This article was downloaded by: [University of Sunderland]

On: 29 December 2014, At: 19:31

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK

Production Planning & Control: The Management

of Operations
Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/tppc20

A multi-dimensional classification of production

systems for the design and selection of
production planning and control systems
a

Bart L. Maccarthy & Flavio C. F. Fernandes

Division of Maufacturing Engineering and Operations Management , University of

Nottingham , University Park
b

Department of Production Engineering , Federal University of Sao Carlos , via

Washington Luiz, KM 235-13565-905, Sao Carlos, SP, Brazil
Published online: 15 Nov 2010.

To cite this article: Bart L. Maccarthy & Flavio C. F. Fernandes (2000) A multi-dimensional classification of
production systems for the design and selection of production planning and control systems, Production Planning &
Control: The Management of Operations, 11:5, 481-496, DOI: 10.1080/09537280050051988
To link to this article: http://dx.doi.org/10.1080/09537280050051988

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the Content)
contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy, completeness, or suitability
for any purpose of the Content. Any opinions and views expressed in this publication are the opinions
and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of
the Content should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising
directly or indirectly in connection with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or
systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution
in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions

PRODUCTION PLANNING & CONTROL, 2000, VOL. 11, NO. 5, 481 496

A multi-dimensional classi cation of production

systems for the design and selection of production
planning and control systems
Downloaded by [University of Sunderland] at 19:31 29 December 2014

BART L. MACCARTHY and FLAVIO C. F. FERNANDES

Keywords production systems, classi cation, production

planning and control, analysis and design of production systems
Abstract. A primary requirement for improved understanding
of the management of production systems is an appropriate
classi cation of such systems. This paper proposes a classi cation that facilitates a better understanding of real production
systems. It combines all the essential features, e.g. the ow of
materials with new classi cation perspectives with respect to
response time, repetitiveness and work organization. As much
previous work has in uenced the approach proposed here, the
paper also presents a review of relevant literature. The classi cation has four groups of characteristics, comprising eight
dimensions of descriptors, encompassing 12 variables. Choosing
or designing an appropriate production planning and control
system ( PPCS) is a di cult task due to the integrative character

of the production planning and control area: it tends to interface with all functional areas of the enterprise. An important
objective of the classi cation system proposed here is to provide
a tool to assist in undertaking this di cult task. Real production
systems are becoming more hybrid in order to be able to cope
with change. We show that our classi cation can successfully
deal with such systems.

1. Introduction
Scienti c knowledge is based on classi cation. It has
been asserted for instance that expert systems are, in general, classi cation systems (Jain 1988). It is not surprising
therefore that a classi cation of classi cations has been
given. Good (1965) proposed the following classes based

A uthors: B. L. MacCarthy, Division of Maufacturing Engineering and Operations Management,

University of Nottingham, University Park, Nottingham NG7 2RD, UK, and F. C. F. Fernandes,
Department of Production Engineering at Federal University of Sao Carlos, via Washington Luiz,
KM 235-13565-905 , Sao Carlos, SP, Brazil.
BA RT M A CC A RTHY is a Senior Lecturer in the Division of Manufacturing Engineering and
Operations Management at the University of Nottingham. Prior to his appointment in 1987 he
had substantial experience in research and management in manufacturing industry including the
engineering, textiles and clothing sectors. He has been Vice-Dean of the Faculty of Engineering for
3 years and was acting Head of Division during the academic year 1996/1997. His research spans
systems analysis, modelling, optimization, and simulation in manufacturing, operations and logistics. He has researched and published widely on decision support for planning and scheduling.
Current research projects include the analysis and modelling of responsiveness in order ful llment
processes, statistical process control for performance measurement in scheduling and the modelling
of supply chains in the public sector.
F L A VIO C. F. F ERNANDE S has been an Assistant Professor at the Federal University of Sao Carlos
in Brazil since 1991. He has authored more than 30 papers on production planning and control and
operations research, and refereed for several academic and professional journals. During 1988 he
has been a visiting scholar in the Division of Manufacturing Engineering and Operations
Management of the University of Nottingham in the UK. One of his principal current interests
is how to reduce the gap between theory and practice in the production planning and control eld.

Production Planning & Control ISSN 0953 7287 print/ISSN 1366 5871 online # 2000 Taylor & Francis Ltd
http://www.tandf.co.uk/journals

Downloaded by [University of Sunderland] at 19:31 29 December 2014

482

B. L. M acCarthy and F. C. F. Fernandes

on the purposes of producing classi cations: (i) for mental clari cation and communication; (ii) for discovering
new elds for research; (iii) for planning an organizational structure or machine; (iv) as a check list; (v) for
fun. The classi cation we propose in this paper falls principally into the third group but, as our classi cation
argues a new point of view, it is also useful for mental
clari cation and communication.
Burbidge (1985) noted that when an engineer designs a
machine, laws of physics and metallurgy help to produce
an e cient design, but there is not a set of scienti c laws
to help a production engineer faced with the problem of
designing an e cient production system. We contend
that this situation has not changed signi cantly in the
intervening years. One of the reasons for this is that
there is not an appropriate classi cation of real production systems. Existing classi cations are oversimpli ed,
only consider a limited number of aspects or take a speci c perspective. These classi cations tend to be of little
or limited value for analysing complex real production
systems or for aiding operations managers in selecting or
designing appropriate production planning and control
systems (PPCSs). Selecting or designing an appropriate
PPCS is a di cult task due to the integrative character of
the production planning and control function: it tends to
interface with all functional areas of the enterprise. A key
aspect of designing or selecting an e ective PPCS is the
ability to classify production systems. This paper focuses
on this area, as we believe it constitutes a signi cant gap
between theory and practice in operations management.
It would be pretentious to propose a totally watertight
classi cation. Moreover, any such claim is likely to be
easily contradictable. In practice any classi cation is a
trade-o between the level of detail needed for usefulness
and the level of aggregation desirable for usability. We
believe our classi cation here will be a signi cant contribution for understanding complex real production
systems, and for aiding design or selection of production
planning and control systems in practice.
This paper has been in uenced by much previous
work. Therefore, a literature survey is presented in section 2. The multi-dimensional classi cation is presented
in section 3. Section 4 shows how to apply the approach.
The nal section presents the conclusions.

