COURSE MANUAL

C4: Operations Management

Module 2

Open University of Mauritius

Reduit
MAURITIUS
Copyright
© Commonwealth of Learning, 2011

All rights reserved. No part of this course may be reproduced in any form by any means
without prior permission in writing from:

Commonwealth of Learning
1055 West Hastings Street
Suite 1200
Vancouver, BC V6E 2E9
CANADA

Email: info@[Link]

OMAURITIUSReduit
MAURITIUS
Fax: +230 464 8854
Tel: +230 403 8200
E-mail: academic-affairs@[Link]
Website: [Link]
March 2013
Acknowledgements
The Commonwealth of Learning (COL) wishes to thank those below for their contribution to
the development of this course:

Course author David Gardiner

Gardiner Consulting Group
Operations Management Consultants
Christchurch, New Zealand

Course content specialist: Olga Kaminer, PhD

Professor, Supply Chain/Logistics/Operations Management
Faculty of Business
Sheridan Institute of Technology and Advanced Learning,
Canada

Subject matter experts: Fook Suan Chong

Wawasan Open University, Malaysia

Maurice Fletcher
University College of the Caribbean, Jamaica

Araba Intsiful
Kwame Nkrumah University of Science and Technology,
Ghana

S. A. D. Senanayake
Open University of Sri Lanka, Sri Lanka

Educational designers: Symbiont Ltd.

Paraparaumu, New Zealand

Course editor: Symbiont Ltd.

Paraparaumu, New Zealand

Course formatting: Kathryn Romanow

Commonwealth of Learning

COL would also like to thank the many other people who have contributed to the writing of
this course.
 Contents

Contents
Module 2 1 
Balancing supply with demand ......................................................................................... 1 
Unit 3 2 
Demand management and forecasting ............................................................................... 2 
Activity 2.1 ...................................................................................................................... 10 
Activity 2.2 ...................................................................................................................... 11 
Activity 2.3 ...................................................................................................................... 13 
Activity 2.4 ...................................................................................................................... 15 
Activity 2.5 ...................................................................................................................... 15 
Activity 2.6 ...................................................................................................................... 16 
Activity 2.7 ...................................................................................................................... 17 
Activity 2.8 ...................................................................................................................... 18 
Activity 2.9 ...................................................................................................................... 20 
Activity 2.10 .................................................................................................................... 21 
Unit summary 22 
Readings for further study 23 
Unit 4 24 
Capacity planning and management ................................................................................ 24 
Activity 2.11 .................................................................................................................... 35 
Activity 2.12 .................................................................................................................... 41 
Activity 2.13 .................................................................................................................... 42 

Unit summary 42 

Assignment ...................................................................................................................... 43 
References 45 
Further Readings 45 
Activity feedback............................................................................................................. 46 
 C4: Operations Management

Module 2
Balancing supply with demand
Introduction
This module examines at a strategic level, the balancing of demand
with supply. An economist would argue that prices adjust to
balance supply with demand. This is great in theory, but the
operations manager gains no comfort from this statement. Excess
demand means lost revenue and excess supply means wasted
resources. The balancing of supply with demand in the real world is
very difficult. If organisations can (somehow) get it right, then they
should be very effective and very successful.
Upon completion of this module you will be able to:
• Explain the nature of demand.
• Understand the strategic role of forecasting.
• Distinguish between qualitative and quantitative forecasting
and perform basic quantitative calculations.
Outcomes • Outline how capacity is measured and appreciate the
dilemma faced by management in matching variable
demand with variable capacity.
• Calculate various aggregate planning scenarios.
• Identify various strategies for balancing supply with
demand.
• Evaluate the application of yield management.
• Evaluate queues and waiting lines.

1
Unit 3

Unit 3
Demaand man nagemennt and
forecaasting
Introdu
uction
Operattions managers spend money
m on the inputs to thee
transfoormation proocess. These inputs may be raw mateerials, humann
resources, buildinggs, machiness, processes, energy and operating
suppliees, to name a few. A majajor decision is what to buy,
b but
equallyy important decisions innclude when to buy and howh much too
buy. To
T answer theese questionns we need too know, withh some levell
of accuuracy, how m much outputt from the traansformationn process wee
will neeed. That is w
where forecaasting and deemand manaagement enteer
the piccture.
Operattions management persoonnel use forrecasts to maake decisionss
about process
p selection, capacity planningg, facility layyout,
producction planninng, schedulinng and invenntory. Forecaasting is
essentiial in operatiions manageement.
One thhing we can bbe reasonablly certain abbout is that any
a forecast iis
most liikely to be ddifferent from
m what evenntually happeens. Many
managgers are discoouraged from m making foorecasts simpply because
they knnow they wiill be “wrongg”. The essence is to som mehow agreee
on a seet of numberrs and plan around
a thosee numbers annd to developp
a contiingency plann in case the real numberrs are too higgh or too low
w.
This unit starts by defining dem mand managgement and distinguishin
d ng
betweeen short-rangge, medium--range and loong-range foorecasting.
This leeads to a disccussion on thhe strategic role of foreccasting and
the diffferences bettween depenndent and inddependent deemand.
We peerform some forecasting deviation caalculations and a interpret
the ansswers. This iis followed by
b some more calculatioons in
quantittative forecaasting. The inntention is too show the ty
ypes of
calculaations that caan be performed and theerefore the mathematics
m is
quite basic.
b
We deecompose a timet series innto componeents so we caan understannd
where the data is coming
c from
m and this plaaces us in a better
b
positioon to determine future daata. One of tthe demand components
c
is seassonality, so we
w develop the
t seasonaliity index andd perform
detaileed calculations using seaasonal indicees and regresssion analysiis.
We coonclude with a discussionn on alternattive approachhes to
forecasting.

2
 C4: Operations Management

Upon completion of this unit you will be able to:

• Define demand management.
• Explain the nature of demand.
• Understand the strategic role of forecasting.
• Distinguish between qualitative and quantitative
Outcomes forecasting.
• Explain forecast accuracy.
• Define forecast value added.
• Perform basic quantitative calculations on forecasting.
• Define and calculate seasonal indices.
• Use regression analysis to develop long-term trends.
• Discuss other approaches to forecasting.

Cyclical A component of demand that occurs in a cycle

component greater than one year. Examples of cycles
include economic, political and/or business.
Terminology
Decomposition A method of forecasting where the time series
data is split into components of demand
(trend, seasonal and cyclical). Trend measures
the general upwards or downwards direction,
seasonal indicates the effect of different
seasons and cyclical shows the effect of a
longer repeating non-seasonal period cycle.
Each separate component is projected into the
future and the sum of the projections becomes
the new forecast.

Delphi method A qualitative forecasting method that uses a

group of experts to arrive at a consensus about
the future.

Demand A need for a particular product or service. At

the finished-goods level demand may not
equate to sales. Demand is what the customer
wants, while sales represent the ability to
deliver. In demand forecasting there are four
components of demand:trend, seasonal,
cyclical and random.

Dependent The demand for all lower-level items

demand calculated from the product structure of the
end item. Dependent demand should be
calculated and not forecast.

3
Unit 3

Forecaast An estimaate of future demand. A forecast cann

be construucted using quantitative
q methods,
qualitative methods or a combinattion of both..
The four components
c of demand are:
a trend,
seasonal, cyclical andd random.

Forecaast The differrence betweeen the actuall (eventual)

deviattion demand and
a the foreccast value.

Indepeendent The demaand for an iteem that is unnrelated to thhe

demannd demand for
fo other itemms. Examples include
demand for
fo finished items and spaare parts.

Randoom Random component

c iin demand foorecasting haas
compoonent no predicttable patternn. For exampple, sales datta
may vary around a forrecast value with no
specific pattern
p formiing and no way
w of more
accurately
y determininng the actual demand
other than
n by the foreecast.

Season
nal A compon nent of demaand describin ng the
compoonent variation that
t occurs bbecause of thhe time of
year, monnth or week. A seasonal component
generally repeats itsellf at least onnce a year
whereas a cyclical com mponent usu ually takes
longer thaan one year.

Season
nal index A numberr used to adjust data to seasonal
demand.

Trend
d A compon nent of demaand showingg an increasee
compoonent or decreasse of demandd over time.
Terminoloogy sourcedd from Gardinner (2010).

4
 C4: Operations Management

What is demand management?

Demand is the need for a particular product or service. When
customers need a product or service they approach their supplier
and demand sufficient quantity of that product or service to satisfy
their demand.
Without a need to satisfy demand, firms and organisations have
little reason to exist. Production firms make tangible products that
can be consumed by customers and service firms deliver services
that are experienced by customers. It is up to the organisation to
decide how much of that demand they will deliver.
It is a strategic decision for organisations to decide how much of
that demand they want to supply. When a production firm stores
products in a warehouse, they are anticipating future demand.
When a service organisation occupies a facility and employs and
trains staff, they are anticipating future demand.
Demand occurs at all stages in a supply chain and at all stages in a
service chain. Raw materials are demanded by the manufacturer,
fabricated and component parts are demanded by the assembler,
finished products are demanded by the wholesaler/distributor,
finished products are also demanded by the retail customer, and
services are demanded at all stages of the supply chain as well as
the service chain.
Predicting the future is an art as well as a science. It does not really
matter how specific firms and organisations arrive at their forecast.
What does matter, however, is exactly what the organisation does
with the forecast. The predictions are useful for organisations
before they make decisions which may give them a competitive
advantage in the market place.
A forecast is an estimate of future demand and it can be developed
using quantitative methods, qualitative methods or a combination
of both. The forecasting process is the business process that
attempts to predict demand for products and services so that
capacity, resources and materials are available in time to meet the
need.
It will cost money to buy capacity; it will cost money to occupy a
facility; it will cost money to buy raw materials and components; it
will cost money to train and educate employees to perform
production and service activities; and it will cost money to deliver
products and services to customers.
Depending on how much money is spent and when it is spent will
determine how well, or how poorly, the organisation is able to
satisfy customer demand. If the organisation spends the right
amount of money, the organisation will satisfy customer demand at

5
Unit 3

the minimum cost. If the organnisation spennds too muchh, the

organiisation may find
f it has unnder-utilisedd resources.

Short--, medium
m- and longg-range foorecastingg
Demannd forecastinng operates within
w three time horizonns, short-,
medium
m- and longg-range.
Long--range foreecasting
Long-rrange forecaasting generaally covers thhe period req quired to
replacee major resoources. Thesee forecasts aare strategic in
i nature andd
cover applications
a such as faciility and cappacity planninng,
technoology and deesign planninng, research and developpment, and
processs planning.
Forecaasting is aggrregated into total groupss of similar products
p andd
shouldd indicate treends and resoource imbalaances. So a steel
s produceer
might forecast tonnnes of outpuut per year, aan airline migght forecast
the nummber of passsengers and tonnes of freeight per rouute per montth
n electricity pproducer migght forecast megawatt hours (MW.h
and an
= 106 watt
w hours) oof electricity y per month.
Forecaast data for loong-range foorecasting shhould be braacketed in
monthhs, quarters or
o even yearss. If you werre planning the
t electricityy
generaation requirements for a significant rregion you might
m examinne
closelyy the peak seeasons for th
he next 10 orr 20 years. Daily
D
requireements woulld not be impportant, but possible daily peak loadds
would be importannt. The actuaal day that thhe peak load will occur
ot be consideered, but thee year that thhe new peak level is
will no
achievved would bee important.
The foorecasts at thhis level need
d to be in thee “ballpark”.. In other
words,, absolute acccuracy is noot relevant ass long as it iss relatively
close and
a clearly pprovides the trend data annd resource requirementts
that wiill be requireed.
Mediu
um-range foorecasting
Mediuum-range forrecasting appplications incclude sales planning,
p
operatiions planninng, budgetingg, yield mannagement andd aggregate
staffin
ng plans.
Mediuum-range forrecasting moodels use som me aggregatiion. A
hospitaal, might grooup all patien
nts into the type
t of care they are to
receivee, for exampple: maternityy, paediatricc, psychiatricc, elective
surgeryy or accidennt and emerg gency.
The level of detail is greater th
han with longg-range foreecasting and
the reqquired accuraacy is also greater.
g An oorganisation planning
p in
this tim
me frame usuually will noot have the capability of acquiring
additioonal resourcees in time to
o meet any siignificant chhanges in
demannd. This shouuld have beeen highlighteed in the long g-term planss.

