You are on page 1of 18

An assessment of the role of design in the improvement of changeover performance

Richard McIntosh, Steve Culley, Graham Gest, Tony Mileham and Geraint Owen
School of Mechanical Engineering, University of Bath, UK
Introduction Manufacturing organizations of all sizes are under continual pressure both to increase their productivity and to improve their flexibility and responsiveness to customer demands. One of the ways of achieving this is to initiate a rapid changeover improvement programme. The methods to achieve faster changeovers, or set-up reduction (SUR) as it is sometimes referred to, have been promoted actively by a number of consultants and particularly by the Japanese engineer/consultant Shigeo Shingo. Within the methods described by these experts has come the view or perception that a changeover initiative can be a kaizen or bottom up shopfloor initiative with significant benefit being achieved with little financial commitment. This paper analyses in detail the approach to a changeover programme and discusses the problems which may arise. To assist this a definition of the major terms associated with changeover is specified so that the ambiguity prevalent in the current literature is overcome. Another purpose of this paper is to analyse in detail the work of Shingo[1], which is widely accepted as being a reference book for a changeover programme, and to discuss the requirement to expand the single minute exchange of dies (SMED) concept to a more comprehensive approach. The authors have investigated the design/methodology interface in some detail and the paper concludes by promoting the role of design in a changeover programme. The need for a specific technical response to support such an emphasis will be discussed, and a case study will be presented. This work has been undertaken over a three-year research programme in association with a multinational packaging company and has involved major studies with 11 factories with the sponsoring company and eight factories from external organizations. Defining terms Changeover improvement has been discussed by many authors. Unfortunately there is no standardized nomenclature for the terms associated with the
The financial assistance of the EPSRC (Grant Number GR/H 21364) is gratefully acknowledged.

Role of design in changeover performance 5

International Journal of Operations & Production Management, Vol. 16 No. 9, 1996, pp. 5-22. MCB University Press, 0144-3577

IJOPM 16,9

changeover procedure. Those terms which are used commonly are not applied necessarily with any degree of precision. For this reason it is important to set out definitions on which an understanding of changeover is based. A representative graph of line output against time is shown in Figure 1. Some examples of common terms are: internal time, external time, single minute exchange of dies (SMED) and one touch exchange of dies (OTED). These were originally proposed by Shingo[1] and have found widespread use. Changeover, set-up, run-up and set-up reduction (SUR) are also widely used. There is, however, a large degree of ambiguity in the use of these terms and their use within this report is therefore defined next. Changeover. A changeover is the complete process of changing between the manufacture of one product to the manufacture of an alternative product to the point of meeting specified production and quality rates. The total elapsed time for a changeover, Tc , is shown on Figure 1 and is equal to the set-up period plus the run-up period. Set-up period. The set-up period is the readily defined interval when no manufacture occurs. This interval is directly analogous to internal time. It is
Manufacture of product B commences Manufacture of product B commences at a set output and quality rate Run-up period (line reaching full production)

Manufacture of product A ceases

Set-up period (line halted)

Line output Lost production during setup period Total elapsed time for a changeover Tc Internal time Figure 1. Line output during changeover External time Lost production during runup period

Time

important to differentiate between activities which take place during the set-up period and those which are required to set a machine up (i.e. adjust for production): the two need not be the same. For example, pre-setting of tooling is an adjustment activity and occurs during external time. Run-up period. The run-up period starts when production is commenced again, and continues until consistent output at full capacity occurs. It is often difficult to determine at what point full production capacity is reached as, typically, the line performance is still fluctuating at this time. It is quite likely that the production line will be halted at intervals during the run-up period as fine tuning of the adjustments takes place. Internal time. The distinction between activities which occur while the line is halted (as opposed to those which can be conducted while the line is running) has been made previously by Shingo[1]. Internal time is therefore directly analogous to the set-up period. External time. External time is time allocated to preparatory changeover activities prior to halting the production equipment. The main thrust of changeover improvement to date has been to isolate and move into external time those activities which are needlessly conducted in internal time. SMED and OTED. Shingos terms single minute exchange of dies (SMED) and one touch exchange of dies (OTED) are frequently used in industry, but are excluded here from the primary definitions. In strict definition these are time determinants and nothing more, and yet the term SMED is often encountered as a verb or noun to smed a factory; the benefits of using smed. Of SMED, Shingo[1] states :
It is a scientific approach to setup time reduction that can be applied in any factory to any machine.