2. Review of literature on the classi cation of

production systems
Here we survey the most important literature directly
relevant to the aim of this paper. We have selected references that are illustrative of approaches taken to the
subject. We present a review of classi cations of production systems in section 2.1 and a fairly brief review of

classi cations of production subsystems in 2.2. In section

2.3 we highlight the limitations of existing classi cations,
especially from the perspective of PPCS. In particular we
outline the major de ciencies that need to be addressed
with respect to contemporary manufacturin g enterprises
in section 2.4.
2.1. A review of production system classication
2.1.1. Pioneering classications
The pioneering classi cations are well known. Mallick
and Gaudreau (1951) identi ed three types of production: (i) continuous process (with disintegration, e.g.
petrol re ning; or with integration, e.g. synthetic rubber
processing) ; (ii) mass production; and (iii) intermittent
process. Wild (1971) splits the last type, names some of
the classes di erently and presents the following similar
classi cation: (i) process manufacture; (ii) mass production; (iii) batch production ; (iv) jobbing manufacture
(unitary production) . Many textbooks have followed
this type of scheme.
Burbidge (1962) de ned the following production
types: (i) line production [batch quantity (BQ) : 1; type
of ow (TF) : line] ; (ii) batch production (BQ: more than
1; TF: functional) ; (iii) jobbing production (BQ: same as
order quantity, generally small; TF: functional) ; (iv) process batch production (BQ: more than 1; TF: line) ; (v)
process jobbing production (BQ: same as order quantity,
generally small; TF: line). The following types of layout
were presented : (i) functional layout; (ii) group layout;
and (iii) line layout. Burbidge (1971), relating these types
of layout and some characteristics of production control,
de nes seven types of production systems: (i) line production; (iii) line batch production; (iii) group batch production; (iv) functional batch production; (v) line jobbing
production; (vi) group jobbing production; (vii) functional jobbing production.
Insu cient attention has been given to empirical data
on real production systems. Woodward (1965, 1980) conducted research on manufacturing rms in a region in the
UK. From information collected from 92 rms their production systems were then classi ed into 11 categories: (i)
production of units to requirements; (ii) production of
prototypes ; (iii) fabrication of large equipment in stages;
(iv) production of small batches to customers orders; (v)
production of large batches; (vi) production of customers
large batches on assembly lines; (vii) mass production;
(viii) intermittent production of chemicals in multi-purpose plant; (ix) continuous ow production of liquids,
gases and crystalline substances; (x) production of standardized components in large batches subsequently
assembled diversely; (xi) process production of crystalline

M ulti-dimensional classication of production systems

substances, subsequently prepared for sale by standardized production methods. Eight rms did not t any of
the categories, a further four were extremely mixed and
the other two were in transition.
The classi cation of Conway et al. (1967) should also
be noted as it is used particularly in operational research:
(i) single machine; (ii) parallel machines; (iii) ow-shop;
and (iv) job-shop. As noted by Botta et al. (1997), this
was expanded by Lenstra to include hybrid organizations
characterized by parallel machines at any processing
stage.

Downloaded by [University of Sunderland] at 19:31 29 December 2014

2.1.2. Classi cations derived by attributes

Here we consider classi cations based on attributes
that are perceived as important in production systems
or manufacturing enterprises more generally. Johnson
and Montgomery (1974) classify production systems
based on the types of products and processes as follows.
(i) Continuous system : few families of similar products
produced on a large scale. (ii) Intermittent system : frequent changes in the production stages from one product
to another, as a consequence of the large variety of manufactured items. They identify two subclasses: (a) intermittent ow-shop system : the ow pattern of all items is
the same; (b) intermittent job-shop system: items do not
have the same ow pattern. (iii) Large project system:
products are complex and special, and in many cases, are
produced in unitary quantities. They also consider a
fourth type (iv) the pure stock system where items
are bought, warehoused, distributed and sold, without
a processing phase.
Observing that cellular manufacturing is intermediate
in terms of application between the job-shop and the
ow-shop, Black (1983) proposed the following classi cation: (i) large project system; (ii) job-shop; (iii) cellular
manufacturing; (iv) ow-shop ; (v) continuous system.
Putnam (1983) summarized the basic di erences
between job-shop and ow-shop systems.
Constable and New (1976) consider three characteristics in their approach: the structure of products (simple
or complex) ; the layout (in line, functional or group layout) ; and the nature of the customers orders (for stock or
by order). Bu a and Miller (1979) adopt a classi cation
with four types of inventory production systems: (i) continuous system for stock; (ii) continuous system by order;
(iii) intermittent system for stock; and (iv) intermittent
system by order. An intermittent system indicates that
production occurs in lots. Large projects are included
in category (iv).
Nys (1984) de ned the technological system (TS) as
the part of a manufacturing system that comprises a set of
equipment for executing the technological process, the

483

equipment being linked by a material ow and an information ow. He presented two equivalent classi cations
on the basis of features of use a parallel classi cation
and a morphological block classi cation.
Schmidt et al. (1985) proposed classes derived from the
relationship between task divisibility, routing restrictions
and production rate uniformity. Frizelle (1989) presents
a categorization for plants by means of three letters (V, A
and T) that resemble the `shape of the plant. The `V
plant `is characterized by few raw materials subdividing
into many nished products. The `A plant presents
`many raw materials being assembled into few nished
products . The `T plant has a number of components
that `can be assembled in a multiplicity of ways .
Sipper and Shapira (1989) classify production systems
in accordance with the inventory control policy as: (i)
pure WIP system; (ii) modi ed WIP system (planned
to partly satisfy expected shortages) ; (iii) modi ed JIT
system (without any intermediate bu er stock and lots
greater than one) ; and (iv) pure JIT system (unitary
lots) .
In order to clarify the meaning of repetitive and intermittent manufacturing, De Toni and Panizzolo (1992)
de ne six categories of manufacturing systems: (i) individual [type of plant (TP) : yards] ; (ii) unique (TP:
laboratories) ; (iii) intermittent (TP: job-shops and
cells) ; (iv) discontinuous (TP: batch plants) ; (v) repetitive (TP: discrete lines) ; and (vi) continuous (TP: process
lines). The classes were obtained by combining classi cation of manufacturing systems, production volume and
how the product is produced.
Wild (1995) in his well-known textbook rst classi es
the operational system by function (manufacture, transport, supply and service), and then by structure. Wild
classi es manufacturing systems as: (i) make from stock,
to stock, to customer; (ii) make from source, to stock, to
customer; (iii) make from stock, direct to customer; (iv)
make from source, direct to customer.
Pyoun et al. (1995) present a classi cation of the realizable exibility in automated manufacturing systems.
They identify: (i) mass production; (ii) mid-variety and
mid-volume; and (iii) multi-variety and small-volume
production systems. Jichao (1996 ) classi es production
systems for the purpose of variability detection, as: (i)
simple production system (includes either a single process
or independent multiple processes) ; and (ii) complex production system (many processes with close inter-relationships). Dulmet et al. (1997) propose a classi cation of
processes according to the relationship between process
and product. In their approach the degree zero object is
the product, degree one object has direct leverage on the
product (e.g. tools and pallets) up to the degree n object,
which has direct leverage on the degree n 1 object. T hey

484

B. L. M acCarthy and F. C. F. Fernandes

state that it is su cient to dene ve levels for a description of a

given shop.

