Computers in Industry 49 (2002) 9–23

A non-linear redesign methodology for

manufacturing systems in SMEs
J. Bradford*, S.J. Childe
Manufacturing and Business Systems Research Group, School of Computing, University of Plymouth,
Drake Circus, Plymouth, Devon PL4 8AA, UK

Abstract

SMEs often have to adapt their manufacturing systems to changing circumstances such as technology advances and business
growth. This paper presents a new methodology for the redesign of manufacturing systems within SMEs.
This paper will show that most approaches to manufacturing systems redesign are linear and it will explain why this is not the
best option for SMEs. Foremost among these reasons is the uncertainty of the SME environment. Redesign planning that is
conducted at the outset of a change episode may be superseded by events that are impossible to foresee. A further issue for
conventional redesign approaches is the resource poverty that is a widely reported feature of SMEs. The most obvious impact is
on the financial support available for costly technology-led solutions. Resource poverty also includes the available expertise and
managerial time to implement complex methodologies.
The alternative approach proposed uses an iterative model as its basis for change. The basic helical model is extended to
provide four perspectives on systems change. These perspectives focus on structural, people, technology and process issues. It is
these perspectives that make the difference between continuous improvement and systems redesign. The model has been
developed through action research in manufacturing SMEs. Validation was achieved through a longitudinal case study that
provided empirical evidence to support the methodology. A single case was used for validation to provide depth over breadth.
The development of the methodology had been broad through action research into several diverse SMEs and it was considered
that more detailed case study would add balance to the research. The depth afforded by a single longitudinal study also allowed
for greater scrutiny of the methodology under real world conditions.
# 2002 Elsevier Science B.V. All rights reserved.

Keywords: Design; Redesign; Manufacturing systems; SMEs

1. The development of a redesign approach Without considering the activities that make up the
design process we can consider the manner in which
The methodologies that are traditionally posited for those activities are structured as the design strategy [1].
systems design are derived from those concerned with Jones goes on to state that: ‘clearly a major objective in
design of new artefacts. This is different from redesign design methodology is to make designing less circular
of manufacturing systems. and more linear’ [1]. The principal obstacles to linear-
ity are identified as the unpredictability of the relation-
ships between parts of the problem and the fact that
*
Corresponding author. Tel.: þ44-1752-233706. these relationships are variable over time [2]. Ramirez
E-mail address: jbradford@plymouth.ac.uk (J. Bradford). [3] contends that circularity is a feature of the real world

0166-3615/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 6 - 3 6 1 5 ( 0 2 ) 0 0 0 5 5 - 6
10 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

since the critical sub-processes cannot all be identified enon and the tenacity with which designers will con-
before design begins. In looking at groups as problem tinue to hold the initial assumption, even after it has
solving units, Larson and Christensen [4] consider the been shown to be false. The ‘primary generator’ is
solution of poorly structured problems to require itera- independent of the feasibility study and may guide that
tive or circular approaches. With particular respect to study to find the ‘correct’ problem to which it is the
architectural design, Lawson [5] suggests that designers solution. The linear approach described earlier assumes
move freely between three kinds of activity, analysis, that the designer is able to approach the task with an
evaluation and synthesis. Moreover, Lawson contends open mind and all the facts, this is rarely the case as
that there is no fixed path but that the designer ‘. . . recognised by Larson and Christensen [4].
put[s] it all together for [himself ]’ ([5], p. 38). While Lawson describes a highly non-linear
The importance of these design strategies is that approach to design, he makes no claim to be able
there are different ways of structuring the design to guide the design process, only report which phases
process. Different situations may be more amenable the designer went through. The only strategies that
to certain strategies though there is an assumption that claim to be able to guide and help the designer are
linearity is the preferred strategy. largely linear in nature [46–48,56]. Apart from the
Jones [1] has suggested that a wholly logical meth- similarity that can been found between manufac-
odology does not exist to solve design problems but turing systems redesign methodologies, they also
that this does not prevent solution within the human have a common missing element. There is often little
brain. Jones builds upon the work of Asimow [6] who or no consideration of the human factor. The meth-
identified four stages in design that cover feasibility, odologies recognise that social issues are important
preliminary design, detailed design and planning (see in the redesign of manufacturing systems but tend
Fig. 1). Jones describes the planning stage as evaluat- not to give those issues much consideration when
ing and altering the design concept to suit the require- presenting their design outline. This could be due
ments of production, distribution, consumption and to the technical genre from which these approaches
product retirement. evolved [12].
Several authors (e.g. [7–9]), emphasise the require- Manufacturing systems have been viewed using a
ment for feedback loops within the design process. mechanistic model [13] which presupposes that the
While these are not included in Fig. 1, they are usually system can be fully understood through analysis. This
shown as taking problems that have occurred in one reductionist approach requires that all the elements
stage, e.g. detailed design and feeding them back in to in the system be decomposed and the relationships
earlier stages e.g. preliminary design or the feasibility between those elements defined through largely cause
study. The important point being made is that issues and effect laws. The complexity of manufacturing
may arise that were not foreseen at the early design systems is far greater than that of mechanical, elec-
phases, these issues have to be resolved but this can trical or architectural systems [14].
only be achieved by revisiting those early phases.
While this is seen as a vital element in design, it is 1.1. Helical design
a modification to the basic, linear design process.
In a further proposition on routes for the design As early as 1939, Shewhart [45] had described the
process, Lawson [5] describes what Darke [10] calls need to move from the ‘old’ way of manufacturing
the ‘primary generator’. In this instance, the designer with a linear progression to a ‘new’ way with a cyclic
makes an assumption about the final design and then process. In 1984, Deming [44] had described the
tailors the rest of the design process around this initial cyclic design methodology as being better than the
assumption. Rowe [11] has also reported this phenom- linear model. This is supported by Ramirez [3] and

Fig. 1. A linear approach to design [1].

