of all the activities that are part of nursing and will makeour observations in only two categories: nursing and nonnursing activities. Actually, there is muchdebate on what constitutes nursing activity. For instance, is talking to a patient a nursing duty? (An


Service






e x h i b i t T N 5 . 11 Number of Observations Required for a Given Absolute Error at Various Values of p, with95 Percent Confidence Level

PERCENTAGE OF TOTAL TIME ABSOLUTE ERROROCCUPIED BY ACTIVITY OR

DELAY, P ±1.0% ±1.5% ±2.0% ±2.5% ±3.0% ±3.5%

1 or 99 396 176 99 63 44 322 or 98 784 348 196 125 87 643 or 97 1,164 517 291 186 129 954 or 96 1,536 683 384 246 171 1255 or 95 1,900 844 475 304 211 1556 or 94 2,256 1,003 564 361 251 1847 or 93 2,604 1,157 651 417 289 2138 or 92 2,944 1,308 736 471 327 2409 or 91 3,276 1,456 819 524 364 26710 or 90 3,600 1,600 900 576 400 29411 or 89 3,916 1,740 979 627 435 32012 or 88 4,224 1,877 1,056 676 469 34413 or 87 4,524 2,011 1,131 724 503 36914 or 86 4,816 2,140 1,204 771 535 39315 or 85 5,100 2,267 1,275 816 567 41616 or 84 5,376 2,389 1,344 860 597 43917 or 83 5,644 2,508 1,411 903 627 46118 or 82 5,904 2,624 1,476 945 656 48219 or 81 6,156 2,736 1,539 985 684 50220 or 80 6,400 2,844 1,600 1,024 711 52221 or 79 6,636 2,949 1,659 1,062 737 54222 or 78 6,864 3,050 1,716 1,098 763 56023 or 77 7,084 3,148 1,771 1,133 787 57824 or 76 7,296 3,243 1,824 1,167 811 59625 or 75 7,500 3,333 1,875 1,200 833 61226 or 74 7,696 3,420 1,924 1,231 855 62827 or 73 7,884 3,504 1,971 1,261 876 64428 or 72 8,064 3,584 2,016 1,290 896 65829 or 71 8,236 3,660 2,059 1,318 915 67230 or 70 8,400 3,733 2,100 1,344 933 68631 or 69 8,556 3,803 2,139 1,369 951 69832 or 68 8,704 3,868 2,176 1,393 967 71033 or 67 8,844 3,931 2,211 1,415 983 72234 or 66 8,976 3,989 2,244 1,436 997 73335 or 65 9,100 4,044 2,275 1,456 1,011 74336 or 64 9,216 4,096 2,304 1,475 1,024 75337 or 63 9,324 4,144 2,331 1,492 1,036 76138 or 62 9,424 4,188 2,356 1,508 1,047 76939 or 61 9,516 4,229 2,379 1,523 1,057 77740 or 60 9,600 4,266 2,400 1,536 1,067 78441 or 59 9,676 4,300 2,419 1,548 1,075 79042 or 58 9,744 4,330 2,436 1,559 1,083 79543 or 57 9,804 4,357 2,451 1,569 1,089 80044 or 56 9,856 4,380 2,464 1,577 1,095 80445 or 55 9,900 4,400 2,475 1,584 1,099 80846 or 54 9,936 4,416 2,484 1,590 1,104 81147 or 53 9,964 4,428 2,491 1,594 1,107 81348 or 52 9,984 4,437 2,496 1,597 1,109 81549 or 51 9,996 4,442 2,499 1,599 1,110 816

50 10,000 4,444 2,500 1,600 1,111 816

Note: Number of observations is obtained from the formula E = Z

√p (1 − p)

Nand the required sample (N) is N =

E2Z2p (1 − p)

where E = Absolute errorp = Percentage occurrence of activity or delay being measuredN = Number of random observations (sample size)Z = Number of standard deviations to give desired confidence level (e.g., for 90 percent confidence, Z = 1.65; for 95 percent,

Z = 1.96; for 99 percent, Z = 2.23). In this table Z = 2.

Excel: Sample-size





expanded study could list all nursing activities to determine the portion of time spent in each.)Therefore, when we observe during the study and find the nurse performing one of the duties on thenursing list, we simply place a tally mark in the nursing column. If we observe anything besides nurs-ing activities, we place a tally mark in the nonnursing column.

