3 Capacity Management

Manual Section 3
Capacity Management

Page 1
3 Capacity Management

Table of Contents
1. Introduction .............................................................................................................. 4
2. Capacity Planning ..................................................................................................... 7
2.1. Resource or Aggregate Capacity Management .............................................................. 8
2.2. Rough-Cut Capacity Planning ........................................................................................ 8
2.3. Capacity Requirements Planning .................................................................................. 9
2.4. Material/Manufacturing Requirements Planning (MRPI/II) ........................................... 9
2.5. Capacity Planning Techniques ...................................................................................... 9
Capacity Planning for Services .................................................................................................................... 10

2.6. Conclusion ................................................................................................................. 10

3. Determination of Required Capacity ........................................................................ 11
3.1. Factors of Production ................................................................................................. 11
Additional Factors ....................................................................................................................................... 14

3.2. Adjustments for Defective Output .............................................................................. 15

3.3. Adjustment for Efficiency ........................................................................................... 15
3.4. Learning Curves ......................................................................................................... 17
3.5. Machine Utilisation .................................................................................................... 19
3.6. Statistical Variations .................................................................................................. 21
3.7. Rough Cut Capacity Planning ...................................................................................... 21
4. Capacity Management Strategy .............................................................................. 22
4.1. Strategy 1: Provide for efficient adjustment or variation in capacity ........................... 24
4.2. Strategy 2: Eliminate the need for Adjustments .......................................................... 24
4.3. Evaluating Plans ......................................................................................................... 26
Discounted Cash Flow ................................................................................................................................. 26

5. The Control of Capacity ........................................................................................... 27

5.1. Introduction .............................................................................................................. 27
5.2. Cumulative Graphs..................................................................................................... 27
5.3. Cumulative Cost of Production ................................................................................... 28
5.4. Linear Programming ................................................................................................... 29
Example Linear Programming Problem....................................................................................................... 30

5.5. Duality ....................................................................................................................... 31

6. Summary ................................................................................................................ 35
7. Seminar Questions (Full-Time Students)................................................................... 36

Page 2
3 Capacity Management

8. Appendix: Learning in production ............................................................................ 37

8.1. Introduction .............................................................................................................. 37
8.2. The Learning a Curve Defined ..................................................................................... 38
8.3. Analytical Studies of Production Learning ................................................................... 39
8.4. Transfer of Learning between Parallel Production Facilities ......................................... 39
8.5. Detailed Study of Learning Influences ......................................................................... 40
Data ............................................................................................................................................................. 40

8.6. The Analysis and Results ............................................................................................ 41

8.7. Implications of Results ............................................................................................... 41
8.8. Qualitative Assessments ............................................................................................ 42
References...................................................................................................................... 45

Page 3
3 Capacity Management

1. Introduction
Of all the products we have encountered in our lives, whether it is a trendy new smart phone
or car, it has very likely to have been subjected to some degree of Operations Management.
But, it is not only manufacturing enterprises which have had directions from operation
managers, the service sector too, benefits under their guidance.
Capacity for Original Equipment Manufacturers (OEMs) can include available machine time,
and available equipment etc. For services, capacity could relate to the number of beds in a
hospital or even the number of customers at a restaurant, (Adenso-Diaz, et al., 2002). For
either sector, manufacturing or services, the capabilities or capacity of an organisation is
central to its existence. Therefore, the management of capacity is arguably the most
important activity within the sphere of Operations Management (OM) and we know already
that OM provides the mechanism for running an organisation efficiently. As a reminder, Figure
1 outlines the relationship of OM within an organisation.

Figure 1 - Operations management overview, adapted from (Slack, et al., 2016, p. 3)

This section will look closer at the different areas and issues which arise while managing
Capacity. As Capacity is at the heart of Operations Management, it is inevitable that you will
find the other concepts within Logistics and Operations Management to come together and
overlap with each other.
“The definition of the capacity of an operation is the maximum level of value-added activity
over a period of time that the process can achieve under normal operating conditions.”
(Slack, et al., 2016, p. 351). Within an Engineering context, Capacity is defined as the
maximum amount of output that can be produced with the existing stock of capital.
The overall objective of managing capacity is to match the level of available resources to the
level of demand (See Figure 1). Too few resources and we fail to meet demand, we lose
turnover and essentially profit on the lost business. As a result our customers are dissatisfied
and we are likely to lose market share, in addition we may face penalty costs for late delivery.

Page 4
3 Capacity Management

Too many resources and we incur punitive costs from an idle plant, facilities and manpower.
Effective capacity management is therefore a clever balancing act that seeks out the most
beneficial means of reconciling capacity and demand.
Capacity management tasks may be divided arbitrarily into long and short-term. Long-term
tasks are principally concerned with planning and the short-term with control (Figure 2). The
division in time is a device which simplifies the study of the subject but equally it is an artificial
step since the decisions taken in the long term will interact strongly with those that need to
be taken in the short term.

PLANNING
Long
Selection of products Term
Identification of
markets

Determine capacity
required
Capacity
Develop strategy for Management
resources
Manipulate resources
for changes in demand

CONTROL

Figure 2 - The long-term, short-term and planning/control aspects of capacity management

The chief problems arising in the management of systems capacity often occur due to the
uncertainty about the balance between the demand and activities in the future, manifested
by:
a) The number of orders which will be received and/or
b) The resources necessary to satisfy the demand

It should become self-evident that the uncertainty should be reduced as far as possible by
attempting to make some forecast of customer requirements within the planning period.

Module Link: What is the purpose of a forecast?

You will have previously read that no system of forecasting is totally accurate. However, there
are recommendations which can be carried out in order to minimise the error of your
forecast, and thus reduce the uncertainties. To have no forecast at all would be an
unreasonable attempt to plan against a totally unknown demand. It is not the intention here
to deal with the various methods of producing forecasts. It will be assumed that an estimate
of demand has already been made. It is worthwhile to mention however that the forecast
does influence decisions regarding capacity. This highlights how different topics in the

Page 5
3 Capacity Management

Logistics and Operations module can complement each other within Operations
Management.
Once you have mastered the principles of managing capacity, it will help prepare you for a
modern operations environment, where the planning and control of capacity, as well as
materials is likely to be quite complicated involving thousands of materials as well as products
and hundreds of resources.

Capacity Management Overview

 Capacity is defined as an organisation’s (service/manufacturer) available ability to deliver
their product or service.
 The Objective of Managing Capacity is to balance available resources to the level of
demand.
 Reducing uncertainties will help operations management and allow an efficient execution
of decisions
 Capacity is influenced by the Forecasting and in-turn influences Scheduling.

The following steps outline the key steps for Managing Capacity, with the relevant section of
where each topic will be covered in these notes:

 Strategize and Plan - Section 2

 Determine Level of Requirements - Section 3
 Analyse Current Capacity - Section 4
 Controlling Capacity (Preparing to Allocate Capacity) - Section 5
 Plan for Future Capacity Requirements - Section 6

Page 6
3 Capacity Management

2. Capacity Planning
A typical car contains between 3,000 and 5,000 components and assemblies. A jet engine has
up to 10,000 components and a Boeing 747 has 4.7 million. Whereas a smart phone may only
have a fraction of the mentioned components, nevertheless the manufacturing processes of
these components involves many different types of resources and an array of different
suppliers.
Planning allows us to set the limits of levels of manufacturing operations in the future. More
specifically Capacity Planning is a process which allows a Master Production Schedule (MPS),
or our production plan to be adjusted to balance the due dates of any orders against the
capacity of a plant or a service, (Salvendy, 2007, p. 2042). This allows us to achieve stable
production, which is a very important factor for the overall profitability of an operation.
Due to the vast number of components and products we may have to manage, we cannot
hope to manage this level of complexity without the use of computer systems. The most
widely used computer system for the management of material and capacity is based on a
generic approach called Materials Requirement Planning (MRP).
Figure 3 shows a diagram for an Operations Management system. This is discussed briefly in
the next section before moving on to the detail of Capacity Management.