 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 11

directly contradicts Jones [1]. Pressman [15] develops analogy that while caterpillars and butterflies are
an idea by Boehm [16] for a spiral design model for manifestations of the same creature they cannot be
software engineering. In this model the stages are meaningfully compared with each other as the differ-
identified as being (1) planning, (2) risk analysis, ences are too great. In considering the implementation
(3) engineering and (4) customer evaluation. The of total quality management (TQM), Ghobadian and
customer evaluation feeds in to the next planning Gallear [21] conduct a comprehensive review of litera-
phase. At the risk analysis phase there is a go/no-go ture regarding the implications of organisational size.
decision based upon the analysis. In particular, they suggest that there are ‘. . . significant
structural differences between SMEs and large
1.2. Design as learning organisations, . . .’ ([21], p. 127).
Ghobadian and Gallear identify ‘resource paucity’
Design is primarily concerned with creating some- as the most serious disadvantage faced by the SME
thing new. This means that the designer must begin considering implementation of TQM. The term
with an end-state at some point in the future and ‘resource’ is used in its widest sense to cover not only
determine those activities that will lead to such an financial resources but those of knowledge, technical
end-state [1]. The greater the difference between the expertise and management time [22–24]. This lack of
current-state and the end-state the greater the number spare resources leads to SMEs being very risk aware
of activities or scale of those activities required to with regard to committing resources. This would
navigate between the two. Since all this activity takes suggest that for an approach to be successful with
place in the future there can be no way of knowing SMEs, it should have an explicit concern for resource
that all the planned activities will proceed, or have sensitivity, i.e. risk aware. Wynarczyk et al. [25] builds
outcomes that are exactly as planned. on Penrose’s [20] in suggesting that a primary out-
To this extent the designer is continually learning come of this resource paucity is increased uncertainty
about the system that is being designed. Constructivist for the SME.
learning suggests that a learner begins from a base
of prior knowledge upon which further experience 2.1. Uncertainty
is integrated [17]. This is carried out through a
cyclic process such as the (1) discovery, (2) invention, Storey [26] argues that small firms are subject to
(3) production, (4) generalisation cycle of Argyis and greater external uncertainty and greater internal
Schon [18], the (1) observation, (2) abstract concep- consistency of motivation and action than large
tualisation, (3) test, (4) concrete experience cycle of firms. The external uncertainty stems largely from
Kolb [19] and the (1) observe, (2) assess, (3) design, the relatively large size of many customers of small
(4) implement cycle of Kofman used by Kim [43] in firms, which gives the small firms little bargaining
his model of organisational double loop learning. power with their customers [25,27]. This in turn
A cyclic learning approach allows us to consider the leads the SME to be more reactive to the business
manufacturing system as an ambiguous social orga- environment than their larger siblings [28], thus, an
nisation. We cannot know everything about the rela- SME will have to redesign itself more extensively
tionships between individuals but we can learn about and frequently than a larger business. Joyce et al.
their behaviours through experience. We can never [29] suggest that one coping mechanism to deal with
consider this learning process complete as the indivi- external uncertainty is the phenomenon of ‘niche
duals will be constantly developing and the culture of hopping’. While each niche may only provide tem-
the organisation will develop with them. porary respite, the ability to keep ‘hopping’ ensures
the survival of the company. Since the scale of
change related to each hop (or evolution of the firm)
2. Understanding SMEs is relatively large compared to the spare resources,
this will cause increased uncertainty and the like-
Large and small firms have been identified as being lihood of failure is much greater than for larger
fundamentally different by Penrose [20]. She uses the firms.
12 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

2.2. Redesign approach 3. Redesign methodology

It is clear that one of the greatest constraints that The methodology takes as its basis the assumption
SMEs face, financial resources, is an external one over that strategic intent exists within the company.
which they have little influence. While a methodology Whereas the planning activity in a linear methodology
cannot provide influence over external constraints, it seeks to define the entire manufacturing system to be
should enable the SME to evaluate a proposed change realised, the iterative approach seeks only to define the
against the available resources so that the business next change. Each iteration uses the strategic intent to
does not overextend itself. For this reason, there determine the next focus of effort.
should be some risk- or cost-benefit assessment to There is no evidence to suggest that the first itera-
ensure that the resources of the business are equal to tion should take any particular perspective nor is there
the proposed change. This applies for all resources any evidence regarding the weighting of perspectives
available to the SME: managerial time, managerial [41]. Where the perspectives are described, they are in
skills, technical skills, manpower and money no particular order and the order they appear should
[23,30,31]. These highlight the fact that techniques not be taken as indicating that any preference should
suitable for large companies are not suitable for be shown. The methodology described here consists
smaller companies [32]. of planning, risk assessment, action and evaluation
The twin issues of ‘niche hopping’ and firm evolu- phases. These will be described theoretically and
tion suggest that continuous redesign of the manufac- through the use of a longitudinal case study.
turing system is likely to be a feature of SME existence
[33–35]. While continuous improvement (CI) is 3.1. Planning
widely accepted in the academic literature, Gieskes
et al. [36] have identified that SMEs have difficulty Since this methodology is concerned with the rede-
implementing formal CI approaches. sign of manufacturing system, it assumes a strategic
Wiele and van der Brown [37] identify the unease intent within the business. The methodology is not
that SMEs have with current formal redesign meth- concerned with business or manufacturing strategy
odologies. In looking at how SMEs adopt TQM, they formulation, it seeks to reflect strategy through itera-
cite earlier work that suggests that SMEs are uncom- tive redesign of the manufacturing system. Considera-
fortable with formal methods [38–40]. Any methodol- tion of strategy is used within the planning phase to
ogy that is presented to SMEs should not, therefore, gauge progress towards wider business goals. In this
appear as a large and complex approach that will manner, the current iteration will neither negate pre-
lack immediacy of applicability. The uncertainty that vious work nor advance counter to business strategy.
SMEs find themselves subjected to implies that a Within the planning phase of Fig. 2 are the four
redesign methodology should be capable of rapid perspectives (structure, people, process, technology)
conversion of problem situations into solutions and that have been developed from the work of Leavitt and
the implementation of those solutions. Bahrami [41]. The company will have a view on where
During the planning phase the four perspectives their problems lies with respect to these four perspec-
(structure, people, process, technology), developed tives. This is likely to be expressed as a conceptual
from the work of Leavitt and Bahrami [41], were solution or primary generator [10]. There is nothing
considered. Structural issues relate to the organisation within the methodology to directly challenge this
of the company; where authority lies and who is initial view, except to make explicit the existence of
responsible for what. People issues are such things other perspectives. Care should be taken to ensure that,
as morale, communication and other factors that deal while multiple perspectives are not excluded, appro-
with the humanistic side of business life. Process priate focus is maintained relative to availability of
issues consider the business processes, production resources.
processes and layout of the factory. Technology issues Having determined the focus of the change pro-
consider the technology within in the business and also gramme a change team should be assigned to carry
the information and control of that business. out the changes and to run the iteration. This is no
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 13