We can now plan the study. Assume that we (or the nursing supervisor) estimate that nurses spend60 percent of their time in nursing activities. Assume that we would like to be 95 percent confident thatfindings of our study are within the absolute error range of ±3 percent; that is, if our study shows nursesspend 60 percent of their time on nursing duties, we want to be 95 percent confident that the true per-centage lies between 57 and 63 percent. From Exhibit TN5.11, we find that 1,067 observations arerequired for 60 percent activity time and ±3 percent error. If our study is to take place over 10 days, westart with 107 observations per day.

To determine when each day’s observations are to be made, we assign specific numbers to eachminute and use a random number table to set up a schedule. If the study extends over an eight-hour shift,we can assign numbers to correspond to each consecutive minute. For this study it is likely the nightshift would be run separately because nighttime nursing duties are considerably different from daytimeduties. Exhibit TN5.12A shows the assignment of numbers to corresponding minutes. For simplicity,because each number corresponds to one minute, a three-number scheme is used, with the second andthird numbers corresponding to the minute of the hour. A number of other schemes would also be appro-priate. If a number of studies are planned, a computer program may be used to generate a randomizedschedule for the observation times.

If we refer to a random number table and list three-digit numbers, we can assign each number to atime. The random numbers in Exhibit TN5.12B demonstrate the procedure for seven observations.


e x h i b i t T N 5 . 1 2Sampling Plan for Nurses’ActivitiesA. Assignment of Numbers to Corresponding MinutesB. Determination of Observation TimesC. Observation Schedule

A.

ASSIGNED

TIME NUMBERS

7:00–7:59 A.M. 100–159

8:00–8:59 A.M. 200–259

9:00–9:59 A.M. 300–359

10:00–10:59 A.M. 400–459

11:00–11:59 A.M. 500–559

12:00–12:59 P.M. 600–659

1:00–1:59 P.M. 700–759

2:00–2:59 P.M. 800–859

B.

RANDOM CORRESPONDING TIME

NUMBER FROM THE LIST IN TN5.12A

669 Nonexistent

831 2:31 P.M.

555 11:55 A.M.

470 Nonexistent

113 7:13 A.M.

080 Nonexistent

520 11:20 A.M.

204 8:04 A.M.

732 1:32 P.M.

420 10:20 A.M.

C.

OBSERVATION SCHEDULE TIME NURSING ACTIVITY(√) NONNURSING ACTIVITY(√)

1 7:13 A.M.

2 8:04 A.M.

3 10:20 A.M.

4 11:20 A.M.

5 11:55 A.M.

6 1:32 P.M.

7 2:31 P.M.






e x h i b i t T N 5 . 1 3 Deriving a Time Standard Using Work Sampling

INFORMATION SOURCE OF DATA DATA FOR ONE DAY

Total time expended by Computer payroll system 480 min.operator (working time andidle time)

Number of parts produced Inspection department 420 pieces

Working time in percent Work sampling 85%

Idle time in percent Work sampling 15%

Average performance index Work sampling 110%

Total allowances Company time-study manual 15%

This procedure is followed to generate 107 observation times, and the times are rearranged chrono-logically for ease in planning. Rearranging the times determined in Exhibit TN5.12B gives the totalobservations per day shown in Exhibit TN5.12C (for our sample of seven).

To be perfectly random in this study, we should also “randomize” the nurse we observe each time.(The use of various nurses minimizes the effect of bias.) In the study, our first observation is madeat 7:13 A.M. for Nurse X. We walk into the nurse’s area and, on seeing the nurse, check either a nursingor a nonnursing activity. Each observation need be only long enough to determine the class of activity—in most cases only a glance is needed. At 8:04 A.M. we observe Nurse Y. We continue in this way to theend of the day and the 107 observations. At the end of the second day (and 214 observations), we decideto check for the adequacy of our sample size.

Let us say we made 150 observations of nurses working and 64 of them not working, which gives70.1 percent working. From Exhibit TN5.11, this corresponds to 933 observations. Because we havealready taken 214 observations, we need take only 719 over the next eight days, or 90 per day.

When the study is half over, another check should be made. For instance, if days 3, 4, and 5 showed55, 59, and 64 working observations, the cumulative data would give 328 working observations of atotal 484, or a 67.8 percent working activity. For a ±3 percent error, Exhibit TN5.11 shows the samplesize to be about 967, leaving 483 to be made—at 97 per day—for the following five days. Another com-putation should be made before the last day to see if another adjustment is required. If after the 10th dayseveral more observations are indicated, these can be made on day 11.