6
 C4: Operations Management

Peak periods such as holidays, special events and changes in shift

capacity will, however, need to be considered.
Short-range forecasting
Short-range forecasting usually involves detailed planning and
scheduling for purchasing, job scheduling and staffing rosters as
well as production allocations.
Planners and operators working with short-range data will be
endeavouring to resolve short-term demand and capacity issues.
These issues will be current and have to be resolved immediately.
Planning is performed by named personnel for individual tasks at
specific times on given days. This can become very detailed and
this level of detail is required to make the tactical decisions of the
organisation.

Strategic role of forecasting

Forecasting demand for some products is more difficult than
forecasting the demand for others. New products do not have
established patterns of demand; fashion items are influenced by
intangible forces such as perceived desirability; and commodity
products are sold on a regular basis. This, however, does not make
forecasting any less worthwhile.
One thing certain about forecasting is that it will almost always be
wrong. It may deviate from actual performance by only a small
margin, but it will be different. This is enough to put some
organisations off the concept of forecasting. They wonder why they
should put valuable resources into an activity that is always wrong.
In a strategic sense, each organisation should understand the
dynamics of the demand patterns and be prepared for whatever
happens. Sudden demand changes may occur because of external
factors such as natural disasters, terrorist activities, weather
patterns, major events and accidents. Gradual changes may occur
as a result of social, demographic, political, legal and
environmental change.
Some of these changes can be expected and some are unexpected.
One could argue that the unexpected should, in fact, be expected
and that it is the magnitude and timing of the occurrence that is in
doubt. Organisations should be prepared for whatever happens to
their demand patterns. This does not mean that they should be able
to react satisfactorily to all changes in demand. They may, for
instance, elect as a matter of policy not to service demand. They
may have in their planning a statement suggesting that should
demand exceed a certain value then they will choose to reject it and
not attempt to satisfy it. This is not a statement that says that all

7
Unit 3

demannd must be saatisfied. Rath her, it is a sttatement thatt says that thhe
firm caan handle ceertain deviatiions from exxpected demand but onlyy
up to a limit, at whhich point th
he excess is uunsatisfied.
Organisations shouuld considerr the consequuences of thee forecast
being too
t optimistic or too pesssimistic. Whhat would haappen to the
busineess if the foreecast figuress were not acchieved? Or the forecast
figuress were exceeeded? What are the choicces of actionn?

Reflecction 1
Excludding the channging behavviours demonnstrated by thhe customer,
what factors
fa wouldd influence demand?
d

Reeflection

Activitty feedback can be foundd at the end of this moduule.

Depenndent and independdent demaand

Depenndent demandd is the dem mand for all loower-level ittems
calculaated from the product strructure of thhe end item. Dependent
D
demannd should be calculated anda not foreccast. Indepenndent demannd
is the demand
d for aan item that is unrelatedd to the demaand for otherr
items. Examples innclude demaand for finishhed items an nd spare partts.
Deman nd forecastinng processess should be applied
a to inddependent
demannd items and groups of in ndependent demand
d item
ms. Most
produccts and serviices sold exhhibit indepenndent demannd as these caan
be soldd independenntly of all otther sales. Occasionally, an item is
sold inn associationn with anotheer item. For example, bread and
butter,, socks and shoes,
s motorr vehicles and petrol, fishh and chips,
and paaint and a paint brush. Thhe demand ffor these typees of items iis
still coonsidered to be independdent since thee quantity off each is nott
directlly related to the other item in the pairing.
Depenndent demandd relates to thet componeent items of a product
structuure where thee componennt items are inn relatively fixed
quantitties of the paarent item. For
F example,, the meat annd vegetablees
of a reestaurant meaal, the purchhased and fabbricated item
ms of a
refrigeerator and the quantity off steel and concrete in a building.
Depenndent demandd should be calculated bby multiplyinng the
quantitty per parentt by the nummber of parennt items beinng demandedd.

Forecaast deviattion calcu

ulations
Forecaast deviationn is the differrence betweeen the actuall (eventual)
demannd and the foorecast valuee. Organisations use the output
o of
forecasting for speending decisiions. It mustt be obvious that if the
actualddemand diffefers from the forecast vallues then speending will

8
 C4: Operations Management

not be at the optimum level. A forecast deviation would suggest the

firm is about to either spend too much and have excess products
and capacity available, or not spend enough and be unable to
satisfy demand for its products and services.
Most organisations plan their supply chains assuming forecasts will
actually happen. They start with this assumption and express regret,
or blame market conditions, when they end up with product
shortages or surpluses. Instead, they should start with a range that
represents the likely upper and lower bounds and plan the demand
and supply risk of those limits.
While it is not necessary (or even practical in all cases) for the
forecast to be 100 per cent accurate, it would benefit the whole
forecasting process if the reasons for the variation were understood
and if a learning process took place. This allows the forecasting
process to improve and, in turn, allows the organisation to aim for
the optimum expenditure of resources.

Reflection 2
Think of three or four reasons why a forecast value would be
inaccurate.

Reflection

Activity feedback can be found at the end of this module.

There are two types of forecast deviation, bias and random. Bias
deviations occur when a consistent mistake such as always too high
or always too low is made. Random deviations simply cannot be
explained; they just happen and are sometimes referred to as
“noise”.
We can measure forecast deviation in various ways. Most
measurements examine the difference between the actual demand
and the forecast value. Sometimes we look at the algebraic
difference which allows for high values to cancel out low values
with a net summation of close to zero. It is possible to have wildly
fluctuating forecast values and still conclude that a good forecast
model is being used.
An absolute difference allows the magnitude of the over- or under-
forecast to be measured. Whether the forecast value is over or
under does not matter; the absolute measurement examines the
relative distance of the actual demand from the forecast value.
The algebraic deviation and the absolute deviation are very popular
measurements but the significance of the deviation relative to the
actual observation quantity is required. A deviation of 50 when

9
Unit 3

forecasting demannd in hundredds is significcantly differeent from a

deviatiion of 50 whhen forecastiing demand in millions.
Mean absolute deviation
d (M
MAD)
The mean
m absolutee deviation (MAD)
( meassures the abssolute
disperssion of the deviation.
d It is
i calculatedd as the meann of the sum
m
of the absolute difffferences betw ween the acttual demandd values and
the forrecast valuess. An absolutte value doees not have anya sign. If thhe
differeence is negattive, it is wriitten just as a number; siimilarly, if itt
is posiitive it is wriitten just as a number. Thhe mean abssolute
deviatiion measures the averagge distance of demand vaalues from
forecast values. Thhe formula for f mean absolute deviatiion (MAD)
is:
mean absoolute deviatioon

MAD = ∑ D −F
n
where D iss the actual demand
d valuue for each period
p
F is
i the forecasst value for each
e period
n is the numbeer of periods or observatiions

Activiity 2.1
Use the data in thee following table to calcuulate mean absolute
a
deviatiion MAD.

A
Activity Mo
onth
De
emand       Fo
orecast        Deviation     
D A
Abs deviation 
D F (D‐F ) |D‐F |
Ja
an 500 550 ‐50 50
Fe
eb 550 600 ‐50 50
M
Mar 420 490 ‐70 70
A
Apr 500 530 ‐30 30
M
May 610 530 80 80
Ju
un 600 550 50 50
J ul 680 610 70 70
A
Aug 670 670 0 0
Se
ep 720 690 30 30
O
Oct 750 730 20 20
Su
um 6
6000 5950 50 450
A
Ave 600 595 5 45

Activitty feedback can be foundd at the end of this moduule.

10
where Dis the actual
F forecast
demand
for each
forperiod
each period

F
A
M n
Ja
e
b
r
y
p 2
4
5
0
1
6 6
9
4
0
3
5 5
7
‐3
0
8 5
7
3
0
8

C4: Operations Management

Bias
Bias indicates whether a method of forecasting tends to favour a
higher or lower value. It is calculated as the sum of the algebraic
differences between the actual demandvalues and the forecast
values divided by the sum of the demand values. Thus, the pluses
may offset the minuses. It is a useful measure especially when
expressed as a percentage of actual demand.
The formula for bias is:

Activity 2.2
Use the data in the following table to calculate bias.

Activity

Activity feedback can be found at the end of this module.

Mean absolute percentage deviation (MAPD)

Mean absolute percentage deviation (MAPD) is the mean of the
absolute deviation between actual demand value and forecast value
divided by the mean of the demand values expressed as a
percentage. This is similar to mean absolute deviation (MAD) but
considers the significance by dividing by mean demand.

11
Unit 3

The MAPD
M is desccribed as a mean
m value but
b the formuula does not
divide by the numbber of observ vations. This is simply because
b n, thhe
numbeer of observaations, is thee divisor for both
b numeraator and
denomminator and cancels
c itselff out. The formula couldd be written asa
mean absolute
a devviation divided by mean demand.
The fo
ormula for m
mean absolutee percentagee deviation MAPD
M is:

mean absollute percentaage deviationn

MAPD = ∑
D-F x100
0
∑D
where D iss the actual demand
d valu
ue for each period
F is
i the forecast value for each period

Mean absolute percentage

p variation (MAPV)
Mean absolute perrcentage variiation (MAP PV) is the avverage of the
absoluute deviationn between acttual demandd value and meanm demannd
value divided
d by thhe mean dem mand expressed as a perccentage. Thee
MAPV V is describeed as a meann value, but thhe formula ini total does
not div
vide by n, thhe number off observationns. This is sim mply becausse
n, the number
n of oobservations,, is the divisoor for both numerator
n annd
denomminator and cancels
c itselff out. MAPV V measures thet variabilitty
in the demand.
The fo
ormula for m
mean absolutee percentagee variation (M
MAPV) is:
mean absolute percentage variatiion

∑ D −∑n x100
D

MAPV =
∑D
where D is the actual demand vaalue for each
h period
n is the number of periodss or observattions

Mean absolute perrcentage devviation (MAP PV), on its own,

o is not
strictlyy a forecastinng measure. The usefulnness of MAP PV comes
when comparing
c thhe forecast deviation
d witth the volatility of the
actual demand. A dynamic dem mand patternn is considerrably more
difficuult to forecasst than a stab
ble commodiity demand pattern.
p

12
a
J
e
F 0
5
n
b 0
5
6 0
5

C4: Operations Management

Activity 2.3
Use the data in the following table to calculate mean absolute
percentage deviation MAPD and mean absolute percentage
variation MAPV.
Activity

Activity feedback can be found at the end of this module.