Fig. 2. Methodology diagram.

different to conventional systems change methodolo- important to know the activities that are being carried
gies; a change champion, owner and team should be out by the process, it is the ability of the people to
named in the planning phase. Any budget that is carry out that process that is assessed with the people
available should also be specified together with any perspective. There may be informal teams or social
constraints that the change has. It is vital that a time groups within the business that enable information to
frame is specified as it is easy with any approach be transmitted more effectively than the formal report-
to let actions slip, especially when business situations ing structure, in a similar manner, peer group pressure
can change quickly and resources are committed may be more powerful than the formal disciplinary
elsewhere. structure.
As will be described, the fourth phase is an evalua- There are numerous tools for analysing the people
tion one. To be able to carry out that evaluation some and their issues within an organisation. The important
form of metrics are required, which must be specified feature of note is that there is an explicit requirement
at the planning phase so that they can be deployed in for the people perspective to be considered at the
the action phase and reflected upon in the evaluation planning phase and for it to be taken into account
phase. There is no theoretical requirement to specify when determining the perspective to continue the
qualitative metrics (value based, i.e. better than, easier, redesign activity.
less confusing) over quantitative metrics (objective,
i.e. 51%, 5 min, 136 parts/min) providing that all are 3.3. Process
happy with the measurements chosen.
In Leavitt and Bahrami’s work [41] this perspective
3.2. People is labelled ‘task’. They see this as the external focus
of managing change, of matching the tasks of the
In adopting a people perspective the change pro- business to market requirements. However, current
gramme is looking at the skills, competencies, morale methodologies focus almost exclusively on the inter-
and degree of job satisfaction experienced. While it is nal activities and technology within the system under
14 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

investigation. There is a demonstrated requirement There may be further impacts on the organisatio-
from the field work to understand the internal activ- nal structure as decision making moves between
ities and processes within the system. people and the traditional authority and accountabil-
In considering the manufacturing system it is ity structures no longer reflect the practice of the
argued that a systemic consideration should adopt a business.
holistic focus. This leads to the consideration of the The redesign of information systems and the tech-
manufacturing system as a series of connected wholes nology of manufacturing systems is well developed.
or processes. An example is the CIM-OSA reference The important feature to note here, as with the pre-
model of a business system that identifies three prin- vious two perspectives, is that this is but one perspec-
ciple processes; those of manage, operate and support. tive that should be considered in concert with the
These are further sub-divided such that operate con- others. Whichever is chosen as the focus for any
tains ‘get order’, ‘develop new product’, ‘fulfil order’ particular iteration should not be chosen to the exclu-
and ‘support product’. These processes contain the sion of the others.
activities which enable the tasks of the system to be
undertaken. Thus, considering the tasks of the busi- 3.5. Structure
ness from a systemic viewpoint suggests a process
perspective. For these reasons, the process perspective In considering a structural perspective the change
is concerned with the business processes and activities programme is looking at the areas of responsibility,
that the business undertakes internally. location of authority and the route of decision making.
This is the organisational structure of the business and
3.4. Technology how it provides the support structures for the pro-
cesses, people and technology. This may not be for-
The technology perspective considers the control mally expressed in smaller businesses but there will
and information elements of the system and how still be reporting channels, lines of authority and
technology facilitates their implementation. The responsibility. Where these are informal or have been
information element of this perspective is concerned superseded over time, the organisational approach will
with the flow of information around a business and help the business to gain clarity over these issues and
how that information is used to control the processes to determine the exact structure required. Larger
of that business. It is primarily concerned with control businesses may find that they have changed signifi-
of the process and the information and technology that cantly while their organisational structure has not kept
is used to administer that control. There is an asso- pace. This can lead to excessive managerial structures,
ciated requirement to consider all the technological unclear job roles, confused authority remits and infor-
facets of the system. Since this is a manufacturing mal power structures being formed.
system, there will be manufacturing machinery that Organisational change is a well established disci-
will form a significant technology base within the pline that specialises in the analysis of business orga-
business. nisations, their strengths, their weaknesses and the
This manufacturing technology will have a pro- optimal method for getting from one state to the other.
found impact on the ability of the system to react This methodology will not evaluate the different
to market needs and the perceptions of customers approaches to organisational change except to point
regarding the processes that should be undertaken, out that they should be used with the iterative
the time scales that the business should operate on and approach described here. That will ensure that the
the costs involved. There will also be people issues redesign is systemic and not entirely focussed on one
arising from a technological perspective that sees a perspective to the detriment of the others.
requirement to introduce new technology and infor-
mation management systems. The introduction of 3.6. Risk assessment
new information management technology may also
have a profound impact on the organisational struc- The principal purpose of this phase is to reach a go/
ture of the business as communication patterns alter. no-go decision for the proposed change identified in
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 15