Figure 3 - A schematic of a system for the management of materials and capacity

Figure 3 shows the typical configuration of such a system. This typically involves three levels
of capacity management as shown:

Page 7
3 Capacity Management

 Resource or Aggregate (Level 1)

 Rough Cut Capacity Planning (Level 2)
 Capacity Requirements Planning (Level 2)

2.1. Resource or Aggregate Capacity Management

This section discusses each of these levels in turn and then briefly discusses the difference
between Material Requirements Planning (MRPI) and Manufacturing Resource Planning
(MRPII).
Resource or Aggregate Capacity Management is the long-term strategic planning task. The
exact time scales depends on the infrastructure of an industry. It deals with long-term
questions such as:

 How long does it take to provide a new production facility, sales team or distribution
network in your industry?
Aggregate capacity planning takes an estimate of the likely market demand for products or
services and seeks to convert this into the required level of resources to meet this demand.
It is approximate, it looks several years into the future and it deals with large buckets of time
(i.e., next year, year after, year after that). If aggregate capacity management is done
effectively, when a company comes to launch a new product or service the required operating
systems and capacity should already be in place.

Industry Example: Aggregate Capacity Planning

In the last decade, car makers identified that some of their international markets saw sales
grow by 40%. Reports from as early as 2010 suggested that Jaguar and Land Rover (JLR) was
strategizing to accommodate for the increasing demands for luxury automobiles for markets
in Brazil, China, Russia and India. These are collectively classified as emerging “power-house”
industrial markets. In order to satisfy the future demands, JLR as well as Mercedes-Benz and
Audi have constructed manufacturing plants in these continents to meet future demand. JLR’s
plant in Itatiaia, which opened in 2016, will have a capacity to produce 24,000 cars per year
in 2016. (Burgess, Rachel, Autocar, 2016)

2.2. Rough-Cut Capacity Planning

The next stage in the planning process involves the creation of a Master Production Schedule
(MPS) for a facility. This major product plan identifies the quantity of each product or variant
to be made in each week or month for the next few months. The time horizon varies for each
facility and depends on the lead-time of raw materials and the internal manufacturing lead-
time. Consider the shortest lead time.
The MPS is crucial for effective planning and control as it acts as the driving force for the
provision of all externally and internally manufactured components and assemblies needed

Page 8
3 Capacity Management

for final assembly. Rough-Cut Capacity Planning should be used to check the validity of the
MPS before it is used to initiate supply of component parts and sub-assemblies.
Rough-Cut Capacity Planning is generally based on a small number of critical or bottleneck
resources, it uses buckets of time, and capacity limits are based on past experience. The
production of the MPS and the Rough-Cut Capacity Plan is an iterative process. Contracts or
quantities are moved or changed in the schedule until resource requirements fall within the
pre-set limits. Once the MPS has been validated and authorised it can be used to drive the
MRP Process.
MRP takes the MPS and works out the total requirements for each of the component parts
that make up the final products based on the Bill of Materials for each product. It works out
the required dates for material by stepping back in time by fixed lead-times for each
component or sub-assembly. Where one component is used on a number of products, it will
also add these quantities together. It is not unusual for MRP to create over 100,000 orders
for purchase or internal manufacture of components.

Module Link: Is the MPS only useful to Capacity Managers?

2.3. Capacity Requirements Planning

Once the individual manufacturing orders have been created by MRP, Capacity Requirements
Planning (CRP) can be used to check the availability of specific resources for specific machines.
The overall validity of the plan (MPS) has already been checked using Rough-Cut Planning.
CRP is about identifying detailed problems in the implementation of the plan and their
solutions. It is much too late and much too complicated to try and change the top level plan
at this stage. CRP is about solutions.

2.4. Material/Manufacturing Requirements Planning (MRPI/II)

This is not the place to enter into a detailed definition of Materials Requirement Planning
(MRPI). However, most people have heard the term at some stage in their engineering
careers. Figure 3 forms a useful way to differentiate between MRPI and MRPII. MRPI was the
early version of the material control system and is shown down the left hand side of Figure 3.
It purely calculates material requirement dates and quantities. Manufacturing Resource
Planning (MRPII) is an enhancement on the original system as it encompasses the hierarchy
of capacity planning modules and is identified down the right hand side of the diagram.

2.5. Capacity Planning Techniques

Planning our Capacity can further help us to execute decisions regarding Scheduling.
Vollmann et al (2010) summarise the 3 types techniques used to plan Capacity.

Capacity Planning using Overall factors (CPOF) – otherwise known as a Rough-Cut Capacity
Plan, Section 2.2. For this method, inputs are sourced from a MPS rather than a MRP.
Workload levels are sourced from performance standards or historical data for end

Page 9
3 Capacity Management

productions. Fabrication times for components which are included in the end item are also
added to the totals. Overall the data is used to derive workload levels. However this method
does not consider the time-shift associated with the lead times.
Capacity Bills – This technique provides a relatively greater linkage between different end
products being produced and the respective capacities required by these different end items
in various work centres. This method is receptive to changes in the product mix of final items
made, additional data however is needed to use this method, for example lot sizes for each
end product and their respective components must be known as well as setup and run times
for each lot should be defined, for each work centre.
Resource Profiles – This third approach refines the capacity bills procedure. It considers lead-
time requirements associated with each node in the parts explosion diagram, all data from
the previous method is used but is further defined to occur in the specific period during which
the work on a specific part/sub/assembly is scheduled to take place. This method is the most
detailed and arguably the most robust of the three mentioned here.

Capacity Planning for Services

When planning capacity in a warehouse planning involves the following aspects. 1) An
assessment of the existing capacity, 2) forecasts of future capacity requirements, 3) an
evaluation of ways to build extra capacity and finally 4) a financial evaluation.
It is further suggested that when developing a long-range plan, a service based firm must
make a basic economic trade-off between the cost of capacity and the opportunity cost of
not having adequate capacity. Capacity costs include both the initial investment into the
facilities as well as the annual cost of maintaining and operating the facilities.
Output measures for service organisations can be difficult to interpret because humans differ
in the rate that they work in comparison to machines (see Learning Curves, Section 3.4).
Therefore input measures are more commonly used for service firms. An example would be
the number of car parking available in the University. As a recommendation: Service
organisations should forecast the human effort involved in providing the necessary services.

2.6. Conclusion
This section has covered some of the popular techniques to manage capacity for
manufacturing and also has shed light into service based organisations. It is advised that you
follow up these techniques with examples from industry. A suitable source to examine, are
the Industrial Engineering Handbooks which cover topics such as Capacity Planning. These are
available in the library in most reference sections.
You will begin to realise as you plan, that you can begin to prepare for eventualities which
may occur when managing capacity. By planning your capacity, you will find it easier to carry
out the scheduling based tasks required for your operations. This relationship between
capacity and scheduling is another example of linkages within LOM. The next step before we
can allocate capacity is to understand the required capacity for a given function, which is
covered in the next section.