Forecast value added

Forecast value added is the change in forecast accuracy due to an
activity in the forecasting process. The forecasting process starts
with demand history. This is usually quite difficult to obtain since
most organisations have sales history but not demand history. The
difference is in the detail of quantities, products and timings. The
customer demands products in certain quantities and certain times,
and the organisation supplies quantities of products at certain times.
The demand may not equal the supply, but it is the supply that is
measured. Supply is the firm’s ability to deliver.
The quantitative forecast is now examined by various functions and
individuals in the organisation for their qualitative inputs. These
functions and individuals may represent sales, marketing, product
development, finance, production and anyone else who can make a
meaningful contribution to the forecast output. Additionally,
executive management would almost certainly want to have some
inputs.
This is where the forecast value added analysis becomes effective.
All the inputs supplied by these functions and individuals are noted
and a consensus forecast is developed by a consensus process.
Power, influence, hierarchy, knowledge, lack of knowledge,
history, experience and politics all play their part in developing the

13
Unit 3

consennsus forecastt. Often, in power

p politiccs, it is the chief
executtive that has the strongesst voice, sincce few emplooyees
reportiing below thhe chief execcutive actually stand up and a argue
strongly against thhe chief execcutive. Yet, iit could concceivably be
that the inputs from
m the chief executive
e aree making thee forecast
worse..
When actual obserrvations are available theey are compared to the
consennsus forecastt. Ideally, the consensus forecast shoould mirror
the acttual observattions. If it dooes not, and this outcom
me would
normaally be expeccted, then eacch of the inpputs to the coonsensus
processs is examineed to see the effect each input had inn modifying
the staatistical foreccast. In otherr words, each modificatiion to the
quantittative forecaast is noted and
a analysedd to see if it added
a value
to the forecast or made
m it worsse.
If, for example, thee marketing function is aable to exertt pressure
duringg the consenssus process and
a this resuults in a posittive
contribbution to thee consensus, then marketting will assuume a
stronger position inn the next coonsensus rouund. If a function, or an
individdual, contribbuted to the consensus
c annd made it worse,
w their
inputs will be discounted in fuuture and maay even be ig gnored.

Quanttitative forrecasting
Quantiitative forecaasting assum
mes the histoory of past daata about thee
item being forecasst can, in somme way, be used
u to prediict the futuree.
Quantiitative technniques includde time series models succh as simplee
movinng average, weighted
w mooving average, exponentiial smoothinng,
decom
mposition of a time seriess and regresssion analysiss.
When using past data d one assuumes the dem mand patternn will be
repeateed. If it is knnown that the demand paatterns will not
n be
repeateed, then the historical daata must be cchanged. Thiis may occurr,
for exaample, whenn a supplier is i unable to ddeliver raw materials
m
resultinng in no salees of finished products. In
I this exam mple, demandd
is occu
urring but thhe sales are not
n and mostt firms recordd sales (and
not demmand). Otheer examples of when passt data becom mes unreliable
include changes inn legislation and known changes
c in thhe
environment.
Firms often strugggle with this concept of changing
c thee actual data
to reacch a better foorecast. Theyy argue that the past dataa is history
and thee forecast shhould be based on historyy. The analyyst response
shouldd be that it dooes not matter how the fforecast is deeveloped.
What does
d matter is that the fo
orecasting prrocess is impproved.
Simple moving avverage
The simmple moving average foorecast is a fo
forecasting method
m that
adds toogether the most
m recent actual
a observvations and divides by
the nummber of obseervations.

14
t=
tF ......
t−n
+ D t−2chosen number
D
n
+ ∑
t−1
D
n
= of periods
D

where Fis the demand

D forecast value for each period

tis the chosen

n period number
number of periods

C4: Operations Management

The formula to calculate the forecast using simple moving average

is:

Using simple moving average, the forecast for a period is the

average of the actual demand for the n most recent periods. The
choice for the number of periods, n, is arbitrary and is usually an
odd number of periods such as 3, 5, 7 or 9. Greater numbers of
periods makes the forecast less responsive as this tends to smooth
out any temporary ups and downs in the demand pattern.

Activity 2.4
Calculate the forecast for month six using a three-month simple
moving average given historical demand for months one to five
as follows: 120, 130, 110, 135 and 145.
Activity

Activity feedback can be found at the end of this module.

Activity 2.5
Calculate the forecast for month six using a five-month simple
moving average given historical demand for months one to five
as follows: 120, 130, 110, 13, and 145.
Activity

Activity feedback can be found at the end of this module.

15
Unit 3

Weighhted movinng average

The weighted movving averagee is an averagging techniq que that
assigns varying weeights accordding to theirr significancee to selectedd
data vaalues.
The fo
ormula for w
weighted movving averagee is:
sum of (each h period's deemand valueex each perio
od's weight)
Ft =
summ of the weig ghts
D w + ..... + Dt −2 wt −22 + D t −1wt −1
Ft = t −n t −n
wt −n + ..... + wt −2 + wt −1
wherre F is the foorecast
D is the demand
d value for each peeriod
w is the weight
w applied
d to each period
t is the peeriod numbeer
n is the ch
hosen numbeer of periodss

A weigghted movinng average alllows any weighting (or influence) to t

be placced on the demand for each of the n most recentt periods. Thhis
method recognisess that all periiods should not be treateed equally.
Assumme, for exam mple, that the forecast covvers nine moonths. It might
be unrreasonable too allow the same emphassis (weight) to be appliedd
to the first period (most
( distannt) as is used with the lasst period
(most recent).
The sim mple moving average treeats all perioods equally and
a applies an a
even weighting
w to each period. Weighted moving
m averrage does noot
treat alll periods eqqually. The sum
s of all weeights usuallly adds up too
one.

Activiity 2.6
Calculaate the forecaast for monthh six using a three-montth weighted
moving g average givven historicaal demand foor months on ne to five
as follo
ows: 120, 130, 110, 135 and 145. Appply weights of 0.2, 0.3
and 0.55. (In other words,
w apply weights of 220 per cent, 30 per cent
A
Activity and 50 per cent.)

Activityy feedback ccan be foundd at the end oof this moduule.

16
where Fis the demand
D forecast value for ea

tαisisthe
thesmoothing α
0<
period number
constant

C4: Operations Management

Exponential smoothing
Exponential smoothing is a time series forecast that adjusts a
previous forecast by a percentage of the forecast deviation. The
method is called exponential because data points are weighted in
accordance with an exponential function of their age.
The formula for exponential smoothing is:

For this method, only three pieces of data are required — namely
the most recent forecast, the actual demand that occurred for that
period, and a smoothing constant, alpha (α). Alpha is given a value
between 0 and 1. This method gives the weight of α to the demand
of the most recent period; α(1- α) to the demand one time period
2
older; α(1- α) to the demand two time periods older; and so on.
Thus the method applies exponential weighting such that each
increment in the past is decreased by (1- α).
For stable demand, a small alpha is desirable for lessening the
effects of random changes. For increasing or decreasing demand, a
large alpha is desirable. Adaptive smoothing refers to approaches
for controlling the value of alpha.

Activity 2.7
Calculate the forecast for month six using exponential smoothing
given historical demand for months one to five as follows: 120,
130, 110, 135 and 145. Use an alpha factor α equal to 0.2 and you
are given a forecast for month five equal to 130.
Activity

Activity feedback can be found at the end of this module.

Regression analysis
Regression analysis enables us to determine the relationship
between a variable of interest, called a dependent variable, and one
or more independent variables. The equation that describes a
straight line, or linear function, takes the form ŷ = a + bx.

17
Unit 3

yˆ is the bestt estimate off y = a + bx

n∑ xy − (∑ x∑ y)
y
b=
n∑ x 2 − (∑ x)
2

a = ∑ −b∑
y x
n n
where y is the
t dependennt variable
x is the
t independdent variablee
a is a constant where
w the linee on the grapph cuts the y - axis (the y intercept)
b is a constant giiving the graadient, or sloope, of the linne

Simplee linear regreession proviides a mathem matical wayy of estimatinng

a and b.
b It is also kknown as thee line of bestt fit or least squares lineaar
regresssion.
Simplee linear regreession defines the trend line by deterrmining the
values of a and b from
f the past data such tthat the sum of squares of
o
the verrtical differeences betweeen actual vallues (y) and values
v
obtained from the line (ŷ) is a minimum.
In demmand forecassting, a trendd line using llinear regression is
derived by fitting a straight lin
ne through thhe time seriees demand
data, with
w time on the x-axis and a demand oon the y-axiss.

Activiity 2.8
Given the
t six montths sales dataa in the folloowing table, develop a
trend line using leaast squares reegression anaalysis. Use the
t trend
line to forecast
f the next three months.
m
A
Activity Mo
onth Demand x y xy x
2

Janu
uary 115 1 115 115 1
Fe brruary 123 2 123 246 4
Ma rch 132 3 132 396 9
pril
Ap 130 4 130 520 16
May 140 5 140 700 25
Jun
ne 150 6 150 900 36
Sum 21 790 7
2877 91
e
Average 3.5 31.6667
13

Activityy feedback ccan be foundd at the end oof this moduule.

18
 C4: Operations Management

Decomposition of a time series

Decomposition of a time series occurs when the time series data is
split into the components of demand (trend, seasonal and cyclical).
The trend component measures the general upwards or downwards
direction, the seasonal component shows the effect of different
seasons, and the cyclical component shows the effect of a longer
repeating non-seasonal period cycle. Each separate component is
projected into the future and the sum of the projections becomes
the new forecast.

Seasonal index
The seasonal index is a number used to adjust data to seasonal
demand.
The seasonal index for each season is derived as the average of all
the demands for that period divided by the average demand for all
periods.
The formula to calculate the seasonal index is:
period average demand
seasonal index =
average demand for all periods
The deseasonalised demand is calculated by dividing the observed
demand for the period by the seasonal index for the period.
The formula for deseasonalised demand is:

observed demand for the period

deseasonalised demand =
seasonal index for the period

19
Unit 3

Activiity 2.9
Using the
t observedd demand datta for two yeears is shownn in the
followiing table; perrform a regression analyysis on deseaasonalised
demandd to forecast demand forr the winter season
s in yeaar three.
A
Activity
Observed
d 
Yea
ar Season
demand
d
1 n
autumn 205
winte r 140
spring 375
summe r 570
2 autumn
n 475
winte r 270
spring 685
summe r 960
Sum
m of de mand
d 3680
Ave
e rage  de mand 460

Activitty feedback can be found at the end of this moduule.

Alternative approaches to
t forecassting
Moderrn technologgy allows firm ms to improvve their interrnal and
external processes, to change behaviours
b aand subsequeently arrive ata
a betteer forecast off future dem
mand. Most foorecasting syystems acceppt
the cusstomer demaand as given n when it is qquite feasiblee for firms too
be proactive and innfluence dem mand.
Four very
v effectivee alternativee approachess to forecasting are:
1. customer collaboration
c n
2. supply chaain engineerin
ng
3. demand sm
moothing
4. proactive ccollaboration
n.
With customer
c colllaboration, it
i is possiblee for supplierrs and
custom
mers to workk together an nd share information relaating to
demannd. If a custoomer is plannning on increeasing demaand for a short
period
d by promotinng and advertising a prooduct, then itt would makke
sense to
t have suffiicient supplyy arrangemennts in place before
b the
demannd increases.
Supplyy chain enginneering usess standard opperations maanagement
techniqques to imprrove the effeectiveness off the supply chain.