Role of design in changeover performance 7

The imprecise, inconsistent use of the term SMED can be highly misleading. A further objection to the use of both SMED and OTED is that, by definition, they are strictly confined to press systems involving the exchange of dies. Organization and design. In the context of this paper organizational changes are those changes which are people or procedure-based. This might be specifying what particular operators undertake and in what sequence, they tend to rely on high levels of training and motivation. Design-based changes are those which depend on new equipment or machine changes/improvements. These are dealt with in a later section. Improved changeover performance Improved changeover performance will embody faster changeovers and/or better changeovers. Better changeovers arise by attending to the quality of the changeover, particularly in precision location and the optimum setting of all adjustments. The quality of a changeover is likely to impinge on line performance measures such as scrap, efficiency and reliability.

IJOPM 16,9

As shown in Figure 1, lost production occurs both during the run-up period and during the set-up period. It is important to attempt to minimize both the total elapsed time for changeover and production losses overall. These objectives can be compromised by attempting to minimize the set-up period in isolation. Deriving benefit from improved changeover performance As changeover times fall the opportunity normally will be taken to conduct changeovers more frequently. In general, a business will benefit by using more frequent changeovers to manufacture in smaller batch sizes. Approaching rapid changeover Before a changeover improvement initiative is started a business should understand: What level of improvement can be expected. How the business will benefit from this improvement. What structure a successful improvement initiative should have. It is not at all clear that all initiatives are embarked on with this understanding fully in place. Often changeover improvement is tackled as a flavour of the month exercise, with the degree of support and forethought that this implies. The three issues raised above are now considered in the context of current practice within manufacturing industry. This is then followed by an analysis of typical problems which a changeover improvement initiative faces, many of which may not be apparent to the business. Expected levels of improvement Changeover performance, like many other elements of the manufacturing process, is a variable. However, because changeover times are often historically perceived as being fixed, it can be difficult for production managers to grasp what scope for improvement exists. For example, the authors have attempted to persuade a manager of one factory that an eight-hour (one shift) changeover time for a particular machine could be significantly improved. This was not accepted even though an identical machine at an alternative factory had had its changeover time reduced to approximately 90 minutes. Once the hurdle that changeover times can be reduced has been overcome it is important to know what reductions to existing changeover times are possible. No mechanism currently exists which allows a manager to do this, yet alone estimate the cost and effort involved in achieving different levels of improvement. Although claims have been made for the level of improvement which may be reached, these have been of a highly global nature. Hay[2] for example asserts that:
Any setup time can be reduced by 75%.

Shingo[1] promotes similar expectations as the basis of his SMED system:


Although not every setup can literally be completed in single-digit minutes, this is the goal for the system described here, and it can be met in a surprisingly high number of cases.

Role of design in changeover performance 9

Shingo discusses improvement beyond this by refining the SMED concepts further and achieving OTED one touch exchange of dies, defining changeovers of less than one minute. Sekine and Arai[3] have set out a methodology for changeovers of three minutes or less, which they refer to as instantaneous changeovers. Claims of changeover time reductions of 70 per cent or more are supported by a number of reported case studies[1,3-6]. The benefit to a business from changeover improvement Rapid changeover is widely regarded as a fundamental component of JIT[7,8]. Within JIT, rapid changeover is cited as a stockless production technique[9,10]. Shingo[1] and other authors variously claim rapid changeover to increase productivity and reduce lead times. Schonberger[11] gives consideration to rapid changeover and the necessary degree of control which is imparted to the manufacturing system in doing so as a means to increase product quality. Oishi[12] reports a case study where specific improvement to turnover, scrap and rework is isolated as resulting directly from the standardisation and elimination of adjustment. This instance is rare, however. Most authors are unable to directly attribute quality and performance improvement in this way and instead offer generalized, unquantified claims that some degree of improvement will ensue. Rapid changeover will normally be used by a business to reduce batch sizes and thereby to enhance manufacturing responsiveness. In this way it can be seen as a tool to enhance manufacturing flexibility[6,13,14]. Flexibility is also considered by Shingo[1], who incorrectly claims:
The SMED system offers the only path to high diversity, small lot production and minimal inventory levels.

This assertion is incorrect because precisely the same results could be achieved by having a surplus of dedicated parallel equipment, which is used only as and when required. The relative merits of attaining flexibility by these two alternative routes are discussed by McIntosh et al. [15]. There are strong arguments to suggest that even if instantaneous changeover was possible, a degree of surplus capacity would still be beneficial to the organization[16]. Structuring a changeover improvement initiative Shingos SMED system is based on a simple methodology that of separating and streamlining internal and external work elements. It was originated in 1969 when Shingo was a consultant at Toyota. It was subsequently applied during the 1970s by numerous other companies in Japan and has since formed the core of consultant training packages for faster changeovers.