the planning phase. The risk assessment should be 3.8. Evaluation

appropriate for the perspective being adopted and in
line with company norms in managing risk. Some Having carried out some actions in accordance with
companies will accept higher levels of risk in antici- the plan, there is a requirement to evaluate the out-
pation of greater payoffs if successful, others will come of those actions. This is where the metrics
adopt a more conservative approach preferring to become important. If the measurement system is
minimise their risk exposure. not accurately thought out then the changes will be
Any change carries some degree of risk or cost for evaluated against incorrect criteria. It is important that
the business. The purpose of the risk assessment the evaluation is carried out while the project is still
phase is to identify those risks and determine the fresh so that objectivity can be used. Too long a delay
probability that the benefits will out-weigh the costs. may result in people taking an overly optimistic or
In planning the change, consideration should be taken pessimistic stance in analysing the change.
of the likely costs of changing the organisation, Redesign is a learning activity and, therefore, each
introducing training, new technology or re-organising iteration will be constructed upon the learning that
activities. These will be estimated costs but they occurs during previous action phases. The evaluation
should be sufficiently accurate for the business to phase is an opportunity to reflect on the actions that
be satisfied that they are not undertaking an unduly have gone before and the perspectives adopted to
risky venture. This level of perceived risk will be determine what can be learnt for the next planning
unique to individual businesses as will the acceptable phase. As the company gains in experience, the knowl-
level of perceived risk beyond which actions are edge base upon which choices about the appropriate
considered too risky to undertake. perspective will be made will grow. While this may not
Where a particular primary generator for ideas is lead to more accurate forecasting of the appropriate
proving to be unfruitful then this phase is there to perspective, it will lead to greater understanding
catch that and suggest that a different perspective be regarding the importance of the different perspectives,
adopted. Where previous work has been carried out their interrelationships and the implications for the
(through previous iterations) the benefits should be manufacturing system.
weighed against the costs involved incurred. Pareto
[49] analysis may be used to judge when the 80% 3.9. Planning
benefit level has been reached and the remaining 20%
can be left for another day [42,50]. Future iterations Having carried out an iteration of the methodology,
with a different focus may realise this remaining subsequent planning phases will have a slightly dif-
benefit as part of their change. ferent composition. In addition to the strategic input
there will be the results of the evaluation phase. These
3.7. Action will lead the discussion on focus and aims for the
iteration to come.
It is important to carry out those actions proposed in Where the evaluation may have identified a change
the planning phase and justified in the risk assessment episode that is beginning to lose momentum, it would
phase. The actions carried out should be recorded so be appropriate to investigate a different perspective
that they are available for evaluation and the continual to frame the following iteration. This is a valuable
improvement of the redesign process. The planning element as it prevents stagnation and self-limiting of
phase will have specified a time plan and metrics the change process.
against which the action phase can be measured. It
may, at this point, be worth employing a linear project
planning aid to ensure that the tasks identified in the 4. The methodology in action
action plan are carried out according to the plan. This
may prevent slippage and ensure that the action phase The case study was conducted at a manufacturing
does not cause the project to grind to a halt because no facility based in the south-west of England. The case
action has been taken. study spanned eight iterations over a period of 9
16 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

months. The time spent on any particular iteration production cell. Within this development there were
varied depending upon the prevailing business climate six identifiable iterations demonstrating the methodol-
and resource availability. ogy in practise. Rather than describe the case in its
A single longitudinal case design [51,52] was chosen entirety, the salient points from each iteration will be
to provide the depth of research data that would not grouped under the relevant methodological headings.
be present with other designs. This depth and richness
of data is a primary reason for choosing the case 4.2. Planning
study method in the first place and ‘. . . single case
studies can be influential, especially when they are The first iteration developed a plan for relocating
purposely non-representative, perhaps reporting major the production process that manufactured doors. The
innovations . . .’ [53]. change was concerned with the business process of
The longitudinal approach also provides for an manufacturing doors. The plan was developed by the
understanding of the phenomenon under investigation maintenance engineer with much of the detail being
as a dynamic rather than static process [54]. Design is decided by the team leaders. In this manner, the plan
as an ongoing process that has much in common with gained from significant ‘buy-in’ from the shop floor
cognitive learning theory. The depth gained from a operators. In parallel with the process focus was a
single, longitudinal case study would provide more recognition that technology issues would be vital
useful evidence on the validity of the methodology since the control of the new door line would be
than a quantitative survey approach. entirely under the management of one team leader.
A more practical reason for a single study was that The production of doors had previously been dis-
the field being addressed is huge, estimates put the tributed between two different team leaders. The co-
proportion of SMEs in the economic environment as ordination between the leaders had been managed by
being >99% [55]. To gain access to a realistic sample the factory manager, this activity was no longer
population and to conduct a quantitatively meaningful being carried out but the leader of the door produc-
analysis of application of the methodology in such a tion line had to control the whole line. The new
huge sample would require more resources than were layout would allow visual control to be applied since
available. all the stages in the manufacture of doors would be
in one controlled area. The scheduling of door
4.1. Example change initiative production would not be effected by the change,
at least until the capacity of the new line was fully
Within the case company, there had long been a realised.
concern to improve the efficiency and productivity of The second iteration modified the layout of the
various product lines. For numerous historical reasons, door line so that suggestions from the shop floor and
the door line was distributed around the factory with team leaders were included. This produced a floor
machines being fitted in as well as possible. The two plan that was a refinement of the original. The new
significant effects of this was that the production of movement would be carried out during normal pro-
doors came under the responsibility of three team duction since the changes were minor and no new
leaders, none of which had ownership of the final services were required. Disruption to production
doors and individual doors were subject to high levels would be minimal. With the continued downturn in
of material handling while being transported around orders there was sufficient capacity within the line to
the factory. While improvements had been made to cope with minor down time while the modifications
individual machines the disjointed production system were carried out.
for making doors was incapable of significant The third iteration was to remove glazing from the
improvements until it was reunited as a recognisable main shop floor and include this activity in the new
system. door line. This suggestion led to others and a free
Developments within the business had cleared a discussion ensued around the organisation of the
space on the shop floor that provided an empty space activities in the process that would provide the most
into which the door line could be established as a appropriate use of company resources. In addition to
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 17