Downloaded by [University of Sunderland] at 19:31 29 December 2014

2.1.3. D escriptive classications

These are classi cations based on a description of the
attributes of rms or production systems in a nite number of classes. Ingham (1971) de ned eight types of business according to the relationships between marketing
and production. From his work, it is possible to identify
the following classi cation: make to stock; customize to
order, i.e. the client can specify its needs in terms of
design of certain class of products; make to order, i.e.
the items among a wide range of options are manufactured after the con rmation of customers orders; make to
order and to stock, i.e. make from orders for a wide range
of products and make for stock in the case of standard
products with considerable and continual demand; make
products to order and major components to stock (this
class corresponds to what is now known as assemble to
order) ; totally customized to order, i.e. `the rm does not
o er a range of products but a production service within
the limits of its equipment .
Barber and Hollier (1986a, 1986b) developed a classi cation of engineering batch manufacturin g companies
based on questionnaire evidence. They propose six
groups according to the level of production control complexity. It is clear that several of the measures have been
di cult to estimate or interpret for many rms and
responses were necessarily subjective.
McCarthy et al. (1997) consider the evolution of manufacturing changes in order to show how to construct a
classi cation based on cladistics (Kitching et al. 1998)
and give an example using the automotive sector. They
emphasize that with cladistics `it is possible to examine
the way in which characters change within groups over
time, the direction in which characters change, and the
relative frequency with which they change . . . Thus,
organizational cladograms can be used as a tool for
achieving successful organizational change. Unfortunately, other sectors are not nearly as well documented in the literature as the automotive sector.

system; (iv) multi-product single-machine system ; (v)

mixed-product sequential ow line I, where all the products have the same sequence of operations and there is no
necessity for machine resetting; (vi) multi-product
sequential ow line where products have the same operational sequence but they have to be produced separately
in batches; (vii) mixed-product bypass ow line I where
some products are not processed in all machines and
equipment resetting is not necessary ; (viii) mixed-product bypass ow line II, where some products are not
processed in all machines and equipment resetting is
necessary; (ix) multi-product backtracking ow line
where equipment resetting occurs, production is in
batches and the variation in operational sequences is
due to omitted and/or backtracked operations and the
product ow is bidirectional; and (x) multi-product
multidirectional backtracking system where the products
are batched, the operational sequences are so varied that
ow is multidirectional, and therefore, line production is
not feasible.
Burbidge (1970) identi ed four categories of group
technology systems: (i) single-machine system; (ii)
group layout system; (iii) total group layout system
(group layout plus classi cation and coding system,
value analysis, variety reduction, standardization ) ; and
(iv) line ow system, with characteristics between class
(iii) and mass production.
Many classi cations of exible manufacturing systems
(FMS) have been presented, e.g. Groover (1980), Kusiak
(1985), and Maimon and Nof (1986). Browne et al.
(1984) classi ed FMSs as: exible machining cells, exible machining systems, exible transfer lines and exible
transfer multi-lines. Stecke and Browne (1985) added the
descriptor `type of material handling system to the previous classi cation, so the resulting classi cation is based
on the ow pattern of parts and more speci cally on the
routing exibility. MacCarthy and Liu (1993) and Liu
and MacCarthy (1996 ) presented a classi cation scheme
for FMSs based on a consistent set of de nitions. They
distinguished between : (i) a single exible machine
(SFM) ; (ii) a exible manufacturing cell (FMC) ; (iii)
a multi-machine exible manufacturing system
(MMFMS) ; and (iv) a multi-cell exible manufacturing
system (MCFMS) , and show the relationship between
them.

2.2. Classications of production subsystems

Important contributions were made by Petrov (1966 ),
Wild (1972) and Carrie (1975). Aneke and Carrie (1984)
integrated existing classi cations of simple ow lines
(mass production) and group technology ow line using
six criteria. The classi cation reduces to: (i) single-product single-machine system; (ii) single-product multimachine system; (iii) mixed-product single-machine

2.3. General comments on existing classications

The only signi cant previous paper that we have found
that has reviewed production system classi cations is that
by McCarthy (1995) with 14 references. He noted that
previous classi cations, with the exception of Barber and
Hollier (1986a), were subjective and that cladistics pro-

Downloaded by [University of Sunderland] at 19:31 29 December 2014

M ulti-dimensional classication of production systems

vided a way of making an objective classi cation. We
would argue however that all classi cations are necessarily subjective and that the important issue is not the
subjectivity of the classi cation but rather its usefulness.
Classi cation has to be subjective because it represents an
author s perspective on production systems. Although
Barber and Hollier (1986a) use a quantitative approach
in arriving at their classi cation scheme, it is also subjective because the number of groups, the characteristics
the system may possess and the factors utilized to measure
the characteristics are all chosen. Moreover, McCarthy
et al. (1997) also present an essentially subjective classi cation, e.g. there are many subjective choices in the
chosen clade.
All attempts at classi cation are necessarily approximations and can always be criticized on this basis. Any
classi cation involves choosing between the level of detail
and the level of aggregation. However, if the perspective
and the rationale are clear then the bene ts of a good
classi cation are many. Some of the classi cations discussed here are useful in providing insight and understanding of production systems, and some have clearly
demonstrate d their usefulness in particular domains.
We argue, however, that there is a real need for classi cations with de ned objectives and areas of application.
Unfortunately this is not the case in many of the classi cations above.

2.4. D eveloping a classication for the design or selection of

PPCS
The need for improved classi cations for the study,
analysis, design and management of production systems
in today s global economies is clear. For example,
Banerjee (1997) shows evidence that `in spite of many
millions having been spent in manufacturing planning
and control systems . . . no real solution to the need for
greater responsiveness and exibility has been found. In
the context of contemporary manufacturing organizations, and in particular their control systems, the following are drawbacks that can be applied in some measure
to the classi cations discussed above.
. Too general or very super cial to be of use in any
context.
. Disregard of important concepts, e.g. ow pattern.
. Consideration of ow-shop characteristics only
within group technology cells.
. Placing the concept of job-shop in the context of
ow lines.
. Failing to distinguish between assembly lines and
production lines.

485

. Consider assembly as a process related only to

assembly lines, ignoring the fact that other types
of installations may also carry out assembly operations.
Many further issues are crucial in contemporary
manufacturing organizations and need to be addressed
in e ective classi cations. In the past, changes in business
needs occurred at low speeds and simple classi cations of
production systems were appropriate as a basis for
designing a production system. In today s environments,
in order to accommodate more rapid changes, manufacturing systems have become more and more hybrid.
There is a real need for a classi cation that treats hybrid
production systems in depth. A further issue is that of
work organization . Previous classi cations do not consider this issue but it is vital in contemporary organizations. Response time and repetitiveness are also critical
variables from a control viewpoint and an e ective classi cation must address them. It is not su cient just to
add a new variable to an existing classi cation. The overall scheme must be coherent in order to be applied e ectively.
The classi cation we present in the remainder of this
paper addresses these issues and most of the points noted
above in a rational way. Our principal objective is to aid
the design and selection of production planning and control systems. The previous classi cations that were most
in uential to us are: Wild (1971) , Burbidge (1971),
Ingham (1971), Johnson and Montgomery (1974),
Constable and New (1976 ), Black (1983) and Aneke
and Carrie (1984).