Page 10
3 Capacity Management

3. Determination of Required Capacity

Once an estimate of demand has been made, the next stage in the process of drawing up an
operational plan is to determine the required capacity. This involves translating the forecast
requirements for finished goods into the demands that they create for the various "Factors
of Production". Once the required capacity has been calculated we can then go on to
compare this with the available capacity and make decisions about the level of resources to
be provided. As a result this will help us with scheduling and help us to allocate the capacity.
This process is summarised in Figure 4. The calculation of required capacity is also discussed
in this section.

Figure 4 - Hierarchical flow diagram of Managing Capacity

3.1. Factors of Production

The term ‘plant capacity’ has somewhat different meanings which depend upon the type of
industry being considered. In process (or flow) production, or to some extent mass
production, the capacity is expressed in terms of the maximum rate of output that the plant
can achieve. When the output units are non-homogeneous it may be more appropriate to
express the capacity in units of input rather than output. For the service industry it can mean
the maximum level to accommodate customers, whether it is a restaurant, hospital or a
theme park. This section will focus on manufacturing.
For most manufactured products the finished item requires bringing together various sub-
assemblies. In assessing total requirements, the forecasted products are exploded into a Bill
of Materials (BOM). A BOM comprises of a list of parts needed and the number of those parts
in each unit, cascaded into sub-assemblies of sub-assemblies and so on, illustrated in Figure
5.

Page 11
3 Capacity Management

Figure 5 - A typical hierarchical bill of materials (BOM)

The total requirement of each item in the BOM is summed over the entire product range
(some parts are likely to be common to more than one product).

Industry Example: Shared Components and Product Variants

Giving customers a choice in the product/service range is vastly important to securing greater
customer satisfaction and profitability. In order to diversify product portfolios, manufacturers
can create products which to the customer may be seen as a brand new market introduction,
(e.g., a new car) but most likely the vehicle will have shared existing platforms/architecture
from an existing vehicle. One example can be seen with cars made by the VW group. A typical
model such as the Golf will be related to no fewer than 8 other vehicles within the VW family,
with shared components. The VW Group owns brands such as Audi, Bentley, Porsche, SEAT
and Skoda.
Another example can be seen within the computing and mobile telecommunications industry.
In spite of patent related lawsuits over the past few years, Apple is still heavily reliant on
Samsung produced parts. In 2017, Samsung made more money selling components to Apple
for the iPhone X than it did selling its own Galaxy S8 product. (Johnson, L., TechRadar, 2017)
These examples highlight how some components within a product range are likely to be
common for more than one product.

Page 12
3 Capacity Management

For each item the Process Planning Department should have completed an Operation Sheet
or Routing Sheet (Figure 6) which indicates the steps to be followed and resources used at
each stage of production.

Part Number : T2064 Drawing Number : T2064012

Description : Axle Issue Number : 4/2
Where Used : T1014. Assy Located : DD1224
Material: BS970:1971045M10 R/S Number : PP64/20/4
Raw Form : 1.5cm bar Planner : JCB
Raw Matl./Unit : 0.85m length Date : 18 Aug 93
Op M/C Operation Description Gauge Time
No. Tool Loc. SU Run Scrap

100 Saw Cut-off to 83cm length 10 2 Bar end

200 Lathe Feed stock to stop. Turn 50mm LG320 15 2.2 1.5%
length to 10mm dia +/- 0.1

300 “ Reverse Stock and Repeat LG320 - 2.3 1.5%

400 TM2 Thread both ends TG197 15 7

500 Inspect. ( Check threads, length) TG205 1.5 2%

600 Plate Stop off both ends. Cu plate

700 H/T Case harden bearing surfaces

800 Inspect. ( Check Hardness ) Vickers 5 2%

Figure 6 - Operation or Routing Sheet for T2064

The principal resources to be calculated from the information may be summarised under the
three M’s:

 Materials
 Machine Hours
 Man Hours

Most computerised packages for Capacity Planning accumulate the resources by product, and
will then total them across the entire range of products to provide a gross requirement for
the Factors of Production.
The degree of detail provided by these calculations varies with the time-scale over which the
plans are being prepared. For the long to medium-term the process is described as ‘Rough
Cut’ Capacity Planning. This provides an outline plan which considers the block resources
required over a block of time. However this implies that the demands being anticipated may
be met at any time during the planning period. A simple example of the number of machine
hours and man hours required is shown in Figure 7.

Page 13
3 Capacity Management

W ork Centre M 18

M achine M an
Part No Q uantity M aterials Hours Hours

T2064 1000 924m 80.51 40.50

T2095 500 673m 50.25 50.25

B3134 600 833m 24.75 36.25

Total 2430m 155.51 127.0

.
Figure 7 - A Gross Requirements Plan

As the time-scale is reduced (i.e., short-term), the level of detail increases and the capacity
cannot be considered without also taking into account the product-line schedules.
We can calculate capacity requirements based on.

 Demand
 Bill of Materials quantities
 and process times

Additional Factors
There are a number of additional factors which need to be considered. These also consume
capacity and will lead to inaccurate planning unless we make allowances for them in our
calculations. The additional factors include:

 Defective Output
 Efficiency
 Learning Curves
 Machine Utilisation
 Statistical Variations

The next five sections discuss these potential capacity losses. Once these have been
identified, it is possible to compare needed capacity with available capacity and make
decisions about the use of resources. This comparison is called Rough Cut Capacity Planning
and is discussed in Section 3.6. In turn Section 4 discusses possible actions once over-loads

Page 14
3 Capacity Management

or under-loads have been identified. Once these factors have been taken into account, these
issues will help you plan while scheduling.

3.2. Adjustments for Defective Output

Generally it is possible to divide defective output into two parts:
a) Rework
b) Scrap
Items that are suitable for reworking will normally involve extra labour-hours and/or
machine-hours but not additional material. It will be self-evident however, that the resources
required are specific to the product and the form of defect.
When items are completely scrapped the effect is to increase total demand for all of the
factors of production. Whenever it is possible to make reasonable estimates of the level of
scrap, (e.g., from past records), then the necessary adjustments to the production target may
be made very simply:
Scheduled Production = Required Number of Units
100% - Predicted Scrap %

Figure 8 illustrates the allowance for scrap losses over a number of operations.

Op. No. Operation Input Scrap Output

100 Saw 6mtr bars . 1078 round off 1075

take whole number

200 Machine: Self Inspect 1075 1.5% 1059

300 Machine: Self inspect 1059 1.5% 1043

400 Thread 1043 1043

500 Inspect 1043 2.0% 1022

600 Cu Plate 1022 1022

700 Heat Treat 1022 1022

800 Inspect 1022 2.0% 1000

Figure 8 - Adjustment for scrap

3.3. Adjustment for Efficiency

Once the scheduled target has been established, it is possible to calculate the capacity
requirements in units which are related to those used by the planners. Most commonly this

Page 15
3 Capacity Management

information will indicate the time needed to carry out each stage of production in Standard
Hours (or Standard Minutes).
The Standard Time is a measure of the work involved in carrying out an operation, by a
‘standard worker’ working at ‘normal’ speed. However, most workers will deviate from this
by some amount which depends upon their own skills and abilities and upon the payment
and incentive schemes being applied. Once again, the target figure will need to be amended
to take account of the predicted difference between the standard and the typical level of
performance. Figure 9 shows a typical job ticket and the calculations for operator efficiency.