the purely process issues there were training issues 4.3. Risk assessment
that resulted from the need to move personnel towards
a multi-skilled environment and the freedom to plan The major objective was to improve the layout so
and organise their own work patterns depending on the that the production line would be more efficient. It was
jobs present. planned to produce savings of several minutes per
A people perspective was quickly identified as door. Production lost whilst moving equipment would
being the forth change focus since the people within be recovered through reduced scrap and production
the system did not have the skills to carry out the new times. Individual times for door production were not
tasks that were being assigned to them. The new recorded as several operations were shared with other
business process would require operators to have production lines. This made it impossible to separate
the confidence and authority to stop work and move out the door production figures.
to any bottlenecks to help out as required. There was a While there was no budget available for the move,
degree of animated discussion around the subject of there were considerable savings to be reaped from the
the change, scope and associated issues, which was new layout. These savings could be offset against any
finally resolved when it was clarified that this was an materials that would be required. The only significant
iterative process and they were not expected to get it cost was to be the labour input from the maintenance
100% correct the first time. There would be a period of engineer. Having established that there were no
learning and development, during which time other inhibiting risks the maintenance engineer was
suggestions that had been floated could be incorpo- authorised to begin the next phase of the redesign
rated into the plan. and implement the actions determined in the plan-
The training could be carried out using in-house ning phase.
staff and cross-training could be used to extend the Throughout the risk assessments, costs were mini-
skills within the production workforce. The metrics mal since the internal maintenance engineer would
chosen were capacity (time to manufacture, queuing carry out the work. Spare capacity within the line
time) and reduction in waste. These were already meant that any disruption could be quickly recovered.
known through previous analysis projects and could The new production layout only called for 85% uti-
easily be measured in the new system. They also lisation of the available space; the remaining 15%
represent clear and visible measurements that could would be used to absorb any work in progress (WIP)
be related to the changes made on the production and allow some manoeuvring room when re-position-
line. ing the machinery.
The fifth iteration required further modification of In assessing the risks with the people focussed
the shop floor. This was to provide more working change it was recognised that a reduction in capacity
space with less travelling between stations. The plan would occur. The benefits would be that the equip-
was two-fold. To move the glazing area slightly and to ment would be used properly and, in addition to
re-structure it to follow a more orderly flow pattern. increased productivity, scrap levels would be
The new layout would establish a ‘U’ shape from reduced. It was also a voiced opinion that the
stores to goods outward. increase in flexibility of the staff would have a dual
The final iteration, in the study case, was to transfer improvement. The possibility for job rotation, and
the new best practice from the door line to the rest of hence, enrichment was seen as a significant supple-
the factory. This changed the focus to more structural mentary benefit from this cross-training programme.
and technology issues concerning the control of the The current manufacturing and assembly operations
new layouts and staffing structures to take account of were relatively repetitive and some variation was
new responsibilities. The activities of making win- considered to be a good thing. The flexibility of
dows and doors was not fundamentally altered, these the staff also reduced the dependency of the system
were largely dictated by machine capability and on a small number of key personnel. If one operator
agreed specifications. The layout, management and was taken ill or had holiday booked, the others would
control of the people and machines was a structural be able to adapt and maintain the efficiency of the
and technological design problem. system.
18 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

4.4. Action important by the operators it was recognised that some

would not wish to learn all of the equipment. It was
The first iteration was conducted in the order that made clear that this would not be the subject of
material would flow around the final system. All the negative appraisals, though this may change in the
work was carried out by the maintenance team with future. This action was being continued externally to
the assistance of the operators. The move was spread the change programme since it took some time for all
over 1 week of production time. In this manner, the operators to progress through the whole door line.
maintenance engineer estimated that only half a day Although the change programme had achieved its
had been lost in production. While the move had gone aims for that iteration, the training continued over
smoothly the methodology required an evaluation several weeks. The evaluation of that iteration and the
phase to check the new situation against the forecast planning for the next began before all the operators
gains and benefits. had been fully trained.
The floor plan was implemented immediately. The changes were conducted over a period of 2
Minimal alterations were required to the services since days. No major incidents were reported and everything
the machinery was not being moved far. Machinery proceeded according to the plan developed. Each jig
was typically moved a couple of metres in one direc- was moved separately and production shifted to a
tion or another. The short distances coupled with the temporary jig or table until the equipment was ready
new cell layout allowed for the moves to take place in the new location. All the temporary jigs and tables
during machine slack times. During these times, the functioned as required. During the change there was
operator would move to another machine. In this no measurable change in production efficiency or
manner, the production output of the door line was output.
not affected by the action phase. The planning board was installed and the new cards
The door line equipment was moved around such released. The approach worked well with all orders
that the beading and glazing was carried out as part of being tracked to ensure that they are not getting too
the door line and not at a separate glazing station. This old. Once an order was on the shop floor for more than
led to equipment being freed up which in turn led to 4 weeks, it was scrapped. No scrapped orders were
more suggestions regarding the cascading of this later called for. These were usually orders that had
improvement throughout the factory. The changes been duplicated through an office error, or for a job
were overseen by the factory maintenance engineer. that had been cancelled without the shop floor being
The training was carried out by the existing glazing advised.
staff. This took the form of a half day session for the The new storage areas were produced and labelled
operators that were to initially take over the operation. clearly. The areas allowed for large units to be stacked
This training was carried out on the machines in the such that they were not damaged. Smaller units were
door line and using door assemblies. There is no stored above the normal working area so that they
formal appraisal system that is linked to training so were completely out of the way.
it was left to the existing glazing staff to assess that the
required level of competence was achieved. Once 4.5. Evaluation
training was complete the glazing staff handed over
the complete operation to the door line which now An integral element of the relocation was to bring
operated independently from the other lines on the the entire door line under a single team leader. This
shop floor. had major implications for the structure of the orga-
Having completed the training, the operators began nisation. The responsibility for door production had
training their colleagues in each other’s tasks. An previously been distributed between three team lea-
informal job rotation scheme began with operators ders, with each having the opportunity to apportion
moving around once they felt confident on each blame to the others. Now that the line was under a
machine. The aim was to have all the operators single person this was no longer possible.
capable of completing any of the five major activities This change to the organisational structure of the
involved in door manufacture. While this was seen as business was recognised within the project but was not
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 19