20
 C4: Operations Management

Techniques such as building flexibility into processes, shortening

set-up times, minimising lead time, minimising safety stock, using
pull replenishment systems and postponement tend to reduce the
reliance on forecasts by shortening the forecast horizon.
Demand smoothing is a proactive approach that recognises the
volatility of inherent demand caused by normal consumption of the
product or service and the artificial volatility created by the
organisation’s own policies and procedures. End-of-period push,
sales contests, trade promotions, channel-stuffing, bulk discounts,
trading terms favouring start-of-month orders and pricing changes
all contribute to a pattern of demand variability beyond the normal
variability of consumption. This is artificial and is all within the
managerial control of the organisation.
The inherent volatility of demand can be measured using the
coefficient of demand variation (CDV). This is the standard
deviation of demand divided by the mean demand. The value
calculated for both supplier and customer should be the same. If
they are not the same then artificial volatility exists.
Proactive collaboration is similar to customer collaboration but
operates proactively. Business partners work together to smooth
demand and make demand patterns predictable. This results in an
effective supply chain which should lower total supply chain costs.

Activity 2.10
Work through the following questions. You may need to go back
and reread the unit to help you.

Activity

1. Describe the four components of demand.

2. Explain the difference between qualitative and quantitative
forecasting.
3. Describe the strategic importance of forecasting.
4. Describe the use of MAPD and MAPV.
5. Explain how forecasting performance might be measured.
6. Explain the expression, “Forecasting is about understanding
variation”.
7. What is the difference between seasonal variation and cyclical
variation?

21
Unit summaary

8. Discuss seasonal variationn of demandd and how ann organisation

caan respond.
9. Ex
xplain how tthe seasonal index is calcculated.
10. What
W strategiees are used by
b airlines, hotels
h and renntal car
co
ompanies to influence deemand?

Unit summary
In this unit you leaarned how too define dem mand manageement, to
explainn the nature of demand, to understannd the strateg gic role of
forecassting, to disttinguish betw
ween qualitaative and quaantitative
S
Summary forecassting, to expplain forecastt accuracy, tto define foreecast value
added, to perform basic quantiitative calcullations on foorecasting, too
define and calculatte seasonal indices,
i to usse regressionn analysis to
develoop long-termm trends and tot discuss otther approach hes to
forecassting.

22
 C4: Operations Management

Readings for further study

Armstrong, J. S. (2001). Standards and practices for forecasting. In
J. S. Armstrong (Ed.), Principles of forecasting: A
handbook for researchers and practitioners. Norwell, MA:
Reading Kluwer Academic Publishers.
Fitzsimmons, J. A.& Fitzsimmons, M. J. (2008). Service
management: Operations, strategy, and information
technology (6th ed.). New York, NY: McGraw-Hill Irwin.
Gardiner, D. (2010). Operations management for business
excellence (2nd ed.) (pp. 39–78).Auckland, New Zealand:
Pearson Education.
Gilliland, M. V. (2002, July–August). Is forecasting a waste of
time? Supply Chain Management Review, 16–23.
Gilliland, M. V. (2004). New metrics of forecasting performance.
APICS 2004 International Conference Proceedings (pp. A–
04). Alexandria,VA: APICS.
Hanke, J. E.& Wichem, D. W. (2009). Business forecasting (9th
ed.). Upper Saddle River,NJ: Pearson Prentice Hall.
Heizer, J.& Render, B. (2010). Operations management(10th ed.).
Upper Saddle River, NJ: Prentice Hall.
Johnson, R.& Clark, G. (2008). Service operations management:
Improving service delivery (3rd ed.). Harlow,England:
Prentice Hall.
Katsaros, J.& Christy, P. (2005). Getting it right the first time: How
innovative companies anticipate demand. Westport, CT:
Praegar.
Lapide, L. (1998–99, winter). Forecasting is about understanding
variations. Journal of Business Forecasting, 29–30.
Savage, S. L. (2009). The flaw of averages: Why we underestimate
risk in the face of uncertainty. Hoboken, NJ: John Wiley &
Sons.

23
Unit 4

Unit 4

Capaccity plannning annd

managgementt
Introdu
uction
Todayy, providing sservice in a competitive
c world has uncovered
u
issues relating to w
what service actually is aand how to measure
m it inn
the minds of most managers, howh service is perceived d by the
mer and how
custom w the customeer’s future purchasing deecisions mayy
be affeected by the service expeeriences theyy currently encounter.
e
To proovide a service, managem ment must fiirst anticipatee the level
and naature of custoomer deman nd. For instannce, airlines must
anticip
pate the level of demand in order to purchase
p aircraft and
supporrt staff. If cuustomer demmand is low, pprofits will be
b lower
becausse expensivee resources such as aircraaft are not uttilised. If
demannd is high it maym be that customers w will be dissattisfied by noot
gettingg the seats addvertised andd may opt foor a competittor’s offerinng.
There is a need to anticipate deemand and pprovide resou urces and —
at the same
s time — maximise profit. This requires maanagement too
take deecisions. Hoow many cusstomers a dayy can a restaaurant
servicee? How longg should a cuustomer waitt before beinng answered
by a teelephone hottline? How many
m firefighhters and apppliances
shouldd a city emplloy?
Manuffacturing com mpanies may y create inveentory bufferrs in order too
supplyy goods fromm stock even if the factorry is idle. Thhis is not so
with seervice produucts which arre intangiblee and often consumed
c annd
supplieed simultaneeously. Service can rarelly be inventooried althouggh
there are
a exceptionns such as in n emergency services whhere fire
enginees are essentiially inventooried in case they are reqquired. Also,,
most product
p offerrings come bundled
b withh services. When
W you
purchaased a car,waas it presenteed in a showwroom? Did you y require
finance to pay for the car? Waas your old car traded in as part of the
new puurchase arranngement?
Manuffactured prodducts are oftten sold “offf the shelf”, which
w means
custom
mer demand had been an nticipated andd resources organised
o too
make, assemble, shhip and storee the producct. Services cannot
c be
managged in a similar fashion. Errors or thee mismatchees between
actual demand for services andd the resourcces needed to o provide
those services
s musst be manageed without thhe ability to keep stock.
Hencee better manaagerial approoaches are neecessary. Hoow well
compeeting compannies managee resources m may have a significant

24
 C4: Operations Management

impact on both their competitive position in the market and their

profitability.
This unit focuses on answering difficult, often strategic resource
questions and explains the significance of managing capacity and
providing customer service. It begins by considering the problems
facing both manufacturing and service organisations in meeting
their respective customer demand. It considers the nature of
variability and the significance of different resources and their part
in balancing supply with demand. The study unit then investigates
yield management (or revenue maximisation) which attempts to
maximise revenue from (relatively) fixed capacity and is practised
by airlines, hotels and rental companies.
This is followed by a short discussion on flexibility and an analysis
of queues and waiting lines.

Upon completion of this unit you will be able to:

• Outline how capacity is measured and appreciate the
dilemma faced by management in matching variable
demand with variable capacity.
• Calculate various aggregate planning scenarios.
Outcomes • Discuss the strategic planning process.
• Identify various strategies for balancing supply with
demand.
• Evaluate the application of yield management.
• Discuss flexibility.
• Appreciate the customers psychology in relation to queuing.
• Discuss queues and waiting lines.

Aggregate Aggregate planning is the process used to

planning develop tactical plans to support the
organisation’s business plan. In the capacity
Terminology
planning process hierarchy, aggregate
planning follows strategic capacity planning
and is performed before short-term capacity
planning. Aggregate planning is performed for
families or groups of products and usually
includes an analysis of the plans for total sales,
total production, targeted inventory and
targeted customer backlog. The result of the
process is the production plan.

Capacity Capacity is the capability to produce output

for a time period. Capacity required represents
the process capability needed to make a given

25
Unit 4

product mix
m or deliveer a given serrvice mix
(assuming g technologyy, product
specificattionand so onn). The capaacity availablle
and the caapacity requiired can be measured
m in
the short, medium andd long term.

Capaccity Capacity planning

p is the
t process ofo
planniing determining the amouunt of capaciity required
to meet market
m demannd for produucts and
services.

Chase capacity Chase cappacity strateggy is a produuction

strateggy planning method
m that varies produuction to
meet dem mand. Producction resourcces are addedd
and removved as required and this maintains a
stable inv
ventory levell or a stable backlog
b
(queue). This
T suits firrms that expeerience
significan
nt changes inn demand and can add
and removveresources easily and effectively.
e

Deman nd Demand management

m t strategy is a productionn
managgement planning strategy
s thatt attempts to modify
strateggy demand to o meet available capacityy. It is used
in conjuncction with eiither a level capacity
strategy or
o a chase cappacity strateegy. Methodss
employed d include priccing to prom mote “off-
peak”, resstricted serviice at peak tiimes,
advertisin
ng, promotion, reservatio ons and
appointmeents.

Disecoonomies of Diseconom mies of scale refer to thee situation

scale when thiss reduction inn average unnit cost is no
longer possible througgh further inncreases in
facility sizze because co-ordination
c n of materiall
flows andd personnel becomes
b so expensive
e
that new sources
s of caapacity mustt be found.

Econoomies of Economiees of scale reefer to the drrop in the

scale average cost for each unit of outpput as a plantt
gets largeer and each succeeding unit
u absorbs
part of thee fixed costss. Economiess of scale aree
importantt to capacity decisions. The
T best
operating level is the capacity forr which the
average unit
u cost is a minimum.

Level production Level pro

oduction strattegy is a pro oduction
strateggy planning method
m that maintains reesources at a
constant level
l resultinng in a relativ
vely level
production rate. This strategy suitts a firm withh

26
 C4: Operations Management

scarce or expensive resources or when steadily

building up stock levels in anticipation of
seasonal demand.

Master Master production schedule is a line on the

production master schedule grid that reflects the
schedule anticipated build schedule for items assigned
to the master scheduler. The master scheduler
maintains this schedule, and in turn, it
becomes a set of planning numbers that drives
the material requirements planning. It
represents what the company can build and
will build expressed in specific product
configurations, quantities and dates.

Master schedule Master schedule is a report that shows by time

period the forecast, customer orders, projected
available balances, available to promise and
the master production schedule. It takes into
account the forecast, the production plan and
other things such as backlog, capacity,
material availability and management policies.

Production plan Production plan is the agreed plan that comes

from the aggregate planning process,
specifically the overall level of production
output planned to be produced, usually stated
as a monthly rate for each product family
(group of products).

Sales and Sales and operations planning is a business

operations process that helps companies keep demand
planning and supply in balance. It does that by focusing
on aggregate volumes (product families and
groups) so that mix issues (individual products
and customer orders) can be handled more
readily.

Supply Supply is the actual or planned replenishment

quantity created in response to a demand for a
product or a component in anticipation of such
demand.

Yield Yield management is the application of

management or discriminatory pricing to various market
revenue segments so that relatively fixed capacity can
management be utilised to satisfy customer requirements
and simultaneously maximise revenue.

27
Unit 4

Terminoloogy sourcedd from Gardinner (2010).

Capaccity and caapacity pllanning

Capaciity is the cappability to prroduce outpuut over a tim
me period.
Capaciity required represents thhe process capability neeeded to makke
a givenn product miix or deliverr a given servvice mix (assuming
technoology, product specificattion and so on.).
o The cappacity
availabble and the ccapacity requuired can be measured inn the short-,
medium m- and longg-term.
Capaciity planning is the proceess of determ
mining the am
mount of
capaciity required to
t meet mark ket demand for productss and servicees.
The term capacity planning haas traditionallly been usedd in
producction industrries where th
he capacity requirementss of open
manuffacturing ordders were callculated at a very detaileed level.
Producction scheduuling attemptted to scheduule jobs thro
ough the
factoryy at minimumm cost and minimum
m leaad time.
More recently,
r thee term capaccity planningg has been ussed by
inform
mation and coommunicatioons technoloogy organisaations since
they neeed to have sufficient resources avaiilable to meeet current annd
future demand.
Forecaast demand iis not compleetely predicttable and avaailable
capaciity is not commpletely preddictable. Theere are elem
ments of
predicttability in deemand and capacity
c but the summatiion includes
variatiion and unprredictability. The real siggnificance off capacity
planninng is realised when orgaanisations apppreciate the variations of
o
demannd and capaccity.
Set-upp time variatiions, producction rate varriables, unav
vailability off
resources at planneed times, unnexpected maachine break kdowns and
transpoort disruptioons all contribute to the variations
v off output that
governn capacity deecisions.