3. A multi-dimensional classi cation for

production systems
3.1. Overall structure
A production system may be de ned as a set of interrelated elements that are designed to act in a manner that
generates nal products whose commercial value exceeds
the costs of generating them. From the nature of these
elements, two types of subsystem may be identi ed:
physical systems and managerial systems. In the rst
the elements are physical entities (e.g. a machine) and
in the second the elements are procedures that transform
data into information in a decision process (e.g. an MRP
system) . Naturally people design and operate both types
of subsystem and can therefore be considered as elements
of both.
Here we propose a multi-dimensional classi cation for
production systems. We identify four main groups of
characteristics, comprising eight dimensions (A/B/C/D/

486

B. L. M acCarthy and F. C. F. Fernandes

Downloaded by [University of Sunderland] at 19:31 29 December 2014

E/F/G/H). Where appropriate, important variables

within some of the dimensions have also been identi ed.
The four groups and their associated dimensions are as
follows.
(1) General characterization: encompasses the following dimensions: enterprise size (A) ; response time
(B) ; repetitiveness (C) ; and automation level (D) .
(2) Product characterization: encompasses the product description (E) .
(3) Processing characterization: encompasses the processing description (F).
(4) Assembly characterization: encompasses the
dimensions: types of assembly (G) and types of
work organization (H).
The choice of these groups, dimensions and variables
was determined by our principal objective, i.e. that the
classi cation be a valuable tool for designing or choosing
a PPC system. Our selection is in uenced by our industrial and academic experience in PPC. It is our view that
the dimensions, variables and levels represent a su cient
set in terms of breadth, depth and level of detail to capture the salient characteristics of most production systems
from the perspective of production planning and control.
In particular we have addressed the de ciencies of existing classi cations with respect to the reality of contemporary manufacturing organizations. The relevance of
the scheme to production planning and control systems
is discussed in section 4, and examples of applications are
presented in section 4.2.
In the following section we present a description of the
scheme. We enlarge on the discussion where we feel it
necessary to justify our selection of dimensions and variables. We also emphasize some parts of the classi cation
because of their importance and novelty. In particular
we present new insights in relation to the existing classi cation literature on response time, repetitiveness level,
ow of materials and types of work organization. The
latter we have highlighted under assembly but in some

cases it may also be necessary to consider it under processing characteristics.

Table 1 illustrates the structure of the classi cation
scheme. We use the forward slash symbol = to separate
the dimensions and the underscore symbol to separate the variables. Letters and numbers are used as shorthand notation for levels or categories of each within each
dimension.

3.2. D imensions and variables of the classication system

3.2.1. General characterization
3.2.1.1. First dimension ( A) : enterprise size
Several descriptors may de ne the size of an enterprise:
turnover, number of employees, market share, etc. For
the purpose of analysing the PPCS, turnover or revenue
are not good descriptors because either or both may be
large if, e.g. raw materials are very expensive. The more
relevant descriptor of enterprise size is the number of
employees. By the number of employees, a rm in the
UK is considered large if it has more than 250 employees
(in Brazil, more than 500) and it is considered medium
sized if the number of employees is between 50 and 250
(between 100 and 500 in Brazil).
3.2.1.2. Second dimension ( B) : response time
Simply stating that a production system is make-tostock or make-to-order , etc. is of little use from the
point of view of designing modern production planning
and control systems. Here we identify the dimension,
response time, which speci es how, strategically, an
enterprise wants to attend to its customers needs.
Figure 1 illustrates three important parameters of an
industrial production system: supplier lead time (SL),
production lead time (PL) and distribution lead time
(DL). The ` rst tier refers to the rst vendor in the
supply chain. The response time (RT) of the production
system is the sum of SL, PL and DL. E ective manage-

Table 1. The multi-dimensional classi cation system ( MDCS).

General characterization

Enterprise size
Response time
* Repetitiveness level
* Automation level
Product description
*
*

Product characterization

*
*
*

Processing characterization

Processing description

*
*
*

Assembly characterization

*
*

Product structure
Level of customization
Number of products
Types of bu er
Type of layout
Types of ow

Types of assembly
Types of work organization

M ulti-dimensional classication of production systems

487

least 75% of the items are non-repetitive and to be

semi-repetitive if at least 25% of the items are repetitive
and at least 25% of the items are non-repetitive.
Undoubtedly these cut-o points are somewhat arbitrary, but they do re ect our experience of real production systems. Using these de nitions we identify a range
from maximum repetitiveness (the pure continuous
system) to the minimum repetitiveness (large-scale
projects) .

Downloaded by [University of Sunderland] at 19:31 29 December 2014

Figure 1. Response time (RT) of a production system.

ment of response time is central to the achievement of

competitive advantage. In the extreme case where the
enterprise maintains stocks of all purchased materials
and all nal products, RT equals DL. Essentially
response time is a policy decision in uenced by technological and operational constraints, marketing, and
customers requirements and strategy.
We identify the following values for the parameter B of
the classi cation:
. B SL PL DL if the system produces by order;
. B DLa (P%) if the system produces for stock and
the service level is equal to P%;
. B DLb (P%) if the system does not produce (only
buys, stocks, sells and delivers items) and the service
level is equal to P%;
. B PL DL if the system produces to order but
maintains stocks of raw materials;
. B SL DL if the system does not produce but it
sells to order.
3.2.1.3. T hird dimension ( C) : repetitiveness
In the report by APICS (1982), the term `repetitiveness is associated with the production volume of discrete
items: the larger the volume, the more repetitive the
production system is considered to be. However, in an
environment where the production volume is very low
due to very large processing times, e.g. one item per
month, and the system only produces this item, then, in
spite of the low production volume, clearly the system
must be considered to be repetitive. Repetitiveness must
therefore be considered to be a function of more variables
than just production volume. The approach we adopt is
to rstly de ne what we mean by a repetitive product
and then de ne what we mean by repetitiveness of a
production system. We de ne a product to be repetitive
if it consumes a signi cant percentage of the annual
available time of the production unit (we specify at least
5%). We de ne a production system to be repetitive if at
least 75% of the items that it produces are repetitive ones.
We de ne a production system to be non-repetitive if at

. C PC: pure continuous system, e.g. petroleum

re ning.
. C SC: semi-continuous system, e.g. each processing unit is a pure continuous system and there
are combinations of routes through processing units.
In the process industries these are sometimes known
as batch processing production systems.
. C MP: mass production system. Almost all items
are repetitive.
. C RP: repetitive production system. At least 75%
of the items are repetitive. In the case of the metal/
mechanical parts sector, a typical RP production
system is the cellular manufacturing system with
ow-shop pattern of ow.
. C SR : semi-repetitive production system. There
are a considerable number of both repetitive and
non-repetitive items. In the case of the mechanical
parts sector, a typical SR production system is the
cellular manufacturing system with job-shop pattern of ow.
. C NR : non-repetitive production system. The
majority (at least 75%) of items are not repetitive.
. C LP: large projects.
3.2.1.4. Fourth dimension ( D ) : automation level
The importance of the automation level for the control
of production systems has been recognized for a long
time. Bright (1958) demonstrated that the nature of the
control has a close relationship with the levels of automation. We di erentiate the following states.
. N. Normal automation comprises all types of
mechanization where the human has a high degree
of participation at the operational or execution
level. Here we include classical ow-shop and jobshops, cellular manufacturin g systems with owshop characteristics (CM1 ) ; cellular manufacturing
systems with job-shop characteristics (CM2 ). In
CM1 the ow pattern is common and in CM2 the
ow pattern is variable, allowing stages to be missed
out and allowing counter- ows ( gure 2) .
. F. Flexible automation has, at the operational or
execution level, computer control taking the main

488

B. L. M acCarthy and F. C. F. Fernandes

production system with just a few additional
operations.
(4) Standard products: the clients do not interfere
in the product design.
. Number of products: we distinguish between:
S: for a single product; and
M: for multiple products.