If at the end of the study we find that 66 percent of nurses’ time is involved with what has beendefined as nursing activity, there should be an analysis to identify the remaining 34 percent.Approximately 12 to 15 percent is justifiable for coffee breaks and personal needs, which leaves 20 to22 percent of the time that must be justified and compared to what the industry considers ideal levels ofnursing activity. To identify the nonnursing activities, a more detailed breakdown could have been orig-inally built into the sampling plan. Otherwise, a follow-up study may be in order. •

As mentioned earlier, work sampling can be used to set time standards. To do this, the ana-lyst must record the subject’s performance rate (or index) along with working observations.Exhibit TN5.13 gives a manufacturing example that demonstrates how work sampling can beused for calculating standard time.

( )×

( )×

( )

=( )

×( )

= 1.26 minutes

Total time Working time Performancein minutes proportion index 1Standard time � ×per piece

Total number of pieces produced 1 − Allowances

480 × 0.85 × 1.10 1420 1 − 0.15





W O R K S A M P L I N G C O M P A R E D T O T I M E S T U D YWork sampling offers several advantages:

1. Several work-sampling studies may be conducted simultaneously by one observer.2. The observer need not be a trained analyst unless the purpose of the study is to

determine a time standard.3. No timing devices are required.4. Work of a long cycle time may be studied with fewer observer hours.5. The duration of the study is longer, which minimizes effects of short-period variations.6. The study may be temporarily delayed at any time with little effect.7. Because work sampling needs only instantaneous observations (made over a longer

period), the operator has less chance to influence the findings by changing his or herwork method.

When the cycle time is short, time study is more appropriate than work sampling. Onedrawback of work sampling is that it does not provide as complete a breakdown of elementsas time study. Another difficulty with work sampling is that observers, rather than followinga random sequence of observations, tend to develop a repetitive route of travel. This mayallow the time of the observations to be predictable and thus invalidate the findings. A thirdfactor—a potential drawback—is that the basic assumption in work sampling is that all obser-vations pertain to the same static system. If the system is in the process of change, work sam-pling may give misleading results.


The third piece of the job design equation is the paycheck. This section brieflyreviews common methods for setting financial incentives.

B A S I C C O M P E N S A T I O N S Y S T E M SThe main forms of basic compensation are hourly pay, straight salary, piece rate, and com-missions. The first two are based on time spent on the job, with individual performancerewarded by an increase in the base rate. Piece rate plans reward on the basis of direct dailyoutput. (A worker is paid $5 a unit; thus, by producing 10 units per day, the worker earns $50.)Sometimes a guaranteed base is included in a piece-rate plan; a worker would receive thisbase amount regardless of output, plus a piece-rate bonus. (For example, the worker’s hourlybase pay is $8, so this coupled with $50 piece-rate earnings gives the worker $114 for aneight-hour day.) Commissions may be thought of as sales-based piece rates and are calculatedin the same general way.

The two broad categories of financial incentive plans are individual or small-groupincentive plans and organizationwide plans.

I N D I V I D U A L A N D S M A L L - G R O U PI N C E N T I V E P L A N SIndividual and work group plans traditionally have rewarded performance by using output(often defined by piece rates) and quality measures. Quality is accounted for by a quality ad-justment factor, say, a percentage of rework. (For example: Incentive pay = Total output ×[1 − Percent deduction for rework].) In recent years, skill development also has been re-warded. Sometimes called pay for knowledge, this means a worker is compensated for learn-ing new tasks. This is particularly important in job shops using group technology, as well asin banking, where supervisors’ jobs require knowledge of new types of financial instrumentsand selling approaches.

F I N A N C I A L I N C E N T I V E P L A N S





AT&T, for example, instituted incentive programs for its managers—an IndividualIncentive Award (IIA) and a Management Team Incentive Award (MTIA). The IIA provideslump-sum bonuses to outstanding performers. These outstanding performers are determinedby individual performance ratings accompanied by extensive documentation. The lump-sumbonus can range between 15 and 30 percent of base pay.

MTIAs are granted to members of specific divisions or units. Appropriate division or unitgoals are established at the beginning of the year. The goals include department service objec-tives and interdepartmental goals. A typical MTIA could call for a standard amount equiva-lent to 1.5 percent of wages plus overtime for the next three years based on performance inthe current year.

O R G A N I Z A T I O N W I D E P L A N SProfit sharing and gain sharing are the major types of organizationwide plans. Profit sharingsimply distributes a percentage of corporate profits across the workforce. In the United States,at least one-third of all organizations have profit sharing. In Japan, most major companiesgive profit-based bonuses twice a year to all employees. Such bonuses may range from50 percent of salaries, in good years, to nothing in bad years.

Gain sharing also involves giving organizationwide bonuses, but it differs from profit shar-ing in two important respects. First, it typically measures controllable costs or units of out-put, not profits, in calculating a bonus. Second, gain sharing is always combined with a par-ticipative approach to management. The original and best-known gainsharing plan is theScanlon Plan.