The im
mportancee of capaccity
Long-rrange capaciity planning ensures thatt sufficient resources
r aree
availabble to meet the
t long-rangge demand. For
F most org ganisations
this is 18 months aand beyond.
Mediuum-range cappacity planniing uses agggregate sales and
operatiions plans. T
The time horrizon can rannge from six
x months to a
year annd a half deppending on the
t organisattion. Capacitty incrementts
(or deccrements) haave to be in place
p to meeet the medium
m-range
demannd.
Short-rrange capaciity planning determines the requiredd capacity
from current
c time out to aboutt six months but can exteend up to a

28
 C4: Operations Management

year. Twenty or 30 years ago, capacity requirements were planned

in great detail at this level. Today’s business systems do not require
that level of detailed capacity planning since modern methods of
lean thinking and flexibility reduce the need for intense detailed
reports, provided the capacity planning has been performed at the
long-range (resource) level and at the medium-range (aggregate)
level.

Strategic capacity planning

The objective of strategic capacity planning is to specify the overall
capacity level of resources —facilities, equipment and labour —
that best supports the organisation’s long-range competitive
strategy.
If capacity is inadequate, the organisation may lose customers
through slow service or by allowing competitors to enter the
market.
If capacity is excessive, the organisation may have to reduce prices
to stimulate demand or else underutilise its workforce, carry excess
inventory, or seek additional, less-profitable products to stay in
business.

Reflection 3
Imagine you are about to launch a new venture. It may be a factory,
a restaurant, a medical centre, a retail store, a transport company, a
school or a hospital (to name a few examples).
Reflection
All production and service organisations usually occupy one or
more facilities at one or more locations. For this new venture, the
following strategic decisions need to be resolved:
1. Where will each facility be located?
2. How large, or small, will each facility need to be?
3. What process technology will be installed at each location?
4. Will the physical size of the facility be sufficient in the short-,
medium- and long-term?
5. When should capacity increments be installed?
6. What happens if the available capacity is too much?
7. What happens if it is too small?

Activity feedback can be found at the end of this module.

29
Unit 4

Aggregate planning – maatching caapacity annd demannd

Aggregate planninng seeks to fiind the combbination of sales,
s
producction, labourr requiremen
nts, inventoryy levels (prooduction) andd
custom
mer backlog (services) thhat minimisees total produuction-relateed
costs over
o the plannning period. An aggregaate plan could be a form mal
report in one comppany and an informal dirrective in another.
Capaciity policy deecisions coulld also limit demand reqquests whichh,
for exaample, are inn excess of current
c capaccity limits.
The pllanning proccess at the ag ggregate leveel considers families or
groupss of productss. Typically, a firm shouuld have 6–12 2 families orr
groupss of productss. This numbber is quite ssignificant. A number lesss
than siix suggests tthat the firm is approaching a full consolidation of o
the total plans for tthe businesss and this woould suit a buusiness
planninng exercise. A number greater
g than 12 would make the
evaluaation of the aaggregate plaans a very leengthy exercise. Assumee
aggreg gate plans are discussed at a senior eexecutive meeeting and 300
minutees is allocateed to discusss each familyy or group. If the
organiisation has, ssay, 20 group ps, then the meeting
m has a 10-hour
duratioon and the whole
w purposse of the meeeting may bee lost and
irrelevvant.
Sales are
a matched with the oveerall level off productionn output to
best meet
m general business
b objjectives of prrofitability, productivity
p y
and co
ompetitive cuustomer leadd times consiistent with thhe overall
busineess plan. Thee sales and production caapabilities arre evaluated
and a sales
s plan, production pllan, financial budget stattements, andd
supporrting plans fo
for materials and workforrce requirem ments are
develooped.
This plan affects most
m function ns of the firmm and requirres inputs
from marketing,
m saales, producttion, financee, new produ
uct and
servicee developmeent, service and
a distributtion.
The master
m producction scheduule disaggreggates the agggregate
producction plan annd specifies the
t amount and a need dattes for the
producction of speccific end products. It is a statement of
o what can
and wiill be built.

Facilitties and caapacity

Designn capacity iss the amount a firm woulld like to prooduce under
normaal circumstannces and for which the syystem was designed.
d It iss
the rate of producttion the facillity was desiigned to acco
ommodate
he long-term
over th m.
Best opperating level is the leveel for which the process was designeed
and is thus the voluume of outpput at which the
t average unit cost is at
a
a minimum.

30
 C4: Operations Management

Economies of scale refer to the drop in the average cost for units of
output as a plant gets larger and each succeeding unit absorbs part
of the fixed costs. Economies of scale are important to capacity
decisions. The best operating level is the capacity for which the
average unit cost is a minimum.
Diseconomies of scale refer to the situation when this reduction in
average unit cost is no longer possible through further increases in
facility size because co-ordination of material flows and personnel
becomes so expensive that new sources of capacity must be found.
Economies of scale occur as the average cost per unit decreases and
diseconomies of scale occur as the average cost per unit increases.
The best operating level occurs at the transition from decreasing
unit cost to increasing unit cost.
Having sufficient capacity available is a strategic decision for any
organisation. For manufacturing industries, capacity can be stored
in inventory. Service organisations do not have that option. Some
assessment of the required capacity has to be made in advance of
requirements.
Finding the right capacity in service industries has more cost
implications simply because the output cannot be stored.
Underutilised resources, including staff, do not generate as much
incoming revenue and become a serious cost to the firm. Over-
utilised resources also increase costs and impact on flexibility and
lead time.

Strategic capacity planning process

Master planning of resources
Master planning of resources is the group of business processes that
includes demand management (forecasting sales, planning
distribution, servicing customer orders) sales and operations
planning (sales planning, production planning, inventory planning,
backlog planning and resource planning), master scheduling
(preparation of master production schedule and rough-cut
planning).
Sales and operations planning
Sales and operations planning is a business process that helps
companies keep demand and supply in balance. It focuses on
aggregate volumes (product families and groups) so that mix issues
(individual products and customer orders) can be handled more
readily.
It usually occurs on a monthly cycle and displays information in
quantity and monetary units. The organisation’s strategic plan is
linked to its detailed processes. When this process is used properly

31
Unit 4

it enabbles the orgaanisation’s managers

m to view
v the orgaanisation
holistically and givves them a window
w into the future.
Masteer schedulee
The master
m scheduule is a reporrt that showss by time perriod the
forecast, customerr orders, projjected availaable balancess, available to
t
promisse and the m master producction scheduule. It takes innto account
the forrecast, the prroduction plaan and otherr things suchh as backlog,,
capaciity, material availability and manageement policiees.
Once the
t preliminaary demand numbers havve been pressented, the
capaciity choices arre balanced to arrive at w
what is achieevable in
terms ofo capacity. This may reesult in agreeeing that the planned
demannd can be prooduced with existing resources or it may requiree a
modifiication of deemand to creeate the rate oof sales conssistent with
existinng capacity.
If theree is a shortfaall in availab
ble capacity, the organisaation may
increasse capacity, to obtain tem mporary resoources, to ou
utsource
producction capacitty or simply to not deliveer the forecaast demand.
If theree is an excesss of capacityy, the organisation may decide to seell
excesss capacity or close down some resouurce.

Strategies for balancing

b supply w
with demannd
It is diifficult to sayy which shouuld come firrst —supply or demand.
An org ganisation thhat can look forward andd visualise a demand
patternn based on hhistory or theeir perceptionn of what is about to
happen n is in a stronng position to
t determinee the required capacity or
o
supplyy. This can be b determined on the basis of the quaantity and thee
timingg required.
Organisations use production planning
p straategies to deevelop the
overalll productionn output to meet
m customeer demand byy setting
producction levels, inventory leevels and baccklog levels. The main
methods used are cchase capaciity strategy, level producction strategyy
and deemand management strattegy.
The chhase capacityy strategy vaaries producttion to meet demand.
Producction resourcces are added and removved as requirred and this
maintaains a stable inventory leevel or a stabble backlog (queue).
( Thiis
strateggy would suiit firms that experience
e s
significant chhanges in
demannd and can addd and remo overesourcess easily and effectively.
e
The level productiion strategy maintains
m reesources at a constant
level, resulting
r in a relatively level
l producction rate. Thhis would suuit
a firm with scarce or expensivve resources or when steaadily buildinng
stock levels
l in antiicipation of seasonal demmand.
Deman nd managem ment strategyy attempts to modify dem mand to meett
availabble capacity.. It is used inn conjunctionn with eitherr a level

32
 C4: Operations Management

capacity strategy or a chase capacity strategy. Methods employed

include pricing to promote “off-peak”, restricted service at peak
times, advertising, promotion, reservations and appointments.

Reflection 4
For the most part, the production planner is given a sales forecast
and has to use a pure strategy or a combination of strategies.

Reflection
Think of three or four capacity or supply planning decisions that
the production planner could use to increase/decrease the available
capacity.

Activity feedback can be found at the end of this module.

Level production strategy

A level production schedule focuses on holding production and the
work force constant over a period of time. Any difference between
the constant rate of production and the varying rate of demand is
made up by allowing inventory levels to rise or fall, increasing or
decreasing the number of orders in the backlog, or changing the
length of the queue of customers.
Operations managers often prefer this method since production
rates are usually dependable, quality of outputs tend to be
consistently high, and operating costs tend to be low. The emphasis
is on production efficiency and service goals are secondary. Staff
levels can be kept constant and supply lines can be arranged to
deliver at a steady rate.
The disadvantage is that inventory levels do change and this
requires substantial warehousing arrangements to handle periods of
low demand. If a queue of customers is being managed, a process
for handling lengthy queues has to be implemented.
Examples of a level production strategy can be found in electronics
and home appliance assembly plants when high volumes of
common products are produced.
Packaging companies will often build up packaging supplies before
the season for which the packaging is required. This applies to
seasonal produce such as apples, pears, kiwifruit and some
vegetables where the seasonal demand of the product is higher than
the production capability at the time for the packaging. Therefore
the companies manufacture the packaging before it is needed and
are able to meet demand when it happens.

33
Unit 4

Chasee capacityy strategy

The chhase capacityy strategy alllows the prooduction capacity to varyy
each period
p to exaactly match thhe forecast aggregate
a deemand in thaat
time period. Workkers are emplloyed and teerminated to adjust the
level of
o the workfoorce. Overtim me and tempporary staff are
a used to
fill sho
ort periods oof high demaand. Often a subcontracto or is used to
add ad dditional cappacity duringg peak periodds.
The main
m advantagge of this meethod is that the level off finished
goods inventory caan be maintaained at a rellatively low level.
mer queues, if present, are
Custom a held at a constant
c length.
Howev ver, labour aand material costs are muuch higher because
b of thhe
disrupttions causedd by frequenttly scaling thhe work forcce up and
down and
a adjustinng capacity ofo materials ssuppliers. Esssentially, thhe
demannd is being mmatched withh available suupply.
Exampples of this sstrategy occuur in most aggricultural annd
horticuultural seasoonal harvestinng activities where the production
p
requireement occurrs for a very short time. T ff permanently
To have staff
on the payroll just for the seasonal activityy would be ana ineffectivee
use of resources.