acted upon. The business was facing a period during changes. The control of the production cell was largely
which there was the potential for significant changes to visual with doors being progressed through the system
the organisational structure. It was felt that to begin a as the next operation became free with little scope for
redesign looking at structural issues would be pre- improving on this. The organisational structure of the
mature and had the potential to create confusion over cell was functioning adequately and there was general
the direction the business was going in. reluctance to change this given the potential changes
In parallel to the consideration of a structural that were on the horizon. From a people perspective,
change iteration the line was ‘bedding in’ while there was real scope for improving morale and job
problems were ironed out. It quickly became apparent satisfaction through cross-training. This could also
that insufficient room had been allowed for operators improve productivity through flexibility.
to move and handle sections of profile. Production To achieve this flexibility, the new staff required
efficiencies were not being realised since the flow of formal training from the existing staff. This was
material was hampered by the cramped conditions. As suspected at the planning meeting but it was thought
an example, two machines were situated so that the that they might be able to pick it up as they went along.
output of one could be passed directly to be processed While the tasks are not too complex, working on the
by the second. Unfortunately, the parts to be pro- equipment does required specialised training.
cessed were extruded stock of such length that both As a result of their training the operators on the door
machines could be operated together. So, one part line are now more aware of the difficulties inherent in
would have to progress through both machines before glazing. They have also experienced the problems that
the next could start, effectively halving the utilisation used to result from beading being cut incorrectly.
on both. Since they are now carrying out their own beading
The new layout was an immediate success. This they are able to cut the material to fit and correct errors
success was both in terms of productivity, which immediately. They also express greater job satisfac-
increased, and also morale. During a brain storming tion and increased feelings of professionalism since
session after the layout improvement, the issue of they are responsible for the product throughout its
storage arose as the next required change. As production life.
described, storage did not easily fit into the four The changes to the door line included a shift in
perspectives used in the methodology. This caused a responsibility for production and quality. This used to
degree of unease until it became clear that storage was reside with several team leaders since a door would
not a cause of problems within the manufacturing pass through many areas during production. There
system but an effect of other problems. Further dis- was a temptation to abdicate responsibility to one of
cussion revealed that the problem was not the storage the other areas. The new layout placed all the respon-
but the bottlenecks that led to the WIP building to the sibility with a single team leader. It also gave that
point where storage became an issue. The recent team leader the authority to control the production
changes to the layout had improved the efficiency line that made the doors. This was a significant change
of the door line and this was causing problems for to the structure of the business and its organisational
glazing. This now represented an issue that resided culture.
within the process perspective, a bottleneck had In an interview with the maintenance engineer and
moved and was causing the storage issues that were informal talks with the door production operators
highly visible. there was a noticeable shift in the culture. There
The productivity per worker was significantly was a new sense of identity. The increasing flexibility
increased through the integration of beading and of the operators meant that they were able to move
glazing into the main door line. The most significant around the production line. This enabled them to help
improvement was the reduction in WIP. The reduc- out their colleagues when problems arose. Problems
tions in WIP triggered a reduction in scrap and rework could now be tackled by the combined efforts of the
since there was also less material handling and oppor- team rather than simply awaiting the arrival of the
tunity for damaging products. There were few options maintenance engineer. To demonstrate this new team
for improving the productivity further through layout spirit the maintenance engineer recounted an example
20 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

that ‘. . . they now applaud late comers back from Having determined the focus of the change pro-
lunch . . .’. gramme a change team was assigned to carry out the
Prior to the changes there was no formal definition changes and to run the iteration. No budget was
of the responsibilities of each team leader. This made available but commitments were received from the
it difficult for the business to carry out a change to the maintenance engineer. In each instance a time frame
organisational structure. It was recognised that struc- was specified for the change to take place. Measure-
tural issues were of concern and that there were ments were determined as part of the strategic aims of
implications from the changes already conducted. the company. These included 100% on-time deliv-
The business situation at that time was particularly eries, 60% reduction in lead time, 40% reduction in
volatile. It was concluded that formal changes to the scrap and 40% reduction in re-work.
organisation would have to wait until the business
situation resolved itself. 5.2. Risk assessment
It was identified that the changes made to the door
line could be replicated elsewhere within the factory. The principal purpose of this phase is to reach a go/
This was first mooted at an earlier planning phase no-go decision for the proposed change identified in
when the glazing and beading was incorporated in to the planning phase. The risk assessment should be
the door line. There was scope for changing all the appropriate for the perspective being adopted and in
working practices to reflect this improvement. line with company norms in managing risk. In each
The morale of the shop floor had increased and the iteration, the company had to satisfy itself that the
flow of products was noticeably more natural. There risks inherent in the change were outweighed by the
were no measurements prior to the change that estab- identifiable benefits.
lished the time spent moving material but the shop Pareto analysis may be used to judge when the 80%
floor report that it is easier to operate the line. benefit level has been reached and the remaining 20%
can be left for another day [42]. Future iterations with
a different focus may realise this remaining benefit as
5. Summary part of their change. For the most part non-financial
measures were used in evaluating risk. In one parti-
Over the space of six iterations the case company cular instance, a change was planned for the shop
adopted three of the four perspectives while redesign- floor. A previous change to the shop floor had released
ing their manufacturing system. The fourth perspec- 15% of the available space and this, together with the
tive was identified and considered. It was not adopted planned improvement in efficiency of the new layout,
because business circumstances precluded meaningful was used to justify the new changes.
organisational change. The four phases of the itera-
tive methodology have been shown to function in 5.3. Action
guiding the change activities and ensuring considered
decisions. It is important to carry out those actions proposed
in the planning phase and justified in the risk assess-
5.1. Planning ment phase. The actions carried out should be
recorded so that they are available for evaluation
The case company had a strategic objective of and the continual improvement of the redesign pro-
increasing efficiency while holding costs down. They cess. The changes were conducted and costs mini-
were also keen to investigate the possibility of increas- mised by using in-house staff. This also ensured that
ing capacity, though this was less important. The activities were conducted expeditiously. An impor-
management team within the case company decided tant element of an iterative change is ensuring that
to hold two Kaizen days to generate initial ideas. organisational momentum is maintained throughout.
Previous experience with this methodology has shown The use of internal staff helped to maintain buy-in
that this is not always required to being the process of and make sure that changes did not stall or lose
manufacturing systems redesign. momentum.
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 21