Works Order No. JHW 199/95 Works Centre M18

Part Number : T2064 Drawing Number : T2064012

Description : Axle Issue Number : 4/2
Where Used : T1014. Assy Located : DD1224
Material: BS970 : 1971045M10 Raw Matl./Unit : 0.85m
Raw Form : 1.5cm bar Total Raw Matl : 924m
Stores Loc. No. : AS 10/5 Return to Stores: nil
Qty Reqd 1043 Qty Produced 1050
Date Reqd 09/1 Scrap 28
Date Delivered 10/2 Received (sign)
Op M/C Operation Description Gauge Time
No. Tool Loc. SU Run Scrap

200 Lathe Feed stock to stop. Turn 50mm LG320 15 2.2 1.5%
length to 10mm dia +/- 0.1

300 “ Reverse Stock and Repeat LG320 - 2.3 1.5%

Machine Hours Planned 80.51 Man Hours Planned 40.5

Machine Hours Actual 76.75 Man Hours Actual 38.75

NOTE 1: Planned Hours are calculated on the basis of ONE operator

running two machines
NOTE 2: Operator Efficiency = 40.50 / 38.75 = 105 %

Figure 9 - Shop order or job ticket

Page 16
3 Capacity Management

3.4. Learning Curves

It is worth pointing out that the performance of an operator, or of a group of operators, is
rarely constant, it is likely that the actual time taken to perform an operation will decrease
with experience and practice, following a curve of the general form of Figure 10.

T n = T 1n x

Production
Time/Unit
(T n )

No. of Units Produced (n)

Figure 10 - General Form of the Learning Curve

Typically this change in performance is represented by the relationship:

x
Tn = T1n where Tn = time for cycle n
T1 = time for first cycle
n = cycle number
x = a constant

Again, experience shows that x is frequently of the form:

x = 1n c/1n 2

Where c, the Learning Constant, takes a value between 0.5 for complex or difficult tasks and
1.0 for simple operations.
Figure 11 highlights a practical example of a Learning Curve from the rapidly developing
battery industry. This particular example captures both technology learning and expected
productivity learning in manufacturing the battery.

Page 17
3 Capacity Management

Figure 11 - An example of a learning curve present in lithium-ion battery production, (Clamp, 2010)

In any event some allowance should be made for the changing capacity requirements over
the life of the product. The concept of the Learning Curve being borne in mind when
calculating:

Capacity Required = No. of Units x Standard Time/Unit

Operator Efficiency

In practice the form of the Learning Curve is likely to depend upon a number of factors, each
of which will modify the value of the Learning Constant. This subject is explored in more detail
in the Appendix ‘Learning in Production’ which will look at the aircraft manufacturing industry
in greater detail.

Page 18
3 Capacity Management

3.5. Machine Utilisation

In setting out to determine the real capacity of a manufacturing facility, the availability of
machines, labour and jobs must be taken into account. Interruptions to production may occur
because of
 Absenteeism

 Machine breakdowns

 Tool shortages

These are but a few examples. These might well be expected in the natural scheme of things
since no machine is proof against all forms of failure and no operator is certain of avoiding
illness and injury in all circumstances.
However, operations can be halted with starvation arising from inadequate scheduling and/or
insufficient buffer stocks. (Module linkage of capacity, scheduling and Inventory). In addition,
time is lost due to tool changing, setting and adjusting. The minimising of these losses is the
responsibility of production management and the losses substantial if the systems of control
are inappropriate.

Table 1 shows the results of one survey carried out into machine utilisation and it is perhaps
surprising to observe not only how low the percentages are of productive time, but also how
consistent these values are across different sectors of industry.

Table 2 shows the same kind of analysis for the utilisation of labour and again it is interesting
to note the similarity in the two sets of figures. What emerges most clearly is the very high
proportion of time that resources are standing idle with the Management Responsible
indicating that lack of co-ordination and managerial control is a major cause of waste. In
fairness it should be said that the complexity in managing multi-tasking, multi-facility systems
is far greater than is recognised by most people, even including those responsible for
attempting it.

Page 19
3 Capacity Management

Iron and Electrical Non-ferrous Motor Vehicle West

Steel components Midlands
sample

Operating 43% 50% 45% 55% 48%

Attending 8% 6% 16% 10% 11.5%

Handling 24% 23% 18.5% 15.5% 19.5%

Servicing 4.5% 4.5% 5% 4.5% 5.5%

Waiting : 10.5% 5% 6% 4.5% 5.5%

Management
responsible

Waiting: Operator 10% 11.5% 9.5% 10.5% 10.5%

responsible

Table 1 - Utilisation of labour showing the division of productive and non-productive time during a working day

Iron and Electrical Non-ferrous Motor Vehicle West

Steel components Midlands
sample

Productive 42% 30% 45% 45% 41%

Setting 7.5% 4% 2.5% 3% 4.5%

Tool Adjustment 3.5% 1.5% 1% 2.5% 2%

Maintenance 3.5% 1% 3% 1.5% 2%

Idle: 36% 59% 44% 41% 45%

Management
responsible

Idle: Operator 7.5% 4.5% 3.5% 7% 5.5%

responsible

Table 2 - Utilisation of machines showing the division of productive and non-productive time during a working day

Page 20
3 Capacity Management

3.6. Statistical Variations

The design of an operations facility is often based on a simple calculation of the amount of
each type of resource needed in order to meet an average level of demand. In fact if a product
has to be processed in a number of resources, we pride ourselves on our ability to "balance"
the capacity throughout the facility so that levels of utilisation on each process will be
uniformly high.
Both demand and processing time will naturally vary about a mean. When the processing
time on a resource is higher than the mean the resource will be limited by capacity and queues
of work will build up in front of that process. This approach results in each resource becoming
a bottleneck. Conversely, when processing time is lower than the mean a queue will reduce
in length. While we only have one process, these variations cancel each other out and we
achieve the required mean level of output. However when we combine a number of
processes we get a different effect. Higher than normal process times mean reduced output,
lower than normal process times do not redress the balance because processes are restricted
from operating faster by the limited flow of work from the preceding work centres.
Statistically expected fluctuations in process times or demand therefore reduce the possible
output level from a facility.
Eliahu Goldratt illustrates this effect well by discussing a troop of boy scouts walking through
a wood in a line, (Goldratt, 2004, pp. 94-102). Consider cars travelling down a motorway at
different speeds. Queues (bottlenecks) form and disappear all the time as cars speed up,
cruise and slow down at varying rates.

3.7. Rough Cut Capacity Planning

Because of all the factors discussed above, it is usual to produce a Rough Cut Capacity Plan
that only consumes 70 to 85% of the available capacity. This is a rule of thumb and the actual
percentages that apply in each organisation will have to be calculated based from historical
performance.
It is unlikely that Rough Cut Capacity Plans will be produced for every work-centre in a facility.
Plans are normally based on groupings of like machines or on certain critical resources or on
bottleneck resources. These are relatively easy to identify based on previous experience.
Because bottlenecks change as product mix changes, it is wise to review the resources used
for rough cut capacity planning on a regular basis.

Page 21
3 Capacity Management

4. Capacity Management Strategy

Once the required level of capacity has been calculated a company can decide the appropriate
level of resources to provide. In all real operations facilities it is almost certain that the
demand for products or services will vary due to:

 Seasonal or other systematic effects

 Random effects

In determining the level of capacity required to meet demand it is important to understand

the influence of this variation. In brief the level which gives the optimum cost is a balance
between:
a) The cost of unused capacity
b) The cost of ‘lost’ or late production

This trade-off is shown in Figure 12. The optimum point is affected significantly by variation,
in that high values of demand will increase both the total cost and the level of capacity at
which the optimum occurs. This is illustrated in Figure 13, it also shows the fact that the
optimum level of capacity, calculated on the basis of unused resources versus unsatisfied
custom, is strongly influenced by the variation in demand.