Deman
nd managgement strategy
Some organisationns have the ability
a to modify demandd to suit theirr
availabble capacity..
One obbvious way is i to alter prrices. A firm that increasses the price
would normally exxpect demannd to fall; likkewise a firm m that
decreaases the pricee would norm mally expectt demand to rise. Pricingg
can bee an effectivee tool to prommote off-peaak demand.
Telephhone compannies, airliness, hotels and rental car coompanies alll
practisse changing tthe price of their servicee to promote off-peak
demannd. The net eeffect, thoughh, is to levell the supply.
Restauurants may ooffer a restriccted service at peak perioods. This is
not preesented as a negative offfering on theeir part. Rathher they
usuallyy promote thhe restrictionn as a positivve. They mayy offer, for
exampple, a speciall low-priced breakfast menu
m until 100:30 am eachh
day alllowing the reestaurant to focus produuction activitties on a
narroww range of chhoices ratherr than the fulll range. Thee customer
benefitts by havingg a cheaper breakfast
b opttion preparedd in a very
short lead time. Otther restauraants may dispplay a “specials” board
which sounds like it is offeringg something extra and “sspecial” wheen
in fact it just meanns that the chhefs have suffficient quanntities of those
food ittems and cann prepare theem quickly aand easily.

34
n
Ja
eb
F 32
4
0

r
p
A
y
a
M
n
Ju 4
6
0
5

C4: Operations Management

Production planning model

Production planning considers the forecast demand, the available
capacity, the required capacity, cost of production, cost of regular
and overtime labour, cost of hiring and training new staff, cost of
making employees redundant or laying off, cost of holding
inventory, cost of backlogging, cost of managing the queue and
subcontracting costs.

Activity 2.11
The data in the following table represents the demand forecast for
12 months commencing January for an organisation.

Activity

The organisation currently employs 25 employees. For planning

purposes, each employee is capable of making 200 units a month.
The cost of hiring additional staff is $600 per employee and the
cost of making an employee redundant is $300. A storage charge
of $1 per unit is made for inventory on hand at the end of each
month. This is to cover the cost of warehousing.
(Please note that the dollar amounts are nominal amounts for
planning purposes and no attempt has been made to quantify the
actual costs in this example.)
Plan 1: Develop a production plan using a level production
strategy.
Plan 2: Develop a production plan using a chase capacity strategy.

35
Unit 4

3 Develop a production plan using six

Plan 3: s months at 4800 and
six moonths at 52000.

Activitty feedback can be found at the end of this moduule.

Yield managem
ment
Yield managemen
m nt is the appliication of disscriminatoryy pricing to
variouus market seggments so reelatively fixeed capacity can
c be used to t
satisfyy customer reequirements and simultaaneously maxximise
revenu ue.
The obbjective of yyield manageement (also kknown as revvenue
managgement) is too increase rev venues for oorganisationss that operatee
with reelatively fixeed capacity.
The cooncept of yieeld managem ment was firsst applied in the airline
industrry during thee late 1970s but has sincce been appliied across a
numbeer of industriies includingg hotels, renttal cars, retail,
advertiising, electriicity generattion and trannsmission, toour operatorss,
passennger transporrt and freighht carrying. Even
E restauraants are
learninng how to usse yield mannagement to ttheir advantaage.
The appplication off yield managgement is well practisedd in the airlinne
industrry with airlinnes offering flights for as a low as onee dollar.
Airlinees may offerr discounted fares on lighhtly loaded sectors
s and
these bookings
b usuually have too be paid in full,
f are non-transferrablle
to anotther person, and non-reffundable. In some cases they t are
changeeable but witth the paymeent of a channge penalty fee f and by
paying g any applicaable fare adjuustment. Peaak-hour flighhts are
unlikelly to have m many fares att lower rates since the airrlines have
little trrouble sellinng these at hiigher rates.
Hotelss practise yieeld managem ment by offerring lower raates at
weekeends and duriing an off-seeason. They have to be aware
a of bothh
predicttable, seasonnal factors annd unpredicttable, individdual custom
mer
demannd by using a systematic approach w with a combinnation of
knowledge, experiience, undersstanding andd forecastingg. This
combinnes predictaability and un ncertainty.
Yield managemen
m nt is most efffective whenn the followinng exist:
• Relatively fixed capaccity and thee same unit of o capacity
can be useed in varietyy of ways. There would be b no need to
t
manage yieeld when cappacity is flexxible. With relatively
r
fixed capaccity such as an airline seeat or a hotell room, the
organisatioon cannot eassily change tthe capacity and yet
demand fluuctuates widely.
• Demand ccan be segmented by maarket and each segmen nt
has varyinng needs, beehaviour andd willingnesss to pay.
Airlines segment their market into general classsifications of
o

36
 C4: Operations Management

business class, full economy fares and discount fares.

Business travellers are time sensitive and are likely to book
late and require maximum flexibility. For that flexibility
they pay a higher-priced fare.
• Demand is highly variable (seasonal fluctuations) and
uncertain. In periods of low demand (winter for the
tourism industry), demand needs to be stimulated and in
periods of high demand (summer in the tourism industry),
revenues need to be maximised.
• Inventory is perishable (hotel room by night, airline seat
by flight). If an airline seat is not sold by the time the flight
departs, the revenue from that seat is lost and can never be
recovered. Similarly, the revenue from an empty hotel room
for the night can never be obtained.
• Product can be sold in advance. Customers choose to buy
at different times, either well in advance or at the last
minute. Suppliers can take advantage of this buyer
behaviour and they also have an opportunity to influence
that buyer behaviour. When the same price applies
regardless of booking time, the only driver that would
encourage customers to buy earlier would be a fear of
unavailability of supply.
• Product can be forecast with relatively high accuracy.
The organisation will dynamically adjust pricing and will
hold back on some capacity to be sold at the last minute at a
premium. The company needs to know how much to hold
back. If it holds back too much, it may miss out on any
revenue. If it holds back too few, it misses out on the high
revenue-generating last-minute sales.
• Fixed costs are high and marginal costs of selling one
extra unit are low. The cost of adding an additional aircraft
with additional seats to the fleet is high. The cost of adding
an additional hotel room requires a new hotel to be
constructed and this is relatively expensive. The cost of
adding one more passenger to the passenger list is relatively
low. The airline has to handle the reservation, handle the
check-in, carry the luggage and make sure the passenger
arrives at their destination.
• Price is not an indicator of quality. Price should not be
seen as a status symbol and should not be an indicator of
quality. Using airline pricing as an example, most
customers realise that if they pay a high price for their seat
they will not get a higher quality flight when compared to a
discount purchaser who purchases the same class of seat.

37
Unit 4

The differeence is in thee timing of thhe purchase,, not the

delivery off the service..
• Producerss are profit-oriented an nd have freeedom of
action. Yieeld managem ment assumees that the su
upplier of thee
service is profit-oriente
p ed. A hotel ccan charge different
d ratess
for each rooom and can hold back soome rooms in i
anticipationn of receivinng higher revvenue later. This
T
approach would
w not bee feasible in aan emergenccy ward of a
public hosppital.

Yield managem
ment proceess
The yiield managemment process starts by deetermining how
h far in
advancce the system
m will look ahead.
a Airlinnes typically
y use 300
days.
The market
m is segm
mented, baseed on future purchasing behaviour.
b
Airlinees have a cleear segmentaation betweeen leisure andd business
travelllers. Hotels have
h short-teerm, long-terrm, leisure, business
b andd
conferrence guests.. Rental car firms
f have similar segm ments to hotells.
The suupplier prediicts customerr demand baased on foreccast demandd
and caapacity at eacch product/pprice level annd attempts to
t optimise
the priice by matheematically deetermining ccapacity avaiilability and
price that maximisses expectedd profit.
This atttempts to alllocate the riight capacityy to the rightt customer att
the rigght time and simultaneouusly maximisses revenue or yield. It
relies on
o being ablle to predict the expectedd behaviour of specific
markett segments within
w the ovverall markett demand. Su uccessful
implem mentation off yield manag gement requuires the orgaanisation to
be cappable of conttinually mon nitoring and forecasting
f c
changes in
demannd patterns.
The suupplier dynammically recaalibrates and continually monitors
performmance and reacts
r to the updated marrket response. It is a
continuuous process keeping suurveillance on
o response tot the pricingg,
compeetitor reactions and makiing adjustmeents.
The management
m o demand attempts
of a to innfluence buy
yer behaviouur
so thatt it fits in witth available capacity. Opptions to infl
fluence
demannd include:
• Partitioninng demand or segmentiing the marrket based on
o
purchasing behaviourr, not just current or past
classificatiions. The moost commonn example is the
partitioningg of businesss and leisuree travel and
accommoddation.
• Offering off-peak
o priccing incentiives. When off-peak
o
services aree priced at a lower rate than
t full-peaak services,

38
 C4: Operations Management

customers will deliberately delay their demand for the

service until the price incentive takes effect.
• Promoting off-peak demand. This is similar to the off-
peak option but relates to promoting activities and
advertising that promotes off-peak demand almost to the
exclusion of promoting on-peak activities.
• Developing reservation systems. Customers know that
they have to make a reservation or an appointment so they
are encouraged to book early to avoid disappointment. This
has an added effect of controlling demand to the capacity
limit.
The options to manage supply are:
• Share capacity with another supplier. This capacity
sharing occurs with airline companies when they fly one
aircraft on a particular route and it carries two or more
airline flight numbers.
• Increase customer participation. This may occur at all
times or just in peak periods or off-peak periods. On-line
banking and automatic teller machines allow customers to
conduct banking business without the need for bank
personnel to be present while the transaction is occurring.
This allows the fixed capacity of the banking system to be
used for longer periods without needing additional staff.
• Cross-train employees. If demand shifts from one area of
the business to another, trained staff can be directed to meet
that demand. If cross-training were not in place the
organisation may not be able to offer the full range of
services at all times.
• Employ part-time. This allows staff numbers to be
increased and decreased almost at will to meet whatever
demand is placed on the organisation.
• Create adjustable capacity. This occurs when
organisations do not normally use all available capacity and
they open up the additional capacity only in peak periods
and close it in quiet times. Tourist destinations often operate
in this manner with some hotels closing completely for the
off-season.