5.4. Evaluation was able to provide solutions that were sensitive to

the resources available. This ensured that suggestions
Redesign is a learning activity and, therefore, each made were implemented and the change momentum
iteration will be constructed upon the learning that maintained.
occurs during previous action phases. The evaluation During the course of the changes the company was
phase is an opportunity to reflect on the actions that able to make significant improvements to their man-
have gone before and the perspectives adopted to ufacturing system. The new production layout has
determine what can be learnt for the next planning reduced the production times by approximately
phase. As the company gains in experience, the knowl- 2 min. By moving an activity within the production
edge base upon which choices about the appropriate line it is estimated that £5000 p.a. (30%) has been
perspective will be made will grow. While this may not saved through scrap reductions. The operators have
lead to more accurate forecasting of the appropriate been quoted as feeling ‘. . . more professional . . .’ with
perspective, it will lead to greater understanding full ownership over their products.
regarding the importance of the different perspectives, The ideas that were generated during the planning
their interrelationships and the implications for the phase produced plans that addressed process, technol-
manufacturing system. ogy and people perspectives. During the span of the
In evaluating the effect of training, it was agreed case study all the perspectives were identified as
that the operators were now more aware of the whole requiring design action. While structure issues were
production process. They also express greater job identified, the business climate was such that the
satisfaction and increased feelings of professionalism company decided not to address them. In conjunction
since they are responsible for the product throughout with other developments within the business the cur-
its production. rent progress shows a 97.92% compliance with on-
In an interview with the maintenance engineer and time deliveries, 33% reduction in lead time, 30%
informal talks with the production operators there was reduction in scrap and 38% reduction in re-work.
a noticeable shift in the culture. There was a new sense The iterative methodology presented here was suc-
of identity. To demonstrate this new team spirit, the cessfully used to redesign a manufacturing system.
maintenance engineer recounted an example that ‘. . . Issues relating to people, process, technology and
they now applaud late comers back from lunch . . .’. structure were identified and dealt with.
The factory maintenance engineer further commen-
ted that ‘. . . people used to resist change simply
because it was change . . . now operators and leading References
hands are asking when their area will be changed . . .’.
This shift in organisational culture was not planned but [1] J.C. Jones, Design Methods—Seeds of Human Futures,
it does represent a beneficial change. The ease with Wiley, London, 1970.
which the methodology can be communicated is such [2] J. Luckman, An approach to the management of design,
Operational Research Quarterly 18 (4) 1967.
that the shop floor are now driving many of the [3] M.R. Ramirez, Engineering vision: considerations in a
changes with management providing a guiding role meaningful approach to conceptual design, Artificial Intelli-
and assessing the risks of each iteration. gence for Engineering Design Analysis and Manufacture 10
(1996) 199–214.
[4] J.R. Larson Jr., C. Christensen, Groups as problem-solving
units: towards a new meaning of social cognition, British
6. Conclusions Journal of Social Psychology 32 (1993) 5–30.
[5] B. Lawson, How Designers Think, The Design Process
The methodology was successfully deployed within Demystified, Architecture Press, Oxford, 1997.
the company. Activity was initiated through focused [6] M. Asimow, Introduction to Design, Prentice-Hall, New
sessions with shop floor staff and business managers. York, 1962.
[7] D. Bennett, Production Systems Design, Butterworths,
The methodology proved robust in application. London, 1986.
Although the company was experiencing a period [8] D. O’Sullivan, Manufacturing Systems Redesign, Prentice-
of financial resource poverty, the redesign activity Hall, Englewood Cliffs, NJ, 1994.
22 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23