Downloaded by [University of Sunderland] at 19:31 29 December 2014

Figure 2. Typical material ow in the semi-repetitive cellular

manufacturing CM2.

role by means of technologies, e.g. local area networks and computer numerical control, and will
often be accomplished by some form of FMS technology. Here we distinguish between ow-shop exible manufacturing system (FMS1 ) and job-shop
exible manufacturing system (FMS2 ).
. R. Rigid automation is the type found in transfer
lines with highly specialized and dedicated automatic equipment.
. M. Mixed automation occurs where the production
system has processing units with di erent automation levels. For example it could be composed of a
manufacturing cell with normal level of automation
and a FMS ( exible automation) .

3.2.2. Product characterization

3.2.2.1. Fourth dimension ( E) : product description
We identify three variables under the product description.
. Product structure: here we simply di erentiate
between:
SL: denoting single-level products requiring no
assembly; and
ML: denoting multi-level products requiring assembly.
. Level of customization: we distinguish between the
following.
(1) Customized products where the clients de ne
all the parameters of the product design.
(2) Semi-customized products where the clients
de ne part of the product design.
(3) `Mushroom customization. Mather (1998)
describes this concept as delaying product differentiation as late as possible in the production
system. There are a number of standard components or modules that are combined in a
large number ofways in the nal stages of the

Thus, a product characterization of ML_2_M

describes a production system with multiple products
having multi-levels and at least some customer-de ned
product design parameters. The idea of a homogeneous
product range is important when we come to applying
the classi cation. By this we mean that all products fall
into the same product characterization. We have seven
possible sets of homogeneous products (SL_1, SL_2,
SL_4, ML_1, ML_2, ML_3, ML_4). The case SL_3
(single level and `mushroom customization ) is not possible.

3.2.3. Processing characterization

3.2.3.1. Fifth dimension ( F) : processing description
This important aspect of production systems is represented by three variables.
. Types of layout: for this variable we identify the
following types of layout:
P: product layout;
F: functional layout or layout by process;
G: group layout;
FP: xed position layout : the resources (human,
equipment) move and not the product.
. Types of bu er : for this variable we distinguish
between the following types of bu er:
(1) bu ers before the rst production stage;
(2) intermediary bu ers between production
stages;
(3) bu ers after the last production stage.
. Types of ow: the third variable de nes the following types of ow:
F1: single-stage , e.g. a machining centre;
F2: single-stage with identical machines in parallel;
F3: single-stage with non-identical machines in
parallel ;
F4: unidirectional multi-stage processing (e.g. a
classical ow-shop system) ;
F5: unidirectional multi-stage processing that allows
stages to be skipped;
F6 : unidirectional multi-stage processing with equal
machines in parallel;

M ulti-dimensional classication of production systems

Downloaded by [University of Sunderland] at 19:31 29 December 2014

F7: unidirectional multi-stage processing with identical machines in parallel but allowing stages to be
skipped ;
F8: unidirectional multi-stage processing with nonidentical machines in parallel;
F9: unidirectional multi-stage processing with nonidentical machines in parallel, allowing stages to be
skipped ;
F10: multi-directional multi-stage processing (e.g. a
classical job-shop) ;
F11: multi-directional multi-stage processing with
identical machines in parallel;
F12: multi-directional multi-stage processing nonidentical machines in parallel.

489

Figure 6. Multi-directional, multi-stage processing with nonidentical parallel machines.

Obviously F12 is the most complex type of ow and all

previous types are particular cases of this one. A description, e.g. G_1-3_F5 is a processing facility with group
layout and bu ers before the rst and after the last production stages with type of ow F5 (unidirectional multistage processing that allows stages to be skipped).

3.2.4. Assembly characterization

3.2.4.1. Seventh dimension ( H)
We distinguish between nine types of assembly.
Figure 3. Single-stage processing, the general case.

Figure 4. Uni-directional, multi-stage processing with identical

parallel machines.

Figure 5. Uni-directional, multi-stage processing with nonidentical parallel machines.

. A1: Mixing (e.g. chemical ingredients).

. A2: Assembly of a large engineering project (e.g. a
large bridge) typically in a layout of xed position.
. A3: Assembly of heavy products (e.g. a large toolmachine) in a layout of xed position.
. A4: Assembly of light products (e.g. a medical
equipment) in one workstation or in one of a set
of parallel workstations.
. A5: Paced assembly line where a conveyor never
stops and the workers may move to perform their
tasks.
. A6: Paced assembly line where a conveyor stops
for a number of units of time (cycle time) and the
workers remain xed at their individual workstations.
. A7: Semi-paced assembly line where a conveyor
always moves and the worker releases the product
only when he nishes his tasks.
. A8: Unpaced assembly line where a conveyor
always moves and the worker simply attaches the
product to the conveyor when he nishes his tasks.
. A9: Unpaced assembly line where a transporter
only moves when a worker activates it after nishing
his tasks (e.g. an overhead travelling crane).

490

B. L. M acCarthy and F. C. F. Fernandes

Downloaded by [University of Sunderland] at 19:31 29 December 2014

3.2.4.2. Eighth dimension ( J ) : type of organization of the work

For this dimension we adopt a classi cation of the
organization of the work based on Johnson (1991). Our
classi cation is the rst to encompass this aspect. Work
organization may also be relevant to other types of processes, but from a production control perspective it has
much greater impact in assembly operations. We distinguish the following ve types.
. (I ) Individual work: the number of workers is equal
to the number of workstations. In the case of assembly lines, the criterion for allocating each task
to a workstation is the balancing of the whole line.
Two speci c categories can be identi ed.
(Ia) Without rotation. Each worker is xed on a
workstation.
(Ib) With rotation. After each task has been allocated to a workstation, the rst set of workers
of the rst matched set of workstations form
the rst team and the second set of workers
form the second team, etc. Workers in the
same team may rotate their workstations.
. (T) Team work. The workstations (or sublines) are
pre-de ned and each one is operated by a single
team with several workers. The tasks carried out
by each worker of the teams are decided by balancing of the subline. Two speci c cases can be identi ed.
(Ta) Every task is assigned to a speci c workstation.
(Tb) Just some tasks are assigned to a speci c workstation.
. (G) Self-managed work groups. As in the team work
cases the workstations are rst de ned and then the
tasks to be performed by each workstation, but the
group of workers of each workstation has autonomy
to organize the work inside the group.