Capacity flexibility
Capacity flexibility essentially means having the capability to
deliver what the customer wants within a shorter lead time than
competitors can offer. Such flexibility is achieved through flexible

39
Unit 4

plants,, processes, wworkers andd through straategies that use

u the
capaciity of other oorganisationss.
The acctual capacityy is the level of output ffor a process or activity
over a period of timme. The effeective capaciity is the outtput rate thatt
managgers expect ffor a given acctivity or proocess. The demonstrated
d d
capaciity is the proven capacityy calculated from actual performancee
data.
Capaciity control iss the processs of measurinng productioon output and
compaaring it with the capacityy plan, determmining if thee variance
exceedds pre-establlished limits,, and taking corrective action
a to get
back on
o plan if thee limits are exceeded.
e
Set-upp time is one of the determinants of ccapacity flexxibility. Set-
up occcurs at the staart of a prodduction run aand total set--up time is a
functioon of the nummber of prod duction runs executed. A machine iss
set up and the run size quantity y is produced. The machhine is then
set up for the next product. Orrganisations often look at a the
producction rate to determine capacity. Cappacity, thouggh, needs to
considder set-up tim
me as a non-pproductive pperiod as welll as the
processsing speed tto determine capacity off a machine.
A prodduction facillity works beest when it fofocuses on a fairly limited
set of production
p oobjectives beecause a firmm should nott expect to
excel in
i every aspeect of production perform mance —cosst, quality,
flexibiility, new prooduct introdductions, reliaability, shortt lead times,,
and loww investmennt.
Service capacity must
m be availlable to prodduce a servicce at the timee
it is neeeded. Emptyy airline seats, for exampple, cannot be
b transferreed
from ana off-peak fflight to a fulll-peak flighht.
m be locatted near the customer or at least be
Service capacity must
availabble to the cuustomer before the servicce can be dellivered.
Havingg empty hoteel rooms in one
o city doees not help a shortage of
hotel rooms
r in anoother city.
Custom mers interactt with the seervice producction system
m and as a
result the
t processinng time requ uired for each delivery may
m vary.
Services experiencce off-peak anda on-peak as well as inn-season andd
off-seaason demandds and this iss caused by ccustomer behaviour
influenncing demannd.
Service providers m must considder the day-too-day relatioonship
betweeen service uttilisation and
d service quaality. A goodd operating
point is
i near 70 peer cent of thee maximum. This is enou ugh to keep
serverss busy but alllows time too serve custoomers indiviidually and
keep enough
e capaccity in reservve so as not to create tooo many
managgerial headacches.
Low raates are apprropriate wheen both the ddegree of unccertainty andd
the staakes are highh, such as an accident and emergencyy departmennt

40
 C4: Operations Management

at a hospital and fire services. Sporting events and concerts prefer

sell-out crowds that use the maximum capacity.

Queues and waiting lines

Queues are a natural consequence of service delivery. In many
ways a queue in a services environment is equivalent to buffer
inventory in manufacturing.
In production industries, the inventory acts as a buffer to allow the
rate of inputs to be different from the rate of outputs.
In services, the queue acts as the buffer between the arrival rate of
customers and the supply and delivery rate of the actual service.

Activity 2.12
When you are waiting in a queue, it often feels like you are
waiting for a very long time. Make a list of possible reasons why
a customer perceives the wait time is longer than it actually is.

Activity
Activity feedback can be found at the end of this module.

Queuing theory
Queuing theory is used to manage processes. A queue can be
studied in terms of the source of each customer, how frequently
customers arrive, how long they can or should wait, whether some
customers should jump ahead in the queue, how multiple queues
might be formed and managed, and the number of servers required.
The design of a call centre provides an excellent example of
queuing theory in practice. Call centre performance is typically
measured by the cost per call, the resolution rate or fulfilment rate
and customer satisfaction.
When customers are in the queue they may leave and be lost to the
system. They may call back or they may hold one line and use
another telephone to join the queue again. An increase in the
average call duration increases the queue.

41
Activiity 2.13
1. What
W does thee term “capacity” mean?
2. Hoow does cappacity differ from capability?
3. Why
W is capaciity managem
ment strategiccally important?
A
Activity
4. Th ment of capaccity for serviices is more difficult thann
he managem
for manufactuuring. Why?
5. Deescribe the capacity
c considerations ffor a hospital and identiffy
hoow this is diffferent from a manufactuuring unit.
6. What
W are the possible
p consequences of demand raate being
different from
m design capaacity rate?
7. What
W is yield managemennt?
8. W industriees use yield managemennt and why?
What
9. Deescribe threee strategies for
f expandinng capacity.

Unit summary
In this unit you leaarned how caapacity is meeasured, andd appreciatedd
the dileemma facedd by managem ment in matcching variabble demand
with vaariable capacity. We dev veloped various aggregaate planning
S
Summary scenariios and discuussed the strrategic plannning process and
develooped various strategies fo or balancing supply withh demand
We expplored the appplication off yield manaagement (or revenue
managgement) and discussed fleexibility.
We conncluded withh a section on
o the custom mers’ psychoology in
relation
n to queuingg, queues andd waiting linnes.

42
 C4: Operations Management

Assignment
There are three questions in this assignment.

Question 1 40 marks
Assignment a. Lowering prices can increase demand for products or services,
but it also reduces profit margins if the product or service
cannot be produced at lower cost. Briefly discuss how an
operations manager should approach his or her job when
competing on cost.
b. Quality is a dimension of a product or service that is defined by
the customer. Today, more than ever, quality has important
market implications. Briefly discuss how an operations
manager should approach his or her job when competing on
quality.
c. As the saying goes, “time is money.” Some companies do
business at “Internet speed,” while others thrive on consistently
meeting delivery promises. Briefly discuss how an operations
manager should approach his or her job when competing on
time.
d. Flexibility is a characteristic of a firm’s operations that enables
it to react to customer needs quickly and efficiently. Some firms
give top priority to flexibility. Briefly discuss how an
operations manager should approach his or her job when
competing on flexibility.

Question 2. 30 marks
a. Provide three questions that should be considered when
developing the objectives of a forecast.
b. Name three different models that could be developed and tested
during the forecasting process.
c. What does “applying the model” mean?
d. Explain, using an example, the forecasting step “considering
real-world constraints on the model’s application”.
e. Explain how one might “revise and evaluate the forecast”.
f. What is the most important rule of forecasting and what should
we be trying to achieve?

43
Assignmennt

Questiion 3: 30
3 marks
The nuumber of gueests staying at an exclusiive lodge haas been:
Quarter Year Actual
S
Summer 1 73
Autumn 1 104
Winter 1 168
Spring 1 74
S
Summer 2 65
Autumn 2 82
Winter 2 144
Spring 2 52
S
Summer 3 89
Autumn 3 146
Winter 3 205
Spring 3 98
Total 1300

a. Caalculate seasoonal indices using the abbove data.

b. Deeseasonalise the above daata, and deteermine the reegression
equuation.
c. Using the regreession equattion, determiine the foreccasts for yearr
fouur.

44
 C4: Operations Management

References
Gardiner, D. (2010). Operations management for business
excellence (2nd ed.). England, New Zealand: Pearson
Education.
Johnson, R. & Clark, G. (2008). Service operations management:
Improving customer service (3rd ed.). Harlow, England:
Prentice Hall.

Further Readings
Fitzsimmons, J. A.& Fitzsimmons, M. J. (2008). Service
management: Operations, strategy, and information
yechnology (6th ed.). New York, NY: McGraw-Hill Irwin.
Reading Gardiner, D. (2010), Operations management for business
excellence (2nd ed.) (pp. 79–118). Auckland, New
Zealand: Pearson Education.
Heizer, J.& Render, B. (2010). Operations management (10th
ed.). Upper Saddle River, NJ: Prentice Hall.
Johnson, R.& Clark, G. (2008). Service operations management:
Improving service delivery (3rd ed.). Harlow, England:
Financial Times/Prentice Hall.
Netessine, S.& Shumsky, R. (2002). Introduction to the theory
and practice of yield management. Retrieved from
INFORMS Transactions on Education, 3(1), 34–44.
Available online at [Link]
Shy, O. (2008). How to price: A guide to pricing techniques and
yield management. Cambridge, NY: Cambridge
University Press.
Talluri, K. T.& Van Ryzin, G. J. (2005). Theory and practice of
revenue management (International series in operations
research and management science). New York, NY:
Springer Science and Business Media.
Wallace, T. F. (1999). Sales and operations planning: The how-to
handbook. Cincinnati, OH: T. F. Wallace.

45
Activity feedback

Activiity feedbback
Reflecctive activitties
Reflecttion 1
A nummber of factoors influence demand succh as changees in
technoology, compeetitor initiatiives or pricinng levels. Foorecasting
helps firms
f to focuus on the facctors that inflluence demaand and
establiish a relationnship betweeen those facttors and actuual demand.
Reflecttion 2
Forecaast inaccuraccy can be attrributed to thhe following causes:
• The forecaasting model or method eemployed maay not be
suitable forr the demandd being monnitored.
• The informmation in the forecasting process mayy arrive too
late to be of
o significantt value.
• True demaand is not beiing capturedd and it is beiing confusedd
with sales ddata.
• Appropriatte data is nott being used.. This featurre develops
when indivviduals are leeft to source informationn for
themselvess and then thhis data is consolidated in n some way
at an organnisational levvel.
• Forecasts aare calculated from the past
p data thatt may not
hold for projected dataa points.
Reflecttion 3
All of these questions have a major
m bearinng on the succcess, or
wise, of the oorganisation.. If you are capable
otherw c of geetting it right,
you wiill find yourself in an en
nviable positiion of being able to
capitallise on everyy opportunityy that comess your way (assuming yoou
want too take it) andd thus maxim
mise revenuees and profitts. If you gett
it wronng, you mayy find yourseelf searching for additionnal capacity at a
a prem
mium price or being left with
w excess capacity thaat you are
unablee to sell.
Reflecttion 4
• Hiring addditional staff and making staff redund
dant.
• Working vvariable dayss per week.
• Working overtime.
• Varying thhe level of inventory.
• Varying thhe number off orders in thhe backlog.

46
MAD n= =
10

=
=
bias ∑950
D−
(6
5 000

= 6000
.8
0
% 33

C4: Operations Management

• Varying the length of the queue of customers.

• Using subcontractors to supply additional capacity.
• Outsourcing parts of the business to free up resources.
• Adding or removing temporary capacity.
• Adding or removing permanent capacity.
These are reactive measures and the controllability of these factors
depends on union agreements, employment contracts, employment
legislation, short-term constraints on physical capacity levels,
customer requirements and preferences, and the amount of money
that can be tied up in inventories.
Activities
Activity 2.1
Calculate the absolute deviation between demand values and
forecast values for each month. Add them up and find the average
(mean) value. This is the mean absolute deviation (MAD).

Thus the actual demand is, on average, 45 units from the forecast
value.

Activity 2.2
Bias is found by calculating the algebraic difference between
demand value and forecast value for each period. To make sure that
the algebraic sign is correct, ensure you subtract forecast from
demand (D – F). The sum of the algebraic differences is divided by
the sum of the demand values and expressed as a percentage.

Thus the forecasting model has a bias in favour of demand of 0.833

per cent.

47
Activity feedback

Activityy 2.3
MAPD D is the meann of the absoolute deviation between actual
demannd value and forecast vallue divided bby the mean of the
demannd values exppressed as a percentage. In this formmula
descrip
ption, both nnumerator annd denominaator calculatee average
values using the nuumber of observations. IIn the formuula, the
numbeer of observaations, n, couuld appear inn numerator and
denomminator and cancels
c each other out.

D=∑
D-F x100
x 450xx100
MAPD = = 7.5%
%
∑D 60000

Mean absolute perrcentage variiation (MAP PV) is the av

verage of the
absoluute deviationn between acttual demandd value and mean
m demannd
value divided
d by thhe mean dem
mand expressed as a perccentage.

∑ D −∑n x100
D
860x100
MAPV
V= = = 14.33%
1
∑D 6000

Thus the
t mean abssolute deviattion is 7.5 peer cent of thee mean
demannd (MAPD) aand the variaability of demmand (MAP PV) is 14.33
per cennt.

Activityy 2.4
In this example n = 3 and t = 6.
6

n = 3, t = 6
sum oof actual dem
mand values for the chossen number of periods
Ft =
mber of periods
chosen num
D + ...... + Dt −2 + D t −1
= t −n
n
D + D4 + D5
= 3
3
110 + 135 + 145
=
3
= 130

Thus, the
t forecast for month siix using a thhree-month simple
s
movin
ng average iss 130.