DEMAND
COST Capacity less Capacity greater
than Demand than Demand

Cost of Cost of
lost or late Unused
production Capacity

CAPACITY

Figure 12 - The build-up of capacity related cost

Page 22
3 Capacity Management

Demand

Time

Increasing variability in
demand increases the
optimum size of plant
COST and the operating cost

MEAN
DEMAND

CAPACITY

Figure 13 - The impact of variability in demand on optimum capacity

Page 23
3 Capacity Management

In any event the approach to satisfying the demand for resources usually involves some
compromise between two options. These are:
4.1 Provide for efficient adjustment or variation of capacity
4.2 Eliminate the need for adjustments.

4.1. Strategy 1: Provide for efficient adjustment or variation in capacity

Table 3 presents practical means for capacity adjustment when there is a need to make
capacity easy to vary.

Resource To Increase Capacity To Decrease Capacity

Subcontract excess load Retrieve sub-contracted work

All
Buy rather than Make Make rather than Buy

Reduce material content

Substitute more readily available

Rearrange priorities
Material materials
Revise delivery schedules
Rearrange priorities

Revise delivery schedules

Revise Work Schedules Revise Work Schedules

(long runs/big batches) (short runs/small batches)

Machines
Revise maintenance schedules (defer, Revise maintenance schedules (advance,
weekends etc.) day shift working etc.)

Increase overtime Decrease overtime

Manpower Annualised Hours Annualised Hours

Hire Fire

Table 3 - Practical means for adopting strategy 1

4.2. Strategy 2: Eliminate the need for Adjustments

Complex, highly specialised activities or operations which employ large quantities of
resources may make it difficult to provide temporary alterations in capacity. In such cases it

Page 24
3 Capacity Management

might be desirable to minimise the need for such adjustments by the use of some or all of the
following:

 Maintain excess capacity

 Accept loss of custom
 Require customer waiting or queuing
 Hold (output) stocks.
These options and their relationships with the adjustment of capacity are considered for
different types of production systems in Table 4. The apparent choice between providing
capacity for maximum demand or for average demand is not as simple as it appears.
Very few manufacturing companies would choose to provide ‘peak’ capacity since this
involves resources standing idle for a very large proportion of the time. However, in a period
of recession the peak demand might well fall below the capacity already installed. Even in
circumstances of growth or stable demand the meaning of the term ‘average’ must be
interpreted with care.

Strategy (A) (B) Reduce/Smooth Fluctuations

(2) Use Stock to

Maintain Excess (1) Fix Upper Limit on Capacity
Form of Production Absorb

Capacity Loss of Business Customer waiting Fluctuations

S>V>P>V>C Feasible, rarely Feasible.

Feasible. Sometimes necessary
(Raw and FG Stock) necessary Normally used

S>V>P>C
Feasible, may be
(Raw Material used with B(1) Feasible. Normally used Not feasible
Stock Only)

S>P>V>C
Feasible. Often
(Finished Goods Unlikely Unlikely to be used
necessary
Stock Only)

S>P>C Sometimes
Unlikely to be used Not feasible
(No Inventory) necessary

Table 4 - Strategy 2, eliminating the need for adjustments

Key: S = Supplier
P = Production
C = Customer
V = Storage

Page 25
3 Capacity Management

4.3. Evaluating Plans

In most cases, when an appropriate stratagem has been decided upon, there will be a number
of alternative routes by which it may be achieved. A major criterion to be used in deciding
between the alternative routes is the value (or ‘utility’) of each. This may be taken into
account in a number of different ways such as:

 Pay-back (or Amortisation) Method

 Accounting Rate-of-Return Method
 Discounted Cash Flow Method

Of these alternatives just one will be considered here in order to illustrate the approach to be
adopted.

Discounted Cash Flow

Discounted Cash Flow (DCF) is a method of evaluation that takes account of the way that
money changes in value with the passage of time. Since World War II it has been common
for the U.K. economy (in common with most others) to involve a positive rate of inflation.
This means that costs have increased or that the purchasing power of the unit of currency has
declined progressively. On the other hand, banks and financial institutions used to offer rates
of interest that are in line with the rate of decline or approximately so.
This means that in many cases it may be preferable to delay expenditure to take advantage
of the decline. Since this decline is in effect the counterpart of the gains in investments
associated with the compound interest factor it should not be surprising to find that
expenditure at some time in years in the future can be discounted to find its present value by
use of the formula:

PV = 1
____________________

n
( 1+ i ) where i = is the rate of inflation
PV = is the present value factor

Page 26
3 Capacity Management

5. The Control of Capacity

5.1. Introduction
There are some simple but effective, quantitative aids to planning and control of capacity that
are well worth considering. In the short term many of the difficulties that arise are associated
with the fluctuations in demand which are a normal part of the life of an operations planner.
They may arise due to seasonal changes in demand (for example: demand for new cars in te
UK, peaks 2x per year when the registration plates change; electric blankets sell in Winter,
barbecues and lawn-mowers in Summer), due to customers building up or running down their
stocks and often due to financial arrangements (goods delivered at the beginning of a month
are usually paid for at the end of the next month).
Whatever steps have been taken already in order to minimise the impact of fluctuations in
demand, it will almost certainly be necessary to make some adjustments in order to exert the
optimum control over the utilisation of system resources. Just two of the simplest techniques
will be considered here in order to show how some problems may be approached. Essentially
the control of capacity will allow you to execute your scheduling efficiently.

5.2. Cumulative Graphs

The use of a cumulative demand graph is considered. In Figure 14 the monthly demand for a
product is shown for the twelve months of a year and it can be seen that there are some
significant changes month on month. If the data are re-plotted as cumulative monthly
demand values (i.e. as a rolling total) then, as shown in Figure 15 the fluctuations are
smoothed to some extent. Moreover the average rate of demand is indicated by the slope of
the curve between any two, arbitrarily selected points. This makes it possible to superimpose
some steady rate of production on the demand curve that would achieve a balance between
the requirements of the customer(s) and the needs of the production facility. In the example
shown a steady production rate of say 6,000 units per month would mean that some surplus
stock of finished goods would be held until the end of April. Thereafter a rate of 5,375 per
month would lead to a small under-delivery in August and November, but would meet the
forecast demand with very little overstocking during the rest of the year.
The alternative plans prepared by this technique should be drawn up with due consideration
to the Cumulative Cost of Production.

Page 27
3 Capacity Management

cumulative 12 period forecast

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 14 - A typical demand profile with cyclical/seasonal variation and a trend

Forecast by period

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 15 - The cumulative sum graph for the demand pattern shown above

5.3. Cumulative Cost of Production

In most manufacturing systems the fixed and semi-fixed costs represent a very high
proportion of the total manufacturing cost for the first few products made. Indeed, the level
of production at which the ‘break even’ point occurs between income and expenditure is
typically 60% to 80% of capacity. This is illustrated in Figure 16. The figure may look complex
initially, but is quite straightforward. At any vertical point on the chart where the red lined is
to the right of the blue line, profit is beng generated. It can be seen that there is no financial
justification for overtime working or extra shifts to satisfy customer demand which is only just
above the full capacity of present operations.