In the late 1930s, the Lapointe Machine and Tool Company was on the verge of bank-ruptcy, but through the efforts of union president Joseph Scanlon and company management,the Scanlon Plan was devised to save the company by reducing labor costs. In essence, thisplan started with the normal labor cost within the firm. Workers as a group were rewarded forany reductions in labor cost below this base cost. The plan’s success depended on committeesof workers throughout the firm whose purpose was to search out areas for cost saving and todevise ways of improvement. There were many improvements, and the plan did, in fact, savethe company.

The basic elements of the Scanlon Plan are

1. The ratio. The ratio is the standard that serves as a measure for judging businessperformance. It can be expressed as

Ratio = Total labor cost

Sales value of production

2. The bonus. The amount of bonus depends on the reduction in costs below the pre-set ratio.

3. The production committee. The production committee is formed to encourageemployee suggestions to increase productivity, improve quality, reduce waste, and soforth. The purpose of a production committee is similar to that of a quality circle.

4. The screening committee. The screening committee consists of top managementand worker representatives who review monthly bonuses, discuss production prob-lems, and consider improvement suggestions.

Though originally established in small companies such as Lapointe, Lincoln ElectricCompany, and Herman Miller, gain sharing has been installed by large firms such as TRW,General Electric, Motorola, and Firestone. These companies apply gain sharing to organi-zational units. Motorola, for example, has virtually all its plant employees covered by gainsharing. These plans are increasing because “they are more than just pay incentive plans;they are a participative approach to management and are often used as a way to install







TYPE OF PLAN APPLICATION ADVANTAGES DISADVANTAGES

Merit pay Individual • Allows management to • Can be arbitrary andtarget specific unbiased when incorrectlybehavior and to easily administered.evolve criteria over • Often not clearly tied totime. business goals.

Profit sharing Group • Ties business • Often individual or groupperformance to behavior is not correlatedemployee reward. to business performance.

Gain sharing Group • Specific group • Often focuses excessivelyperformance directly on cost control.tied to employee • More applicable forreward. tactical improvements than

strategic changes.

Lump-sum bonuses and Either • Allows management to • Often used for and seen asindividual bonuses vary criteria and deferred compensation.

magnitude of reward; • Not always a tie toable to target specific business goals oractions and behavior. performance.

Pay-for-knowledge Individual • Allows management to • May not impact businesstarget specific types of performance unlessskills and personal management targetsgrowth. correct skills and applies

new skills effectively.

Piece rate Either • Allows management to • May lead to undesirabletarget specific output competition amonggoals. workers.

• Standards must be kept upto date.

SOURCE: MODIFIED FROM C. GIFFI, A. ROTH, AND G. SEAL, COMPETING IN WORLD-CLASS MANUFACTURING, AMERICA’S 21ST CENTURY CHALLENGE

(HOMEWOOD, IL: BUSINESS ONE IRWIN, 1990). © 1990 MCGRAW-HILL COMPANIES, INC. USED BY PERMISSION.

At the outset of this technical note we identified current trends in job design.What will the future hold? One thing is clear: Globalization and the successful applicationof sophisticated process technologies will make the human element even more importantto operations competitiveness than before. Giffi, Roth, and Seal speculate that “the 21stcentury will be marked by the human resource renaissance.” In their view, this renaissancewill be characterized by companies actively cultivating their human resources throughcareful selection and training of the best and brightest employees, implementing innovativeteam-based employee involvement programs, developing genuinely participative manage-ment approaches, and continually retraining their employees.5 What is the future of thetime study techniques also addressed in this technical note? In our opinion, they willalways have application to analyzing work methods, to setting work standards, and tostructuring incentive plans. (See the Breakthrough box “Linking Standards and Incentivesat the Gap Distribution Centers.”)

C O N C L U S I O N

e x h i b i t T N 5 . 1 4Comparison of Common Reward/Incentive Plans

participative management.”4 The typical applications of the plans are discussed, along withmerit pay, in Exhibit TN5.14.






breakthroughLINKING STANDARDS AND INCENTIVES ATTHE GAP DISTRIBUTION CENTERS

Fashion retailer The Gap, Inc., believes strongly in the value ofengineered labor standards, which have been implemented insome Gap distribution centers for more than a decade. Theretailer has engineered labor standards in place for receiving,stocking, order filling, and shipping. These standards “are ournumber one means of communication” with warehouse asso-ciates, explains Jay Ninah, senior planning engineer for TheGap. Reports on individual and group performance are postedweekly and associates meet with their supervisors on amonthly basis to discuss their performance. As a result, asso-ciates always know what level of performance is expected ofthem and how their individual performance compares to thestandard.