4. A ppraisal and application of the multidimensional classi cation system

4.1. Relationships between the dimensions in the classication
scheme
Certain groups of characteristics occur quite commonly in production systems. Here we illustrate some
typical relationships between the general characterization dimensions we have identi ed. For instance, the
numbers of employees, level of automation and size of
the production system tend to be correlated. The size of
the production system is closely related to important features of the production system, e.g. the amount of capital
that can be invested in a PPCS. Also in real production

Figure 7. The level of automation and the level of repetitiveness.

systems there is a correspondence between the volume of

production per product and variety of products, size of
the production system and level of repetitiveness of the
production system.
The relationship between volume of production Q and
variety of products P is well known. For high Q and low
P a product layout is appropriate, for medium Q and
medium P a group layout is appropriate, and for low Q
and high P a functional layout is appropriate. The repetitiveness level essentially combines P and Q in one variable. This is illustrated in gure 7 relating repetitiveness
level and automation level. Here we plot the typical level
of repetitiveness against the level of automation for some
important modern manufacturing systems. The level of
repetitiveness increases from job-shop, to CM2 , CM1 and
ow-shop systems. Naturally FMS1 has greater repetitiveness than FMS2. Transfer lines have the greatest repetitiveness level. What is uncertain is the position of lines I
and II ; this depends very much on the technology and
control system employed in the FMS.
The response time, level of customization and the way
we apply our classi cation enables us to treat the relationship between the production system and the customers in a deeper way than in traditional textbook
classi cations, e.g. Ingham (1971) and Wild (1995).
The response time and the level of customization describe
the relationship between the production system and the
market, and it is important therefore that literature
related to PPC systems speci es this relationship.

491

M ulti-dimensional classication of production systems

Downloaded by [University of Sunderland] at 19:31 29 December 2014

4.2. Using the classication to choose the overall structure of a

PPCS
All the 12 variables considered in our multi-dimensional classi cation have direct impact on the complexity
of the production planning and control (PPC) activities.
Table 2 indicates the typical impact of each variable on
the complexity of PPC activities and the relationship
between each variable and the level of repetitiveness.
Repetitiveness is an important variable in our classi cation and we believe it to be the key variable for choosing
the overall structure of the PPCS. Table 2 also indicates
the relationship between the variables and the choice of a
PPCS. The last line of table 2 is justi ed by the following
reasoning: for discrete items the more repetitive the production system is the more likely that the simplest of all
PPCS the kanban can be chosen; for intermediate
situations period batch control (PBC; Burbidge 1996 ) is
likely to be appropriate and for non-repetitive situations
an MRP-based approach is likely to be necessary.
Kanban and PBC due to their logic and procedures do

not work well in non-repetitive situations. For some cases

there is the possibility to choose OPT type systems. For
large projects, PERT/CPM may be the most appropriate
choice.
While the level of repetitiveness has a strong impact on
the choice of basic PPCS, the other variables have signi cant impact on the complexity of the detailed system
to be de ned. For example, MRP may be chosen as the
basic system but the parametrizati on of the system
depends on the complexity of the PC activities. These
are related to the variables in table 2 and also the constraints that are restricting the production system.

4.3. Applying the scheme

Table 3 gives a complete view of the attributes of the
multi-dimensional classi cation scheme.
If the production system is composed of one set of
homogenous products and one processing unit and/or
one assembly unit, the application of the multidimen-

Table 2. The variables and the choice of a PPC system.

Level of repetitiveness of production systems
Other variables
Enterprise size
Response time
Automation level
Product structure
Level of
customization

Pure
continuous

Semi
continuous

Semirepetitive

Nonrepetitive

Large
projects

For all level of repetitiveness, the PPC activities for multi-level products are much more complex than for singlelevel products
Standard
Standard
Standard
Standard
Mushroom
SemiCustomized
products
or mushroom or mushroom or mushroom or semicustomized
customized
or customized
the PPC activities for multi-products are much more complex than for singleProduct
layout
(i) , ( ii)
and ( iii)

Group
layout
(i) , ( ii)
and ( iii)

Group
layout
(i), ( ii)
or ( i)

The complexity of PPC activities increases from (F1) in direction to (F12)

(A1) or
(A1) or
(A5) or ( A6) (A5) or ( A6) (A7) or ( A8)
disassembly
disassembly
or ( A7) or
or ( A7) or
or ( A9) or
no assembly
no assembly
no assembly

Types of work
organization
Basic production
control system
possible to be
chosen

Repetitive

For all levels of repetitiveness, the larger the enterprise the greater the complexity of production planning and
control ( PPC) activities
DL(a7P%) DL(a7P%) DL( a7P%) DL( a7P%) PL DL
PL DL or
SL PL DL
SL PL DL
Rigid
Rigid
Rigid
Normal or
Normal or
Normal or
Normal
exible
exible
exible

Number of products For all levels of repetitiveness,

product
Types of layout
Product
Product
layout
layout
Types of bu er
(i) and ( iii)
(i) , ( ii)
and ( iii)
Types of ow
Types of assembly

Mass
production

A spreadsheet to
control the
rate of ow

If there is assembly, the type of work organization has

a direct impact on the way you will balance the work
in the assembly
A spreadKanban
Kanban or
PBC or
sheet to
PBC
OPT
schedule
the work

Functional
layout
(i), ( ii)
or (ii)

Fixed
position
Without
bu ers

(A3) or ( A4)
or no
assembly

(A2)

MRP

PERT/
CPM

492

B. L. M acCarthy and F. C. F. Fernandes

Table 3. Summary of multi-dimension classi cation system (MDCS) .

Downloaded by [University of Sunderland] at 19:31 29 December 2014

General characterization
(A)

(B)

(C)

(D)

Processing characterization

Enterprise size
( F)
(L) : large number of employees
(M) : medium number of employees
(S) : small number of employees
Response time
SL PL DL if the system produces by order
DLa P% if the system produces for stock and
the service level is equal to P%
DLb P% if the system does not produce ( only buy,
stock, deliver and sell items) and the services level is
equal to P%
PL DL if the system produces by order but maintains
stocks of raw materials
SL DL the system does not produce and attends the
clients by order
Repetitiveness level
PC ! pure continuous system
SC ! semi-continuous system
MP ! mass production system
RP ! repetitive production system
SR ! semi-repetitive production system
NR ! non-repetitive production system
LP ! large project
Automation level
(N) : normal automation
(F) : exible automation
(R) : rigid automation
(M) : mix automation

Product characterization
(E)

Product description
* Product structure
(SL) : single-level products
(ML) : multi-level products
* Level of customization
(1) : customized products
(2) : semi-customized products
(3) : mushroom
(4) : standard products
* Number of products
(S) : single-products
(M) : multi-products

sional classi cation is straightforward. Otherwise, we

must consider the following general approach ( gure 8) .
We have to make the following explicit.
(1) The number n i of products in each case i n1
jSL_1j;. . . ; n7 jML_4j; where the modulus symbol indicates the number of elements in the set).