Page 28
3 Capacity Management

First Shift Second Shift

Capacity Capacity Revenue

Total
Costs

Value
(£’s)

Semi-fixed Costs (2nd Shift)

Variable Costs (1st Shift)

Semi-fixed Costs (1st Shift)

Fixed Costs

Production (units)
Figure 16 - Break-even analysis

5.4. Linear Programming

In the long history of optimisation models for production planning, many if not all of them are
based on Linear Programming, (Salvendy, 2007).
Linear Programming makes it possible to determine the optimum value of a Linear Objective
Function which is subject to a series of Linear Constraints. The approach may be applied to
problems of capacity planning provided that the production system can be described in terms
of the appropriate linear equations (and/or inequalities). When there are just two variables
to deal with the L-P problem may be solved by graphical methods, and since this is easier to
follow the example shown is approached this way. However, when there are more than two
variables it is necessary to use some numerical method such as the Simplex Algorithm. In
effect, this method traces a path around the boundaries of the feasible space, locating each
intersection of the linear constraints and solving for the values of the control variables at this
point. Since the optimum point will always lie at the intersection of two (or more) constraints
this eventually leads to the solution of the problem (Figure 17).
The application of L-P to the classical ‘Product-Mix’ problem is illustrated, but it must be
appreciated that the simplified form of the system shown is rarely met with in practice.

Page 29
3 Capacity Management

Practical problems may require careful manipulation to make them amenable to solution by
this approach.

Example Linear Programming Problem

A manufacturing plant makes two products A and B. The production process involves buying-
in a casting which is machined, drilled and polished. The running costs of the machines being
shown together with nominal, hourly capacities in Table 5:

------------------------------------------------------------------------
Hourly Capacity A B £/hr
------------------------------------------------------------------------
Machining 25 40 20.00
Drilling 28 35 14.00
Polishing 35 25 17.50
-------------------------------------------------------------------------
Table 5 - Hourly capacities for product A and B, with associated hourly costs

Profits from the sales of A and B are £1.20 each and £1.40 each respectively.

If any combination of products can be sold, what product mix will maximise
profits?

Product B The Three Constraints

40
40A + 25B = 1000
35
35A + 28B = 980
30
25A + 35B = 875
25

20

15

10

0
0 5 10 15 20 25 30 35

Product A

Page 30
3 Capacity Management

The Three Constraint Functions

The Objective Function :-

Product B Profit = 1.2A + 1.4B

40

35

30

The Solution
25 16.9 Product A per hour
12.9 Product B per hour
20 Profit of £ 38.39 per hour

15

10

The
5
Objective
Function
0
0 5 10 15 20 25 30 35
Product A

The Valid Solution Area and Objective Function

Figure 17 - Linear programming solution

5.5. Duality
In the example considered, an optimum product-mix is obtained by fully utilising the capacity
available on two of the three production processes.

 Machining (M) - fully utilised

 Polishing (P) - fully utilised
 Drilling (D) - spare capacity

Imagine a situation in which another firm offers to rent time on the three processes, then the
total income which would be earned can be shown as:
£M + £D = £P = Total Rent (1)
Clearly, it would be unreasonable to make time available to the other firm unless the income
obtained was at least equal to that already available. Therefore this imposes constraints of
the form:

Page 31
3 Capacity Management

M/25 + D/28 + P/35 > £i.20 (2)

and M/40 + D/35 + P/25 > £i.40 (3)

But the potential customer will seek to rent the process capacity for as little as possible. The
point of compromise will be found by solving the L-P problem:
Minimise Equation (1)
Subject to Equations (2) and (3)

This is referred to as the DUAL of the original problem. In fact its solution will be found at the
same point as that of the PRIMAL problem, e.g.
O.F. = £38.39
M = £ 9.03
P = £29.36

These figures indicate the value of the resources at the optimum to the manufacturing firm,
and are known as the “Shadow Prices”. They represent the contribution to income made by
the particular resource at the margin, i.e. the increase that would arise if the process were
run for one extra hour.
If the price of buying extra capacity from outside from some sub-contractor for example, then
it would be advantageous to buy time on Polishing since this would yield:
£29.36 - £17.50 = £11.86 profit per hour
On the other hand the Machining stage would cost £20.00 per hour but only increase income
by £9.03 per hour and so would result in a reduction in profit. There is no shadow price for
the Drill since it is not fully used at the optimum and any further income is profit.
Clearly the additional benefit of buying polishing capacity is not infinite. However, one
method known as the Simplex method is easily extended to allow the limit of this expansion
of profit to be calculated. The Simplex method will be covered in the next section. The method
of finding the shadow prices and the limits of expansion come within the general technique
of ‘Sensitivity Analysis’.
It is worth adding that with sufficient expertise it is possible to deal with Objective Functions
and Constraints which are not linear, and variables that can only take integer values.
However, this is outside the scope of the example mentioned in the next section and it is
recommended that you learn about it from the available textbooks. Another example of linear
programming will be covered in the appendix.

Page 32
3 Capacity Management

Simplex method
Elementary transforms
1. Any (complete) row may be exchanged with any other (complete) row.
2. Any row may be multiplied (or divided) by a scalar.
3. Any multiple of one row may be subtracted from (or added to) the remaining rows.
Optimisation
1. Introduce slack variables to convert inequalities into equations.
2. Set up tableau including Basis Variables and Objective Function.
P
3. Identify variable giving most rapid improvement in O.F. i.e. Pivot Column ( pivot).
4. Determine how far the Pivot variable can be improved without violating any
P
constraint, i.e., select lowest Po/ pivot - (Pivot Row).

5. Identify Pivot Element.

6. Divide Pivot Row by Pivot Element and enter as first row in new tableau.
7. Subtract an appropriate multiple of the Pivot Row from each of the other rows in order
to eliminate Pivot variable and enter in the new tableau (this includes the O.F.)
8. Check all the coefficients in the O.F. If any value is positive return to 3 above,
otherwise the solution is a maximum.

Page 33
3 Capacity Management

Basic Pa Pb Pu Pv Pw Po Po/
Var Ppivot
u 1/25 1/40 1 0 0 1 40
v 1/28 1/35 0 1 0 1 35
w 1/35 1/25 0 0 1 1 25 <Pivot Row
Z 1.2 1.4 0 0 0 0
^
b 25/35 1 0 0 25 25 35
u .0221 0 1 0 -.625 .375 16.9 <Pivot Row
4
v .0153 0 0 -.7143 .2857 18.7
0
Z 0.2 0 0 0 -35 -35
^
a 1 0 45.17 0 -28.23 16.93
5
b 0 1 -32.26 0 45.16 12.90
3
v 0 0 -0.69 -.2822 0.026
5
Z 0 0 -9.03 0 -29.35 -38.39

Table 6 - Using the Simplex Method

All the coefficients in the final row are negative or zero therefore the solution is an optimum.

Page 34
3 Capacity Management

6. Summary
The final step when managing capacity is to plan ahead for any future trends. Your plans may
accommodate your present requirements but will they continue to do so in the future?
Therefore, it is important to revisit Step 1 and close the loop between your forecast,
operational demands, and scheduling (Figure 4).
To conclude, we present key steps extracted from these notes. It is recommended that you
follow up these notes by using examples from industry and academic literature. The further
reading section provides you with a useful starting point towards fine-tuning your practical
know-how of Managing Capacity.