Posting individual as well as group performance data keepswith The Gap’s open-book policy, Ninah says. “It’s not thereto intimidate anyone, but to share our findings. It also cre-ates an expectation that the standards are correctly set. If85 percent to 90 percent are achieving the standard, thosenot meeting the standard can determine whether they’reusing a method that’s not the best one.” The Gap coachesbased on its standards. Ninah states, “As we identify lower-performing individuals, we have a structured coachingprocess,” whereby supervisors work with associates on theareas they need to improve. Associates have an extra push toimprove their work, thanks to an incentive program tied toperformance. The program was pioneered last year, Ninahsays, and is getting a good response from associates. “Insteadof giving an across-the-board increase, we give it based onperformance,” he says.

GLOBAL STANDARDS, LOCAL APPLICATIONThe Gap’s distribution network is made up of 18 facilities thatrange from brand-new warehouses with the latest technolo-gy to others that are 20 years old. To make sure that the laborstandards are appropriate for each facility, The Gap’s centralengineering group in Erlanger, Kentucky, creates global stan-dards. These are then localized to fit each distribution center

B R E A K T H R O U G H

SOURCE: MODIFIED FROM DISTRIBUTION CENTER MANAGEMENT 2002, ALEXANDER COMMUNICATIONS GROUP, INC., WWW.DISTRIBUTIONGROUP.COM.

by engineers that work in that facility. Most of The Gap’slabor standards are derived largely by using predeterminedlabor standards using the MOST system developed by H. B.Maynard Co., Ninah says. They are then validated with stop-watch studies. A handful of standards not included in thepredetermined time measures have to be completely devel-oped by the engineering staff.

STANDARDS TO BUILD ONThe standards are a key planning and scheduling tool for TheGap and have proven to be very valuable when setting up newdistribution centers. “It gives us a good gauge of how manypeople we need to hire,” Ninah reports. They also use thelabor standards to track a new facility’s learning curve. “We’reable to see how long it takes a new distribution center to getup to speed,” he says, and for associates to learn their jobs.The standards have proven to be a powerful communicationtool for new facilities. The engineering staff develops laborstandards for a new distribution center before it opens. Themeasures enable associates to “understand what’s expected ofthem, and where they are with respect to where they shouldbe,” Ninah says. “We used to expect a six- to eight-monthlearning curve. Now, it’s about half that.”

Job enrichment Specialized work is made more interesting by giv-ing the worker a greater variety of tasks or by getting a workerinvolved in planning, organization, and inspection.

K E Y T E R M SJob design The function of specifying the work activities of an indi-vidual or group in an organizational setting.

Specialization of labor Simple, repetitive jobs are assigned to eachworker.





F O R M U L A R E V I E WStandard time

[TN5.1] ST = NT(1 + Allowances) Assumes that allowances areadded to normal time.

[TN5.2] ST = NT

1 − AllowancesAssumes that allowances areapplied to the total work period.

S O L V E D P R O B L E M S

SOLVED PROBLEM 1Brandon is very organized and wants to plan his day perfectly. To do this, he has his friend Kellytime his daily activities. Here are the results of her timing Brandon on polishing two pairs of blackshoes using the snapback method of timing. What is the standard time for polishing two pairs?(Assume a 5 percent allowance factor for Brandon to put something mellow on the CD player.Account for noncyclically recurring elements by dividing their observed times by the total numberof cycles observed.)

OBSERVED TIMES

ELEMENT 1 2 3 4 PERFORMANCE RATING NT

Get shoeshine kit 0.50 125%

Polish shoes 0.94 0.85 0.80 0.81 110

Put away kit 0.75 80

Solution�T T

–PERFORMANCE RATING NT

Get shoeshine kit .50 .50/2 = .25 125% .31

Polish shoes (2 pairs) 3.40 3.40/2 = 1.70 110 1.87

Put away kit .75 .75/2 = .375 80 .30

Normal time for one pair 2.48of shoes

Standard time for the pair = 2.48 × 1.05 = 2.60 minutes.


Sociotechnical systems A philosophy that focuses more on theinteraction between technology and the work group. The approachattempts to develop jobs that adjust the production process technol-ogy to the needs of the worker and work group.

Work physiology Considers the physical demands of a job. Work–rest cycles are set according to the energy expended on the job.

Ergonomics Study of the physical arrangement of the work spacetogether with the tools used to perform a task.

Work measurement Job analysis for the purpose of setting timestandards.

Time study Separation of a job into measurable parts, with eachelement timed individually. The individual times are then combined,and allowances are added to calculate a standard time.