Processing description
* Types of layout
( S) : single workstation
( P) : product layout
( F) : functional layout (layout by process)
( G) : group layout
( FP) : layout of xed position
* Types of bu er
( 1) : bu er before the rst production stage
( 2) : bu ers between intermediary stages
( 3) : bu er after the last product stage
* Types of ow
( F1) : mono-stage
( F2) : mono-stage with equal machines in parallel
( F3) : mono-stage with unequal machines in parallel
( F4) : uni-directional multi-stages
( F5) : variable unidirectional multi-stages
( F6) : uni-directional multi-stages with equal
machines in parallel
( F7) : variable uni-directional multi-stages with
equal machines in parallel
( F8) : uni-directional multi-stages with unequal
machines in parallel
( F9) : variable uni-directional multi-stages with
unequal machines in parallel
( F10) : multi-directional multi-stages
( F11) : multi-directional multi-stages with equal
machines in parallel
( F12) : multi-directional multi-stages with unequal
machines in parallel
Assembly characterization

( G)

( H)

Types of assembly
( A1) : mixture of chemical ingredients
( A2) : assembly of a large project
( A3) : assembly of heavy products in a layout of
xed position
( A4) : assembly of light products in one or in
parallel workstations
( A5) : paced assembly line where the conveyor
never stops
( A6) : paced assembly line where the conveyor stops
for C ( cycle time) units of time
( A7) : semi-paced assembly line
( A8) : unpaced assembly line I
( A9) : unpaced assembly line II
Types of work organization
( I) : individual workers
( T) : work teams
( G) : work groups

(2) If the level of automation is mixed (dimension

B M) then we have to specify for each processing unit the level of automation (ALi ) and
other descriptors (type of layout, type of bu er
and type of ow).
(3) An assembly characterization (dimensions G and
H) for each assembly unit 1;. . . ; n.

M ulti-dimensional classication of production systems

493

Downloaded by [University of Sunderland] at 19:31 29 December 2014

Figure 8. The general approach to apply the multi-dimensional classi cation.

Figure 9. The classi cation for the example problem.

(4) Depending on the complexity of the production

system, we may need to show the relationship
between the processing units and bu ers. A directed graph may be an appropriate approach.
Here we rst present a concise illustrative example and
then apply the approach to four examples from the
literature.
A production system manufactures machines utilized
in the construction of roads. There are 2000 employees.
There are processing units with exible levels of automation and others with normal levels of automation. The
system produces to order but maintains stocks of raw
materials. It is a semi-repetitive production system. The
products are complex, standard, multilevel ones. There
are ve processing units. The rst one has a functional
layout with bu ers between the intermediate stages,
before the rst and after the last stage, a normal level
of automation and a multi-directional multi-stage ow
pattern with non-identical machines in parallel. The second and third processing units are manufacturing cells
with normal levels of automation. The second one has a
bu er before the rst stage and a uni-directional ow
pattern with identical machines in parallel. The third
one has a bu er after the last stage and a unidirectional

ow-shop ow pattern. The fourth and fth processing

units are exible manufacturin g systems with job-shop
type ow patterns and bu er after the last stage. There
is one assembly unit of the unpaced type where the conveyor always moves and the worker only attaches the
product to the conveyor when he nishes his tasks with
normal level of automation operated by work groups.
Figure 9 shows the classi cation for this production
system and gure 10 uses a directed graph to show the
relationship between the units and the bu ers.
We have also applied the approach to four production
systems described in recent literature in production planning and control. Obviously the amount of detail given in
each source is limited and variable, and the results are
necessarily incomplete, but it was felt to be an objective
test of the applicability of the approach. Table 4 summarizes the major characteristics of each of the systems
from the information provided in the source.
Artiba (1994) gives more detail for the processing characteristics a production system composed of six processing units all with ow pattern (F4) and the
transport of the materials between adjacent workstations
is made through pipes. The relationships among the six
units are shown in gure 11 [e.g. the products that pass
processing unit 1 (PU1) follow through to PU4 or PU5].

494

B. L. M acCarthy and F. C. F. Fernandes

Downloaded by [University of Sunderland] at 19:31 29 December 2014

Figure 10. The relationship between units and bu ers.

Figure 11. The relationships among the six units (based on Artiba 1994).
Table 4. Application of the MDCS to some literature examples.
References
Characterization
General

Variables
Enterprise size
Response time
* Repetitiveness level
* Automation level
Product description:
* Product structure
* Level of customization
* Number of products
Processing discription:
* Types of layout
* Types of bu ers
* Types of ow
* Types of assembly
* Types of organization of work
*
*

Product

Processing

Assembly

[1]

[2]

[3]

[4]

*
*
NR
F

*
*
MP
R

*
*
MP
R

*
*
SC
R

ML
1
M

ML
3
M

ML
4
M

ML
3
M

F
2
F11
A4

S
1 3
F1
A4

P
2
F4
A5

P
1 3
**
A1
1

Notes: A trace means that the descriptor considered does not apply ; an asterisk means that the information was omitted; and a double asterisk
means that additional description will be provided on the information cannot be put into a simple table.
References:
[1] Anwar, M. F. and Nagi, R., 1998, Integrated scheduling of material handling and manufacturing activities for just-in-time production of complex
assemblies. International J ournal of Production Research, 36, 653 681.
[2] Maimon, O. Z. and Braha, D., 1998, A genetic algorithm approach to scheduling PCBs on a single machine. International J ournal of Production
Research, 36, 761 784.
[3] Dowlatshahi, S., 1996, An e cient solution for determining location and size of bu er stocks. Production Planning & Control, 7, 282 291.
[4] Artiba, A., 1994, A rule-based planning system for parallel multiproduct manufacturing lines. Production Planning & Control, 5, 349 359.

5. Conclusions
Di erent classi cations and taxonomies are necessary
for di erent purposes as di erent perspectives on manufacturing systems are important for di erent aspects of
analysis and design. However, many previous classi cations have limited application: some are not explicit
about the purpose of the proposed classi cation and/or

are very super cial whilst others disregard important elements. The lack of useful classi cations is one reason that
we believe underlies the lack of progress in operations
management generally and production planning and
control in particular.
The classi cation scheme described here can be successfully applied to real production systems because it
can treat in-depth the typical hybrid cases that arise in

Downloaded by [University of Sunderland] at 19:31 29 December 2014

M ulti-dimensional classication of production systems

practice. Each speci c production system demands a
speci c, if not unique, production control system. For
example, under the general heading of assembly lines,
there are in reality, several di erent types, each of
which requires a di erent production control system.
Our classi cation highlights the essential characteristics
that enable a better understanding of production systems
and their relationship with the production planning and
control function. In future research, we intend to demonstrate further its utility for aiding in the selection or
design of appropriate production planning and control
systems. More generally it may be used for research,
particularly in studying the gap between practice and
theory in the production planning and control area. It
may also be used for teaching in operations management
and industrial engineering. We see it being used to
help students in classifying systems from detailed
descriptions and in classifying production systems studied
in the eld.
A cknowledgement
The authors wish to acknowledge the support of the
FAPESP in funding this work.

References
A NEKE , N. A. G., and C A RRIE , A. S., 1984, A comprehensive
owline classi cation scheme. International J ournal of Production
Research, 22, 281 297.
API CS R E P E TITIVE M A NUFA CTURING G ROUP , SECOND
Q UA RTE R 1982, Repetitive manufacturing. Production and
Inventory M anagement, 23, 78 86.
BA NERJ EE , S. K., 1997, Methodology for integrated manufacturing planning and control systems design. In A. Artiba
and S. E. Elmaghraby (eds) T he Planning and Scheduling of
Production Systems (London: Chapman & Hall), pp. 54 88.
BA RBE R , K. D., and H OLL IER , R. H., 1986a, The use of numerical analysis to classify companies according to production
control complexity. International J ournal of Production Research,
24, 203 222.
BA RBE R , K. D., and H OLLIE R , R. H., 1986b, The e ects of
computer-aided production control systems on de ned
company types. International J ournal of Production Research, 24,
311 327.
BLA CK, J . T., 1983, Cellular manufacturing systems reduce
setup time, make small lot production economical. Industrial
Eng ineering, 15, 36 48.
BOTTA , V., G UINET, A., and BOUL LE , D., 1997, Object-oriented
analysis with structured and integrated speci cations and
solutions (OASISS) for production system control. J ournal
of Intelligent M anufacturing , 8, 3 14.
BRIGHT , J . R., 1958, Automation and M anagement ( Boston:
Harvard University Press) .
BROW NE , J ., D UBOIS, D., R A THMIL L , K., SETHI , S. P., and
STECKE , E., 1984, Classi cation of exible manufacturing
systems. FM S M agazine, 2, 114 117.