Capacity Management Summary: Key Steps

1. Strategize and Plan
 MRP/MRPII. Capacity Planning Techniques.
2. Determine Level of Requirements
 Determine standard unit of work, (e.g., labour/machine hours needed).
3. Analyse Current Capacity
 Take into account scrap, inefficiencies and breakdowns.
 Do not neglect statistical variances and learning curves.
4. Controlling Capacity (Preparing to allocate Capacity)
 Consider Linear Programming and optimise for Scheduling.
5. Plan for Future Capacity Requirements
 Think ahead, forecast and plan for any future requirements.

Page 35
3 Capacity Management

7. Seminar Questions (Full-Time Students)

The following questions are for you to test your understanding of this subject area:

1. Why do we need capacity management tools and strategies at all?

2. What measures can we use for capacity?
3. Explain the three levels of capacity management and their different functions.
4. Discuss the relationship between capacity management and effective delivery
performance in a make to order and a make to stock environment
5. Why are bottleneck resources so important in capacity management?
6. Explain the difference between the two major strategies for capacity management.
7. Think of two capacity management illustrations from everyday life. Discuss the trade-
offs in defining the ideal level of capacity in each illustration.
8. How would you evaluate a number of alternative capacity management strategies?
9. What happens to output if we use overtime for an extended period of time? How
does this differ between a manual assembly environment and a driven
assembly line?
10. Generate a cumulative sum graph from the following demand data. What
production rates are appropriate? When might the rate change?

Week Demand Cumulative

Demand
1 30 30
2 60 90
3 30 120
4 80 200
5 100 300
6 70 370
7 110 480
8 130 610
9 60 670
10 130 800

Page 36
3 Capacity Management

8. Appendix: Learning in production

Extracts from Paper: British Aircraft Corporation Ltd. Supplemented with information from Handbook of
Industrial Engineering, The learning Curve (Hancock, 1971). Figures from ‘Applying learning curves in aircraft
production’, (Jones, 2001).

8.1. Introduction
Logistics and Operations Management involves different topics, such as Psychology, Human
Factors as well as Marketing and Engineering Management. Predicting the rate at which a
person can learn a manual task is a complex process, why? Because every human, differ in
their ability to carry out tasks as well as their learning capabilities. We will not go into the
individual differences in this appendix but it is worth mentioning that as individuals begin to
repeat certain actions in manufacturing, time and time again their performance can be said
to improve and they be learning the task by reinforcement, (Hancock, 1971). Industries in
which learning effects have been observed, studied and reported include the following:

 Aircraft
 Shipbuilding
 Automobiles
 Textiles
 Musical Instruments
 Steel
 Glass Products
 Electric Motors
 Electronics
 Cigars
 Printing and Typesetting
 Machine Tool Manufacture
 Oil Refining
 Electric Power Generation
 Joinery
 Hosiery
 Footwear
 Iron Castings

The diverse nature of the industries listed above leads on to a discussion of the scope of the
present paper and the interpretation to be placed on the term "learning". It will have been
noticed I expect that some of the industries mentioned above are in the nature of process
industries or have a continuous flow of production, e.g., yards of cloth, or barrels of oil. Papers
on these and similar industries generally deal with "industry learning" over a period of many
years or even decades and necessarily select arbitrary, if familiar, units for measurement of
output (the oil might as well be measured in gallons or the cloth in inches or miles). The start
date and hence the initial unit input also is necessarily somewhat arbitrary in this type of
study.

Page 37
3 Capacity Management

It is intended in this paper, to discuss learning in the context of a product in which the unit is
determinate. Some writers seek to restrict the term "learning" only to improved dexterity on
the part of operators and to exclude organisational and environmental factors. It is not
intended to accept such a restrictive definition in this paper. In fact part of the paper is
concerned with the relative importance of the different constituents of "earning" defined as
the reduction in total man hours arising from all causes whether operator skill or
organisational.

8.2. The Learning a Curve Defined

The form of the learning curve with which we have had most experience is defined by the
equation:
b
Mn = Ml n where Mn= man hours to build the nth unit

Ml = man hours to build the first unit

n = number of units
b = an index deflating the "slope" of the curve
When plotted on log-log paper this becomes a straight line with slope "b". This line has the
property that the man hours decrease by a fixed percentage each time the number of units
made is doubled. Thus if the index "b" (regression co-efficient, Table A and Table B) has the
value -0.322 then each time the quantity built is doubled the unit man hours is reduced to
80% and the curve is referred to as "an 80% learning curve". Thus the 2nd unit takes 80% of
the man hours for the first unit and the 200th unit takes 80% of the labour hours for the 100th
unit. The figure below shows an example of 75% and 80% learning.

Figure 18 Learning curves, (Jones, 2001)

Page 38
3 Capacity Management

8.3. Analytical Studies of Production Learning

The analytical part of this paper is devoted to describing two specific studies designed to
improve understanding of the causes of production learning. Some of the questions which it
was hoped to answer were:
1) How much of learning is due to operator skills and how much is due to organisation?
2) Do the skills appropriate to a particular job reside only in the individual operators
or, once learnt, are they readily transmitted to other operators?
3) Does the reduction in man-hour content as learning takes place arise within the
"useful" content of the job, i.e. the actual cutting or forming, or in the extraneous
activities such as procuring material and tools, and dealing with queries, difficulties
or shortages?
4) What is the influence of payment by results schemes?

8.4. Transfer of Learning between Parallel Production Facilities

The first subject was the transfer of learning between parallel production lines; this is a
controversial subject, there is always conjecture about how much learning experience can be
transferred and how much has to be acquired on the new production line.
The data consisted merely of unit man hours, build sequence and in which of the five jigs did
construction take place. The item was a major component of a subsonic aircraft requiring an
assembly effort of about 6,000 man hours towards the end of the period covered. The object
of the study was to examine the effect of progressively opening new jigs with a view to
establishing the existence of transfer of learning, or of interaction on existing jigs.
The first stage of regression analysts was to perform a simple regression of unit man hours on
(a) Set Number (SN) and (b) Number In Jig (NIJ)
Logarithmic transformations were used to achieve linearity, the transformed variables being
designated LMH, LSN, etc. The set number is the number in chronological sequence in the
total production while the NIJ is the number in sequence in the individual jig. Thus the first
assembly in the fifth jig was the twenty-fifth in total production sequence and would have an
SN value of 25 and an NIJ of 1.
The second explanatory variable adopted was Number Of Jigs (NOJ): this started at the value
1 and rose to a maximum of 5 as the jigs were progressively opened. It might appear to be a
rather crude variable but in fact it is quite a plausible general indicator of the dilution of
operator experience as the jigs are opened, regardless of whether it is the new jigs which are
manned with new operators or if there is some depletion of the existing jigs.
A single variable, the Operator Repetition Number (ORN) was constructed. This was the
average number of sets experience of the operators engaged on a particular set, on the
following assumptions. "When a new jig is started it is manned 50% by new operators and
50% by operators drawn from existing jigs, with the proviso that no jig may lose more than

Page 39
3 Capacity Management

50% of its existing manning between consecutive sets on that jig. The experienced operators
withdrawn are replaced by new operators". These results are summarised in Table A.
The conclusions from the study were that (a) there was some transfer of learning between
jigs, (b) it was likely that operator learning was a major factor, (c) because of the likely physical
transfer of experienced operators on to new jigs it was not possible to isolate system from
operator learning.