Work sampling Analyzing a work activity by observing an activityat random times. Statements about how time is spent during theactivity are made from these observations.

Predetermined motion-time data systems (PMTS) Systems forderiving a time for a job by summing data from tables of genericmovement times developed in the laboratory.

Elemental data Used to derive a job time by summing times from adatabase of similar combinations of movements.

Normal time The time that a normal operator would be expected totake to complete a job without the consideration of allowances.

Standard time Calculated by taking the normal time and addingallowances for personal needs, unavoidable work delays, andworker fatigue.





SOLVED PROBLEM 2A total of 15 observations has been taken on a head baker for a school district. The numerical break-down of the baker’s activities is

MAKE READY DO CLEAN UP IDLE

2 6 3 4

Based on this information, how many work-sampling observations are required to determine howmuch of the baker’s time is spent in “doing”? Assume a 5 percent desired absolute accuracy and95 percent confidence level.

SolutionTo calculate the number of observations, use the formula at the bottom of Exhibit TN5.11 becausethe 95 percent confidence is required (that is, Z ∼= 2).

p = “Doing” = 6/15 = 40%

E = 5% (given)

N = 4p(1 − p)

E2= 4(.4)(1 − .4)

(.05)(.05)= .96

.0025= 384

R E V I E W A N D D I S C U S S I O N Q U E S T I O N S1 Why might practicing managers and industrial engineers be skeptical about job enrichment and

sociotechnical approaches to job design?2 Professors commonly complain to their families that book writing is hard work and that they

should be excused from helping out with the housework so that they can rest. Which exhibit inthis technical note should they never let their families see?

3 Is there an inconsistency when a company requires precise time standards and encourages jobenlargement?

4 Match the following techniques to their most appropriate application:

Worker–machine chart Washing clothes at laundromat

Process chart Tracing your steps in getting a parking permit

Work sampling Faculty office hours kept

5 You have timed your friend, Lefty, assembling widgets. His time averaged 12 minutes for thetwo cycles you timed. He was working very hard, and you believe that none of the nine otheroperators doing the same job can beat his time. Are you ready to put forth this time as the stan-dard for making an order of 5,000 widgets? If not, what else should you do?

6 Comment on the following:a. “Work measurement is old hat. We have automated our office, and now we run every bill

through our computer (after our 25 clerks have typed the data into our computer database).”b. “It’s best that our workers don’t know that they are being time studied. That way, they can’t

complain about us getting in the way when we set time standards.”c. “Once we get everybody on an incentive plan, then we will start our work measurement

program.”d. “Rhythm is fine for dancing, but it has no place on the shop floor.”

7 Organizationwide financial incentive plans cover all the workers. Some units or individuals mayhave contributed more to corporate profits than others. Does this detract from the effectivenessof the incentive plan system? How would your incentive scheme for a small software develop-ment firm compare to an established auto manufacturing firm?

P R O B L E M S1 Use the following form to evaluate a job you have held relative to the five principles of job

design given in the technical note. Develop a numerical score by summing the numbers in paren-theses.






POOR (0) ADEQUATE (1) GOOD (2) OUTSTANDING (3)

Task variety

Skill variety

Feedback

Task identity

Task autonomy

a. Compute the score for your job. Does the score match your subjective feelings about the jobas a whole? Explain.

b. Compare your score with the scores generated by your classmates. Is there one kind of jobthat everybody likes and one kind that everybody dislikes?

2 Examine the process chart in Exhibit TN5.5. Can you recommend some improvements to cutdown on delays and transportation? (Hint: The research laboratory can suggest changes in therequisition form.)

3 A time study was made of an existing job to develop new time standards. A worker was observedfor 45 minutes. During that period, 30 units were produced. The analyst rated the worker as per-forming at a 90 percent performance rate. Allowances in the firm for rest and personal time are12 percent.a. What is the normal time for the task?b. What is the standard time for the task?c. If the worker produced 300 units in an eight-hour day, what would be the day’s pay if the basic

rate was $6 per hour and the premium payment system paid on a 100 percent basis?4 The Bullington Company wants a time standard established on the painting operation of souvenir

horseshoes for the local Pioneer Village. Work sampling is to be used. It is estimated that work-ing time averages 95 percent of total time (working time plus idle time). A co-op student is avail-able to do the work sampling between 8:00 A.M. and 12:00 noon. Sixty working days are to beused for the study. Use Exhibit TN5.11 and an absolute error of 2.5 percent. Use the table of ran-dom numbers (Appendix B) to calculate the sampling schedule for the first day (that is, show thetimes of day that an observation of working/idle should be made). Hint: Start random numberselection with the first tour.