48
 =
F
t .sum
n+..... of actual
t−
D 2chosen
t−
+
D demand
number
1values
t− of for
periods
chosen number of perio

=
120 +
1230
D +
4n35
153D
110 5
145

=
128 5

t=
F sum of (each
+t−
nwD +
.....
t−
D period'
s demand
sum
+
t−
2w
D of the1wt−exvalue
ach
weights
period' s weight)

=
6=
tF t−
3xw
nW
(D +
4x)w
.....
(D (4t−
)+1W
2w xW
5D )

=
6
F x0.2(110
135 )+
(135 x0.3) x+
(145 0.5)

C4: Operations Management

Activity 2.5
In this example n = 5 and t = 6.

Thus, the forecast for month six using a five-month simple moving
average is 128.

Activity 2.6
In this examplen = 3, t = 6.

In this example, the sum of the weights adds up to 1. Thus, the

forecast for month six using a three-month weighted moving
average is 135 units.

49
Activity feedback

Activityy 2.7
t = 6, F5 = 13 5, α = 0.2
30, D5 = 145
Ft = Ft −1 + α (Dt-1 − Ft −1 )
= 130 + 0.2 x (145 −130)
= 133
Thus, the
t forecast for month siix using expponential smooothing withh
α equaal to 0.2 is 1333.

Activityy 2.8
The daata as supplieed has demaand data for ssix months. The monthlyy
demannd is the indeependent varriable and is assigned to the x-axis.
The months
m in the x-axis are numbered
n 1 tthrough 6. The
T y-axis is
for thee dependent variable
v andd this is the oobserved dem
mand.
In ordeer to calculaate a,b and thhe trend line,, the values for
f xy and
2
x are required
r as sshown in the table abovee.

n = 6, ∑ xy = 287
77, ∑ x = 21, ∑ y = 790, ∑ x 2 = 91
n∑ xy − (∑ x ∑ y )
b=
n ∑ x 2 − (∑ x )
2

6 x 2877 − 21x 790

=
6 x 91 − 212
= 6.4

n = 6, ∑ xy = 287
77, ∑ x = 21, ∑ y = 790, ∑ x 2 = 91

a=
∑ y −b ∑ x
n n
= 13
31.67 − 6.4 x 3.5
= 10
09.27

yˆ = 10
09.27 + 6.4 x trend line equation
n

By usiing Excel thee calculationns can be auttomated. Thee Excel

softwaare requires the
t data anallysis add-in and the calculation is
initiateed using the Data menu followed by Data Analyysis and thenn
Regresssion. For thhe Input Y Range
R select tthe column of
o y-values
and for the Input X Range seleect the colum mn of x-valuues. The
outputt report contaains more daata than is im
mmediately required
r andd
the perrtinent valuees are shownn below.

50
seasonal index for autumn =
340
460
205 =
0.7average
391 demand for all periods

seasonal index for spring

winter =
530
460 =
1.04457
522

seasonal index for summer =

765
460 =
1.6630

C4: Operations Management

Substituting x=7, x=8 and x=9 into the trend line equation provides
the forecast values as shown below.

Activity 2.9
Start by calculating the seasonal indices for autumn, winter, spring
and summer.
The period average demand for autumn is (205 + 475) / 2= 340
The period average demand for winter is (140 + 270) / 2 = 205
The period average demand for spring is (375 + 685) / 2 = 530
The period average demand for summer is (570 + 960) /2 = 765
The average demand for all periods can be calculated as 3680/8 and
is given as 460.

Calculate the deseasonalised demand for each season by dividing

the observed demand by the seasonal index for that period as
shown in the following table. Then calculate the extended fields for
xy and x2 as shown in the following table.

51
Activity feedback

Observ ed  Seasonal  Deseasonalise

ed  2
Ye
ear Season
n x xy x
nd
deman index demand y
1 autum
mn 205 1 0.7391 277.3644 277.3
3644 1
winte r 140 2 0.4457 314.1126 628.2
2252 4
g
spring 375 3 1.1522 325.4643 976.3
3929 9
summe
er 570 4 1.6630 342.7541 1371.0
0164 16
2 autum
mn 475 5 0.7391 642.6735 3213.3
3675 25
winte r 270 6 0.4457 605.7886 3634.7
7316 36
g
spring 685 7 1.1522 594.5148 4161.6
6036 49
summe
er 960 8 1.6630 577.2700 4618.1
1600 64
3680 36 8 3679.9423 18880.8
8616 204

n = 8, ∑ xy = 18880.8616, ∑ x = 36, ∑ y = 3679.94223, ∑ x 2

= 204
2

n∑ xy − (∑ x∑ y)
b=
n∑ x 2 − (∑ x))
2

6 − 36 x 367
8 x 18880.8616 79.9423
=
2 − 36
8 x 204 2

= 55.22648

a = ∑ −b∑
y x
n n
= 460 − 55.2648 x 4.5
= 211..308 ∴ yˆ = 211.308
8 + 55.2648x

Now calculate
c b, a and the besst estimate of ŷ= a + bx.
Thus the
t trend linee for deseasoonalised dataa isŷ = a + bxx = 211.308 +
55.26448x.
Now substitute
s x== 10 correspoonding to winter in the thhird year to
get thee deseasonalisedvalue foor that periodd.

whhen x = 10,th
he deseasonaalised value for winter in
n the third yeear is
yˆ = 211.308 + 55.26448x
= 211..308 + 55.26
648x10
= 763..956

By usiing a program m such as Ex xcel, the calcculations to get the

regresssion line cann be automatted. The Exccel software requires the
data annalysis add-iin and the caalculation is initiated usiing the Data
menu followed
f by Data Analy ysis and then Regression.

52
 C4: Operations Management

For the Input Y Range select the column of y-values and for the
Input X Range select thecolumn of x-values. The output report
contains more data than is immediately requiredand the pertinent
values are shown in the following table.

The deseasonalised forecast for winter in the third year is:

Now multiply the deseasonalised forecast by the seasonal index to

calculate the seasonalised forecast for winter in the third year:
763.956 x 0.4457 = 340.4952 = 340 (0 dp)

Activity 2.10
All answers are in the learning material.

Activity 2.11
Plan 1: Develop a production plan using a level production strategy.
Start this plan by calculating the level production rate. The annual
demand is 60,000 and there are 12 monthly periods so that makes
5000 units a month.
In January, the beginning inventory is zero, the production is 5000
and demand is 4400, therefore the ending inventory is 600 units.
In February, the beginning inventory (following on from January)
is 600, the production is 5000 and demand is 3200, therefore the
ending inventory is 2400 units.
In March, the beginning inventory (following on from February) is
2400, the production is 5000 and demand is 4000, therefore the
ending inventory is 3400 units.

53
Activity feedback

Contin
nue like this for the rest of
o the year.

Beg
ginning  End
ding  ber 
Numb
mand 
Dem nt 
Redundan
M
Month inve
entory  Prod
duction  inven
ntory  of  aff
New sta
fore
ecast staff
on
n hand on h
hand employ
yees
Jan 0 50
000 44
400 60
00 25
Feb  6
600 50
000 32
200 240
00 25
Mar  2
2400 50
000 40
000 340
00 25
Apr 3
3400 50
000 54
400 300
00 25
May 3
3000 50
000 66
600 140
00 25
Jun 1
1400 50
000 50
000 140
00 25
Jul 1
1400 50
000 40
000 240
00 25
Aug  2
2400 50
000 30
000 440
00 25
Sep 4
4400 50
000 48
800 460
00 25
Oct 4
4600 50
000 64
400 320
00 25
Nov 3
3200 50
000 70
000 120
00 25
Dec  1
1200 50
000 62
200 0 25
60
0000 600
000 280
000

The innventory storrage cost is $28,000,

$ the cost of empploying new
staff iss zero, and thhe cost of terrminating staff is zero too give a totall
cost foor this plan oof $28,000.
Plan 2: Develop a production plan
n using a chasee capacity stra
ategy
In this plan the prooduction ratee varies to m
match the dem
mand patternn
and thee number off employees is increased or decreasedd to match
the pro
oduction ratee.
In Januuary, the dem
mand forecaast is 4400, so productionn is set to
match that rate. Beeginning invventory on haand is zero, production
p
matchees demand fforecast, so thhe ending innventory on hand
h is zero.
To prooduce 4400 wew need 22 staff
s (200 unnits per emplloyee per
monthh) so three emmployees aree made redunndant. Theirr employmennt
contract would speecify the tem
mporary natuure of their emmployment.
In Febbruary, the deemand foreccast is 3200, so productioon is set to
match that rate. Beeginning invventory on haand is zero, production
p
matchees demand fforecast, so thhe ending innventory on hand
h is zero.
To prooduce 3200 wew need 16 staff
s (200 unnits per emplloyee per
monthh) so six empployees are made
m redunddant.

54
J
e
F
a
M n
b
r 0 2
3
0
4 0
2
3
4 0 0
1
2 2
6
0 3
6
4

p
A
M
Ju y
a
0
r
ln 6
5
0
4 6
5
0
4 0 3
2
0 7
3
5
0 7
6
8
5

C4: Operations Management

In March, the demand forecast is 4000, so production is set to

match that rate. Beginning inventory on hand is zero, production
matches demand forecast, so the ending inventory on hand is zero.
To produce 4000 we need 20 staff (200 units per employee per
month) so four employees are hired.
The remaining months are calculated in a similar fashion.

The inventory storage cost is zero, the cost of employing new staff
is $22,200, the cost of terminating staff is $9,300 to give a total
cost for this plan of $31,500.
Plan 3: Develop a production plan using six months at 4800 and six months
at 5200.
The calculations for this strategy follow the same pattern as plan
one and two except that the production rate is set at 4800 for the
first six months, then increases to 5200 for the rest of the year. This
represents a starting position in trying to optimise the plan. The
number of employees is increased or decreased to match the
production rate.

55
Activity feedback

eginning 
Be En
nding  mber 
Num
emand 
De ew 
Ne dant 
Redund
Month in
nventory  Pro
oduction  inve
entory  o
of 
fo
orecast emplo
oyees employ
yees
o
on hand on
n hand emplloyees
Jan 0 4800 4400 4
400 2
24 1
Feb  400 4800 3200 2
2000 2
24
Mar  2000 4800 4000 2
2800 2
24
Apr 2800 4800 5400 2
2200 2
24
May 2200 4800 6600 4
400 2
24
Jun 400 4800 5000 2
200 2
24
Jul 200 5200 4000 1
1400 2
26 2
Aug  1400 5200 3000 3
3600 2
26
Sep 3600 5200 4800 4
4000 2
26
Oct 4000 5200 6400 2
2800 2
26
Nov 2800 5200 7000 1
1000 2
26
Dec  1000 5200 6200 0 2
26
60000 6
60000 20
0800 2 1

The innventory storrage cost is $20,800,

$ the cost of empploying new
staff iss $1,200, thee cost of term
minating stafff is $300 to give a total
cost foor this plan oof $22,300.

Activityy 2.12
Johnsoon and Clarkk (2008) idenntified that thhe customer often
perceivves that the time
t in the queue
q is longger than it reeally is. Theyy
observved the followwing:
• Unoccupieed time feels longer than occupied tim
me.
• Pre-process waits feel longer
l than in-process waits.
w
• Anxiety maakes the waiit seem longer.
• Uncertain w
waits are lonnger than known, finite waits.
w
• Unexplaineed waits seem
m longer thaan explainedd waits.
• Unfair waits are longerr than equitaable waits.
• The more vvaluable the service, the longer the customer
c
waits.
• Solo waitinng feels longger than grouup waiting.
• Uncomforttable waits feel
fe longer thhan comfortaable waits.
• New or inffrequent userrs feel they wait
w longer.

Activityy 2.13
All ansswers are in the learningg material.

56