[9] D. Parish, Flexible Manufacturing, Butterworths, London, [31] R.F. Marsh, S.M. Shafer, J.R. Meredith, A comparison of
1990. cellular manufacturing research presumptions with practice,
[10] J. Darke, The primary generator and the design process, in: International Journal of Production Research 37 (14) (1999)
Proceedings of the New Directions in Environmental Design 3119–3138.
Research (EDRA 9), EDRA, Washington, 1978, pp. 325–337. [32] P.G. Ratcliff, Change catalysts in small manufacturing
[11] P.G. Rowe, Design Thinking, MIT Press, Cambridge, MA, companies, Engineering Management Journal (1997) 137–
1987. 140.
[12] Unesco, Social Sciences and Humanities in Engineering [33] B.B. Flynn, R.G. Schroeder, E.J. Flynn, World class
Education, The Unesco Press, Paris, 1974. manufacturing: an investigation of Hayes and Wheelwright’s
[13] J. Gharajedaghi, R.L. Ackoff, Mechanisms, organisms and foundation, Journal of Operations Management 17 (1999)
social systems, in: H. Tsoukas (Ed.), New Thinking in 249–269.
Organisational Behaviour, Butterworths, Oxford, 1994, [34] A. Gunasekaran, L. Forker, B. Kobu, Improving operations
pp. 25–39. performance in a small company: a case study, International
[14] L.R. Pondy, I.I. Mitroff, Beyond open systems models of Journal of Operations and Production Management 20 (3)
organisations, in: B.M. Staw (Ed.), Research in Organisa- (2000) 316–335.
tional Behaviour, JAI Press, Greenwich, CT, 1979, pp. 3–39. [35] T.I. Savolainen, Cycles of continuous improvement: realising
[15] R.S. Pressman, Software Engineering: A Practitioners competitive advantages through quality, International Journal
Approach, McGraw-Hill, New York, 1992. of Operations and Production Management 19 (11) (1999)
[16] B. Boehm, A spiral model for software development and 1203–1222.
enhancement, Computer 21 (5) (1988) 61–72. [36] J.F.B. Gieskes, H. Boer, F.C.M. Baudet, K. Serferis, CI and
[17] T.G. Sticht, Comprehending reading at work, in: M. Just, P. performance: a CUTE approach, International Journal of
Carpenter (Eds.), Cognitive Processing in Comprehension, Operations and Production Management 19 (11) (1999)
Erlbaum (Lawrence), London, 1976. 1120–1137.
[18] C. Argyis, D. Schon, Organisational Learning: A Theory of [37] T. Wiele, A. van der Brown, Venturing down the TQM path
Action Perspective, Addison-Wesley, Reading, MA, 1978. for SMEs, International Small Business Journal 16 (2) (1998)
[19] D.A. Kolb, Experiential Learning: Experience as the Source 50–68.
of Learning and Development, Prentice-Hall, Englewood [38] A. Ghobadian, D. Gallear, Total quality management in
Cliffs, NJ, 1984. SMEs, Omega, International Journal of Management Science
[20] E.T. Penrose, The Theory of Growth of the Firm, 3rd Edition, 24 (1) (1996) 83–106.
Basil Blackwell, London, 1995. [39] G.L. Lee, I. Oakes, The pros and cons of TQM for smaller
[21] A. Ghobadian, D. Gallear, TQM and organisational size, firms in manufacturing: some experiences down the supply
International Journal of Operations & Production Manage- chain, Total Quality Management 6 (4) (1995) 413–
ment 17 (2) (1997) 121–163. 426.
[22] A.A. Gibb, Small firms training and competitiveness: [40] M.M. McTeer, B.G. Dale, Are the ISO 9000 series of quality
building upon the small business as a learning organisation, management system standards of value to small companies?
International Small Business Journal 15 (3) (1997) 13–29. European Journal of Purchasing and Supply Management 1
[23] J.A. Welsh, J.F. White, A Small Business is not a Little Big (4) (1994) 227–235.
Business, Harvard Business Review (1981) 18–32. [41] H.J. Leavitt, H. Bahrami, Managerial Psychology—Mana-
[24] S.M. Yusof, E. Aspinwall, TQM implementation issues: ging Behaviour in Organisations, University of Chicago
review and case study, International Journal of Operations and Press, Chicago, 1988.
Production Management 20 (6) (2000) 634–655. [42] M. Hammer, J. Champy, Reengineering the Corporation,
[25] P. Wynarczyk, R. Watson, D.J. Storey, H. Short, K. Keasey, Harper-Collins, New York, 1993.
The Managerial Labour Market in Small and Medium Sized [43] D.H. Kim, The link between individual and organisational
Enterprises, Routledge, London, 1993. learning, Sloan Management Review, Massachusetts Institute
[26] D.J. Storey, Understanding the Small Business Sector, of Technology, Boston, Fall 1993, pp. 37–50.
Routledge, London, 1994. [44] W.E. Deming, Out of Crisis, Massachusetts Institute of
[27] P. Westhead, D. Storey, Management training and small firm Technology, Boston, 1984.
performance: why is the link so weak? International Small [45] W.A. Shewhart, Statistical Methods from the Viewpoint of
Business Journal 14 (4) (1996) 13–24. Quality Control, Van Nostrand, New York, 1939 (reprinted
[28] N. Siropolis, Small Business Management, Houghton Mifflin, Milwaukee, ASQC 1980).
Boston, MA, 1997. [46] G. Broadbent, Design in Architecture, David Fulton Publish-
[29] P. Joyce, A. Woods, T. McNulty, P. Corrigan, Barriers to ers, London, 1998.
change in small businesses: some cases from an inner city [47] D.G. Ullman, The Mechanical Design Process, McGraw-Hill,
area, International Small Business Journal 8 (4) (1990) 49–58. London, 1987.
[30] P.-A. Julien, A. Joyal, L. Deshaies, C. Ramangalahy, A [48] D. Wobschall, Circuit Design for Electronic Instrumentation,
typology of strategic behaviour among small, International McGraw-Hill Inc, 1987.
Small Business Journal 15 (2) (1997) 33–49. [49] V. Pareto, Cours de l’économie politique, Paris, 1897.
 J. Bradford, S.J. Childe / Computers in Industry 49 (2002) 9–23 23

[50] M. Hammer, S.A. Stanton, The Reengineering Revolution— re-design theory, business processes, project implementation and
A Handbook, Harper-Collions, NY, 1995. human factors. He currently supports the University of Plymouth
[51] R.E. Stake, The Art of Case Study Research, SAGE in submitting proposals for consultancy and commercial research and
Publications, 1995. development, together with European Commission collaborative
[52] R.K. Yin, Case Study Research Design and Methods, SAGE research bids under FP5.
Publications, 1994.
[53] R. Westbrook, Action research: a new paradigm for research in S.J. Childe is leader of the Manufacturing and Business Systems
production and operations management, International Journal of (MABS) research group at the University of Plymouth. His research
Operations and Production Management 15 (2) (1994) 6–20. interests include production management and manufacturing systems,
[54] S. Chetty, The case study method for research in small- and business systems, intervention methodologies, implementation and
medium-sized firms, International Small Business Journal 15 human aspects.
(1) (1996) 73–85. He has held various research grants including ‘‘a methodology for
[55] DTI, 2000, Statistical Press Release P/2000/561. the re-engineering of business processes’’. He is the author of ‘‘An
[56] B. Wu, Manufacturing Systems Design and Analysis, 2nd Introduction to Computer-Aided Production Management’’, editor
Edition, Chapman and Hall, 1994. of the international journal ‘‘Production Planning and Control’’ and
a member of IFIP’s Working Group 5.7 on Integrated Manufacturing
J.W.G. Bradford is a Project Manager for University of Plymouth Systems. Dr. Childe was the chairman of the second SMESME
Enterprise Ltd. His research interests include manufacturing systems, Conference, held at Plymouth in 1999.