5 The final result of the study in Problem 4 estimated working time at 91.0 percent. In a480-minute shift, the best operator painted 1,000 horseshoes. The student’s performance indexwas estimated to be 115 percent. Total allowances for fatigue, personal time, and so on are10 percent. Calculate the standard time per piece.

6 Suppose you want to set a time standard for the baker making her specialty, square doughnuts. Awork-sampling study of her on “doughnut day” yielded the following results:

Time spent (working and idle) 320 minutes

Number of doughnuts produced 5,000

Working time 280 minutes

Performance rating 125%

Allowances 10%

What is the standard time per doughnut?7 In an attempt to increase productivity and reduce costs, Rho Sigma Corporation is planning to

install an incentive pay plan in its manufacturing plant. In developing standards for one op-eration, time-study analysts observed a worker for 30 minutes. During that time the worker com-pleted 42 parts. The analysts rated the worker as producing at 130 percent. The base wage rateof the worker is $5 per hour. The firm has established 15 percent as a fatigue and personal timeallowance.a. What is the normal time for the task?b. What is the standard time for the task?c. If the worker produced 500 units during an eight-hour day, what wages would the worker have

earned?8 Because new regulations will greatly change the products and services offered by savings and

loan associations, time studies must be performed on tellers and other personnel to determine thenumber and types of personnel needed and incentive wage payment plans that might be installed.As an example of the studies that the various tasks will undergo, consider the following problemand come up with appropriate answers.






A hypothetical case was set up in which the teller (to be retitled later as an account advisor)was required to examine a customer’s portfolio and determine whether it was more beneficial forthe customer to consolidate various CDs into a single issue currently offered, or to leave the port-folio unaltered. A time study made of the teller yielded the following findings:

Time of study 90 minutes

Number of portfolios examined 10 portfolios

Performance rating 130 percent

Rest for personal time 15 percent

Teller’s proposed new pay rate $12 per hour

a. What is the normal time for the teller to do a portfolio analysis for the CDs?b. What is the standard time for the analysis?c. If the S&L decides to pay the new tellers on a 100 percent premium payment plan, how much

would a teller earn for a day in which he or she analyzed 50 customer portfolios?9 Based on a manager’s observations, a milling machine appears to be idle approximately 30 per-

cent of the time. Develop a work-sampling plan to determine the percentage of idle time, accu-rate within a 3 percent error (±3%) with a 95 percent confidence level. Use the random numbersfrom Appendix B to derive the first day’s sampling schedule (assume that the sample will takeplace over 60 days and that an eight-hour shift is used from 8:00 to 12:00 and 1:00 to 5:00).


C A S E : J E A N S T H E R A P Y — L E V I ’ S F A C T O R Y W O R K E R S A R E

A S S I G N E D T O T E A M S , A N D M O R A L E T A K E S A H I T

In an industry notorious for low wages and lousyworking conditions, Levi’s has prided itself on being a grand excep-tion. It offered generous pay plus plenty of charity support in factorytowns—all financed by the phenomenal profitability of its brilliantlymarketed brand name. It clung to a large U.S. manufacturing baselong after other apparel firms began moving offshore, and it oftenwas ranked among the best companies to work for.

But to many of Levi’s workers, the company’s image has notfit for some time. In 1992 the company directed its U.S. plants toabandon the old piecework system, under which a worker repeat-edly performed a single, specialized task (like sewing zippers orattaching belt loops) and was paid according to the amount ofwork he or she completed. In the new system, groups of 10 to 35workers would share the tasks and be paid according to the totalnumber of trousers the group completed. Levi’s figured that thiswould cut down on the monotony of the old system and enablestitchers to do different tasks, thus reducing repetitive-stressinjuries.

At the time, the team concept was a much-touted movementdesigned to empower factory workers in many industries, andLevi’s unions agreed to the effort. But there was more to it than thatfor Levi’s. Faced with low-cost competitors manufacturing over-seas, the San Francisco–based company did not feel it could keepmany of its U.S. plants open unless it could raise productivity andreduce costs, particularly those incurred by injured workers pushingto make piecework goals. Teamwork, Levi’s felt, would be morehumane, safe, and profitable.

Instead, the new system led to a quagmire in which skilledworkers found themselves pitted against slower colleagues, damag-ing morale and triggering corrosive infighting. Many top perform-ers said the first thing they noticed about teams was that their payshrank—and some of them decided to throttle back. They feltcheated because they were making less. Whenever a team member

was absent, inexperienced, or just slow, the rest of the team had tomake up for it. That infuriated some team members who felt theywere carrying subpar workers. With limited supervision fromcoaches, groups were forced to resolve most workflow and person-ality issues themselves.