495

BUFFA , E. S., and M IL LE R , J . G., 1979, Production-Inventory

Systems Planning and Control ( US: Richard D. Irwin).
BURBIDGE , J . L., 1962, T he Principles of Production Control ( rst
edition) (London: MacDonald & Evans).
BURBIDGE , J . L. ( ed.), 1970, Final Report, International Seminar on
Group T echnology ( Turin: Turin International Centre) .
BURBIDGE , J . L., 1971, T he Principles of Production Control (third
edition) (London: MacDonald & Evans).
BURBIDGE , J . L., 1975, T he Introduction of Group T echnolog y
( London: William Heinemann).
BURBIDGE , J . L., 1985, Automated production control. In
P. Falster and R. B. Mazumder (eds) M odelling Production
M anagement Systems ( North-Holland: Elsevier Science BV )
pp. 19 35.
BURBIDGE , J . L., 1996, Period Batch Control ( Oxford: Clarendon
Press).
C A RRIE , A. S., 1975, The layout of multi-product lines.
International J ournal of Production Research, 13, 541 557.
C ONSTA B LE , C. J ., and N EW , C. C., 1976, Operations M anagement,
a Systems Approach T hroug h T ext and Cases (London: Wiley).
C ONW A Y , R. W., M A XW EL L , W. L., and M IL LER , L. W., 1967,
T heory of Scheduling (Reading, MA: Addison-Wesley).
DE T ONI, A., and P A NIZZOLO, R., 1992, Repetitive and
intermittent manufacturing: comparison of characteristics.
Integ rated M anuf acturing Systems, 3, 23 37.
DUL MET , M., L HOTE , F., and O RTIZ -H ERNANDEZ , J ., 1997, A
processes classi cation for the representation of production
systems. Fifth International Conference on Factory 2000. T he
T echnolog y Exploitation Process. IEE Conference Publication
no. 435, pp. 496 500.
F RIZE LL E , G. D. M ., 1989, OPT in perspective. A dvanced
M anuf acturing Engineering, 1, 74 80.
GOOD, I . J ., 1965, Categorisation of classi cation. In
M athematics and Computer Science in M edicine and Biolog y
( London: HMSO), pp. 115 128.
GROOVE R , M ., 1980, A utomation, Production Systems and Computer
A ided Manuf acturing (Englewood Cli s, NJ: Prentice-Hall) .
I NGHA M, H., 1971, Balancing Sales and Production (London: British
Institute of Management).
J A IN, A. K., 1988, EX PSYS: o shore structures expert system.
Proceedings of the 5th ASCE ( A merican Society of Civil Eng ineering)
Specialty Conference, Blacksburg, USA, pp. 25 27
J ICHA O, X., 1996, Variability detection and robustness design in
complex production system. Computers & Industrial Engineering ,
31, 775 778.
J OHNSON, L. A., and M ONTGOMERY , D. C., 1974, Operations
Research in Production Planning, Scheduling and Inventory Control
( New York: Wiley) .
J OHNSON, R. V., 1991, Balancing assembly lines for teams and
work groups. International J ournal of Production Research, 29,
1205 1214.
K ITCHING, I . J ., F OREY , P. L., H UMP HRIES, C. J ., and
W IL L IA MS, D. M ., 1998, Cladisticsthe T heory and Practice of
Persimony Analysis ( 2nd edition) (Oxford: Oxford University
Press).
K USIA K, A., 1985, Flexible manufacturing systems: a structural
approach. International J ournal of Production Research, 23, 1057
1073.
L IU, J ., and M A CC A RTHY , B. L., 1996, The classi cation of
FMS scheduling problems. International J ournal of Production
Research, 34, 647 656.
M A CC A RTHY , B. L., and L IU, J ., 1993, A new classi cation
scheme for exible manufacturing systems. International
J ournal of Production Research, 31, 299 309..

Downloaded by [University of Sunderland] at 19:31 29 December 2014

496

B. L. M acCarthy and F. C. F. Fernandes

M A IMON , O., and N OF , S. Y., 1986, Analysis of multi-robot

systems. IIE T ransactions ( Institute of Industrial Eng ineers) , 18,
226 234.
M A LL ICK, R. W., and G A UDREA U, A. T., 1951, Plant Layout
( New York: Wiley).
M A THE R , H., 1988, Competitive M anuf acturing (Englewood Cli s:
Prentice Hall).
M CC A RTHY , L., 1995, Manufacturing classi cation: lessons
from organizational systematics and biological taxonomy.
Integrated M anufacturing Systems, 6, 37 48.
M CC A RTHY , I ., L ESE URE , M., R IDGW A Y , K., and F IEL L ER , N.,
1997, Building a manufacturing cladogram. International
J ournal of T echnology M anagement, 13, 269 286.
N Y S, D. A., 1984, Classi cation of machine tool equipment used
for machining housing type components. Soviet Eng ineering
Research, 4, 39 43.
P ETROV , V. A., 1966, Flowline Group Production Planning (London:
Business Publications).
P UTNA M, A. O., 1983, MRP for repetitive manufacturing shops:
a exible kanban system for America. Production and Inventory
M anagement, 60 88.

P Y OUN, Y. S., C HOI, B. K., and P A RK, J . C., 1995, Flexibility

value as a tool for improving decision making in exible
automation. Eng ineering Economist, 40, 267 282.
SCHMITT, T. G., K L A STORIN , T., and SHTUB , A., 1985,
Production classi cation system, concepts, models and
strategies. International J ournal of Production Research, 23, 563
578.
SIP P ER , D., and SHA P IRA , R., 1989, JIT vs. WIP a trade-o
analysis. International J ournal of Production Research, 27, 903 914
STE CKE , K. E., and BROW NE , J ., 1985, Variations in exible
manufacturing systems according to the relevant types of
automated materials handling. M aterial Flow, 2, 179 186.
W IL D, R., 1971, T he T echniques of Production M anagement
( London: Holt, Rinehart and Winston) .
W IL D, R., 1972, M ass Production M anagement (London: Wiley) .
W IL D, R., 1995, Production and Operations M anagement ( 5th edition) (London: Cassell Eductional).
W OODW A RD, J ., 1965, Industrial Organization T heory and Practice
( 1st edition) ( Oxford: Oxford University Press).
W OODW A RD, J ., 1980, Industrial Organization T heory and Practice
( 2nd edition) (Oxford: Oxford University Press).