8.5. Detailed Study of Learning Influences

Data
Some detailed information was available on ten individual assembly tasks which were part of
the assembly of a major component of a transport aircraft. The information extended over a
run of more than 100 aircraft and included the following items:
1. Number in build sequence
2. Man hours for task
3. Identification of individual operator(s) involved
4. Set number at which allowed time was confirmed
5. Additional time allowances and reasons therefore

From this information the following parameters were set up:

MH Man hours
SN Set number
ORN Operator Repetition Number
RSP Rating Step
RSL Rating Slope
EA Extra Allowances
OFN Operator forgetting number
N Number of operators

The man hours, number of operators and set numbers are self-explanatory, the operator
repetition (ORN) is the number of sets to date by the individual operator, RST and RSL are
dummy variables which enable a step and/or a slope change to be incorporated in the
relationship, particularly at the point where the allowed time is agreed. Extra Allowances (EA)
is the extra time allowed for the task arising from various circumstances outside the
operator's control. The Operator Forgetting Number (OFN) is the number of intervening sets
since the operator was last involved.
Influenced strongly by the form of the long established simple learning curve model, a model
of the following form was adopted.

a b d f 9MH h
MH = k x SN x ORN x RST x RSL x e x OFN

Page 40
3 Capacity Management

Where a to k are constants and is a residual. A logarithmic transformation was used to enable
a linear regression analysis to be applied. RST takes the value I up to the set at which the time
allowance is agreed (SNR) and 2 thereafter. RSL take the value SNR up to the set at which time
is agreed and SN thereafter. The functional form of the EA term perhaps requires some
explanation. It had been included in a form homogeneous with the other terms and while it
had given a satisfactory result mathematically it was appreciated that the functional
relationship was less than suitable from physical considerations.
The operation of a premium bonus scheme is such that one would expect a linear relationship
between MN and EA. The particular function adopted is an attempt to introduce the required
additive relationship into a multiplicative model.
Please read Jones (2001), for an in-depth understanding of the various models which can be
used to depict the learning curve. Additionally, you may wish to learn about regression
analysis to be able to interpret the results.

8.6. The Analysis and Results

Multiple regression analysis was performed on various combinations of the variables using
MH as the dependent variable. The Analysis was done on all ten tasks simultaneously with
723 data sets. The different tasks were of course of different man hour content so scaling
factors were introduced for nine of the tasks and included in the final regression with all the
parameters present so that they were automatically determined. These factors were then
applied to the man hours of each task and the regressions re-run. Some of the more important
of the results obtained are shown in Table B.

8.7. Implications of Results

This extremely demanding study has been justified by the results obtained. Probably for the
first time some understanding of the relative importance of different contributions to the
learning process in aircraft assembly has been achieved. Probably the most important result
is the establishment of the influence of the individual operators experience as being
substantial but not over-riding.
The influence of the manipulation of the incentive scheme has been measured and is of some
consequence but not the major effect postulated by some observers. The effect of
shortcomings in design and production at the commencement of the manufacturing run has
been established. An attempt has been made to summarise the results visually in the figure
below.

Page 41
3 Capacity Management

Figure 19 - Summary of results, (Jones, 2001)

In terms of the cumulative man hours to 100 sets, one has to distinguish between the actual
achievement with the inevitable movement of operators and the potential achievement if
they could be constant. From this figure one can see a partial explanation of the excess man
hours due to learning and hence the possible improvements in man hour performance which
would result as action could be taken to influence the causes.

8.8. Qualitative Assessments

In some instances you may be able to make assessments based on subjective opinions and
not use statistics. For example, the figure shown below highlights that the return of
investments (ROI) beyond the the ¾ quantity point on the learning curve diminish rapidly.
One interpretation from the figure would suggest that any investment required beyond the
¾ point will less likely provide an acceptable ROI (Jones, 2001).

Page 42
3 Capacity Management

Figure 20 - Relative labour savings (Jones, 2001)

TABLE A: Transfer of Learning, Regression Results

Constant Correlation Regression Learning

Parameter Term Coefficient Coefficient Curve

LSN 10.355 0.913 -0.343 79%

LNIJ 9.90 0.865 -0.372 77%

LSN -0.468 72%

LNOJ 9.95 0.943 +0.576 -

LORN 10.12 0.939 -0.460 73%

LORN 10.10 0.939 -0.493 71%

LSN +0.026 -

Page 43
3 Capacity Management

TABLE B: Detailed Study of Learning Influences

Results of Regression Analysis
Regression Constant Correlation Interpretation
Parameters Coefficient Term Coefficient (Learning slopes etc.)
SN -0.354 5.578 0.792 78%

SN -0.241 5.596 0.831 84%*

+
ORN -0.159 76%

SN 86%*
+
ORN -0.125 5.383 0.851 79%
RSL +0.078
RST -0.450 27% drop

+
SN -0.184 80%
ORN -0.137 5.287 0.857
EA 0.287 250% bonus

SN -0.225 79%
ORN -0.116
ES 0.247 5.088 0.897 300% bonus
RST -0.253 16% drop
RSL 0.122

+
SN -0.232 79%
ORN -0.111
EA 0.247 300% bonus
RST -0.249 5.097 0.868 16% drop
RSL 0.120
OFN 0.027 6% increase after 10
sets gap
* "One Set" operators

+ constant operators

All coefficients significant at 95% confidence level

Page 44
3 Capacity Management

References
Adenso-Diaz, B., Gonzalez-Torre, P. & Garcia, V., 2002. A capacity management model in
service industries. International Journal of Service Industry Management; Bradford, 13(3/4
(Available via ABI Inform at http://webcat.warwick.ac.uk/record=e1000009~S1)), pp. 286-
302.
Burgess, Rachel, Autocar, 2016. Jaguar Land Rover factory opens in Brazil. [Online]
Available at: https://www.autocar.co.uk/car-news/industry/jaguar-land-rover-factory-
opens-brazil
[Accessed 28 September 2018].
Clamp, S., 2010. Lithium Air, Batteries, Electric Engine drive trains:why the automotive
industry may change forever. [Online]
Available at:
https://cleaninvest.wordpress.com/2010/02/16/lithium_air_batteries_and_electric_engine/
[Accessed 28 September 2018].
Goldratt, E. M., 2004. The goal : a process of ongoing improvement. 3rd, available at
http://encore.lib.warwick.ac.uk/iii/encore/record/C__Rb3037272 ed. London: Routledge.
Hancock, W., 1971. The learning curve (in Industrial Engineering Handbook by Maynard). New
York: McGraw-Hill.
Johnson, L., TechRadar, 2017. Samsung stands to make more from iPhone X than Galaxy S8
thanks to parts. [Online]
Available at: https://www.techradar.com/uk/news/samsung-stands-to-make-more-from-
iphone-x-than-galaxy-s8-thanks-to-parts
[Accessed 29 September 2018].
Jones, A. R., 2001. Case study: applying learning curves in aircraft production-procedures and
experiences, in Zandin K Maynard's Industrial Engineering Handbook. New York: McGraw Hill.
Salvendy, G., 2007. Handbook of Industrial Engineering: Technology and Operations
Management. 3rd, available via
http://encore.lib.warwick.ac.uk/iii/encore/record/C__Rb2579545 ed. New York: John Wiley
& Sons, Inc..
Slack, N., Brandon-Jones, A. & Johnston, R., 2016. Operations Management. 8th (available as
e-book via http://encore.lib.warwick.ac.uk/iii/encore/record/C__Rb3022525) ed. Harlow:
Pearson.
Vollmann, T., Berry & Whybark, D., 2010. Manufacturing Planning and Control Systems for
Supply Chain Management. 6th ed. New York: Irwin/McGraw-Hill.

Page 45