The fundamental problem arises from the nature of work atLevi’s factories. Unlike an assembly line for cars or copiers, speedin garment-making relates directly to a worker’s skill and staminafor grueling, repetitive motions of joining and stitching fabric. Theworkers in Levi’s plants operate machines that perform specifictasks: pocket setter, belt looper, and fly stitcher, among others.Some employees work much faster than others.

In 1993 Levi’s hired a consulting firm to analyze the problems.Its conclusion was simply that the company should start fromscratch and involve all parties in a redesign of pay structures andwork processes. As they began discussing the changes, some plantmanagers complained that the sessions were “at times too touchy-feely and not business-based enough.” Some managers just did notlike the idea of having sewing machine operators challenge theirauthority. Costs mounted, and in April 1994 plant managers werewarned that they must cut costs by 28 percent on average by the endof 1997 or face an uncertain future.

By early 1997, Levi’s share of the domestic men’s denim jeansmarket fell to 26 percent from a high of 48 percent in 1990.Burdened by new debt, Levi’s in February 1997 announced plans tocut its salaried workforce by 20 percent over 12 months. Later inNovember 1997, the firm announced the closing of 11 U.S. plantsand layoffs of 6,395 workers. The company said that none of thesejobs were transferred overseas. Still, over the years the companyshifted much of its work abroad. Industrywide in 1991, approxi-mately 15 percent of the jeans for the U.S. market were manufac-tured abroad. Approximately 45 percent of the jeans were producedin foreign plants by the end of 1997.






S E L E C T E D B I B L I O G R A P H YAft, Lawrence S. Work Measurement and Methods Improvement

(Engineering Design and Automation). New York: Wiley-Interscience, 2000.

Meyers, F. E., and J. R. Stewart. Time and Motion Study: For LeanManufacturing. 3rd ed. Upper Saddle River, NJ: Prentice Hall,2001.

Niebel, B. W., and A. Freivalds. Methods, Standards, and Work Design.New York: WCB/McGraw-Hill, 1998.

Parker, S., and T. Hall. Job and Work Design: Organizing Work toPromote Well-Being and Effectiveness. Thousand Oaks, CA: Sage,1998.

Ramsey, G. F., Jr. “Using Self-Administered Work Sampling in a StateAgency.” Industrial Engineering, February 1993, pp. 44–45.

Rutter, R. “Work Sampling: As a Win/Win Management Tool.” IndustrialEngineering, February 1994, pp. 30–31.

Sasser, W. E., and W. E. Fulmer. “Creating Personalized Service DeliverySystems.” In Service Management Effectiveness, ed. D. Bowen,R. Chase, and T. Cummings. San Francisco: Jossey-Bass, 1990,pp. 213–33.

F O O T N O T E S1 M. Apgar IV, “The Alternative Workplace: Changing Where and How People Work,” Harvard Business Review 76, no. 3

(May–June 1998), pp. 121–36.

2 E. E. Lawler III, The Ultimate Advantage: Creating the High Involvement Organizations (San Francisco: Jossey-Bass, 1992),pp. 85–86.

3 Ibid., pp. 98–99.

4 E. E. Lawler III, “Paying for Organizational Performance,” Report G87-1(92) (Los Angeles: Center for Effective Organizations,University of Southern California, 1987).

5 C. Giffi, A. Roth, and G. M. Seal, Competing in World-Class Manufacturing: America’s 21st Century Challenge (Homewood, IL:Richard D. Irwin, 1990), p. 299.

Levi’s says the team approach was the company’s attempt toensure long-term survival for as many U.S. plants as possible.Plant closures might have come sooner, and job losses might havebeen heavier, had teams never been adopted, company officials say.Levi’s vows to persevere with the team strategy at its remainingU.S. plants. But unofficially, much of the approach is beingscrapped as individual managers seek ways to improve productivi-ty. People in the plants are gradually going back to the old way ofdoing things.

Q U E S T I O N S1 What went wrong with Levi’s move to teams in their plants?2 What could Levi’s have done differently to avert the

problems?3 Devise a team incentive plan that you think might work.4 Do you think the need to move jeans production offshore

was inevitable? Could Levi’s have done anything to avert theproblem of increasing labor costs?

SOURCE: R. T. KING JR., “JEANS THERAPY,” THE WALL STREET JOURNAL, MAY 20, 1998, P. A1. © 1998 DOW JONES AND COMPANY, INC. ALL RIGHTS RESERVED WORLDWIDE. USED WITH PERMISSION.

Operations Management for Competitive Advantage 11e Ch05

Documents

Transcript of Operations Management for Competitive Advantage 11e Ch05