IJOPM 16,9

10

The methodology requires analysis and enabling techniques to be applied if it is to be successful. For example, analysis will have to occur in order to: Determine the skills portfolio of the work teams. Determine where internal and external time arise, and where effort to gain improvement should be applied. Consider the impact of, and need for, adjustment. Techniques which allow improvement to occur will include: Issues of standardization. Upgrading clamping techniques for changeover. Simplifying handling and delivery. It is in areas such as these that various consultants have had their greatest impact. In many cases these consultants have improved on, and added structure to, what Shingo originally set out. Although important, however, little more has been done other than provide refinement to a central theme. Demonstrably gaps remain in the science of rapid changeover as prominent failures have been observed, even when there is close collaboration between a business and a leading firm of consultants. Some of the difficulties which a changeover initiative potentially faces, and which can lead to its downfall, are now considered. Problems associated with changeover improvement initiatives While awareness of the significance of changeover has risen, coupled with an understanding that changeover times can be driven down, there are still problems. A study conducted with the sponsoring company has concluded the results presented in Figure 2 for one improvement initiative. This initiative was conducted in conjunction with an external training course which involved all members of the improvement team and a large number of other shopfloor

Changeover time (hours) 6 5 4 3 2 1 Figure 2. Sustainability of changeover times 0 Original changeover time (1989) Best changeover time (1991) Current changeover time (1994)

personnel. As the figure shows, there was a marked degree of improvement in changeover time, but the best levels of performance have now lapsed to approximately those of 1989. Elsewhere this general finding that times are difficult to sustain is supported, although, as noted below, reliable data have proved to be elusive. Potential difficulties faced by those conducting an improvement initiative are now considered. Measuring and reporting changeover It has been observed that few sites record their changeover performance with any degree of accuracy if indeed the performance is formally recorded at all. This means that it is difficult for many businesses to understand exactly what progress has been made, or to know what levels of performance to integrate into current production schedules. A number of difficulties are explained next. Monitoring the set-up period in isolation. With reference to Figure 1, a particular problem is that many sites believe that it is satisfactory simply to monitor the set-up period in isolation rather than to record the total elapsed time for the changeover (Tc on Figure 1), including the run-up period. This is an understandable failing, however, as it is potentially very difficult to determine when the run-up period ends and true volume production begins, where the influence of the changeover is no longer present. Integrating unrepresentative performance data. It is often the case that particular focus is placed on the set-up period rather than considering the changeover as a whole. One instance of this occurs when management teams observe and time the set-up period, and this is concluded more quickly than usual. Often the management team will confuse what has just been observed with the total elapsed time for a changeover and will take away an unrepresentative view of the factorys true changeover performance capability. Any pressure to integrate and sustain this perceived level of performance will lead to numerous difficulties and will reflect badly on the initiative. A further problem in establishing a representative changeover time for work scheduling purposes is that changeover times tend to fluctuate heavily dependent on what problems are encountered and who is present to conduct the task. Attention to adjustment. If undue pressure is placed on the set-up period, the changeover as a whole can suffer. One likely outcome is that insufficient attention is paid to the quality of adjustments, and this lack of attention can then extend the run-up period significantly. This lack of attention is also likely to impact on a number of line performance measures such as product quality and the process production rate. It is important, therefore, that the business understands the way that the set-up period and the run-up period are related. Production and quality measures. It is rare for any site to attempt to link subsequent line performance data to the changeover. Were this to happen it is quite possible that many companies would allow an increase in the time taken

Role of design in changeover performance 11

IJOPM 16,9

for a changeover as a worthwhile trade-off against improved production and quality rates when the line is back in production. Safety. Where artificial pressure is brought to bear, the changeover crews are more likely to perform in an unsafe manner. Benefit analysis There is a majority view that changeover should be tackled as a shopfloor kaizen activity. The emphasis is on exploiting the innate skills and experience of the existing labour force and, through them, achieving improvement at low cost. Noaker[17] reports a view that :
Unless you must address an immediate problem, initial steps of setup reduction should not rely on expensive tooling and equipment. Emphasise the best use of existing resources before looking to automation and large capital expenses.

12

The authors confirm that this view is prevalent in industry. Indeed, the view is so deeply rooted that for many the idea of spending virtually any money is an anathema. In one case the authors observed a request for a single item of hand tooling being turned down, even though it would have paid for itself the first time it was used. The reluctance to spend money on changeover improvement, or a reluctance to give due profile to a changeover improvement initiative, often arises because a business is simply unaware as to what benefit it will derive as changeover times fall. While many authors outline specific areas of benefit, none advocate conducting a formal analysis to quantify the worth of changeover, nor give any guidance as to how this should be done. Initiative fatigue A further advantage of conducting a formal benefit analysis is that it will allow the importance of changeover improvement to be ranked among other potential change programmes. As observed by Beer et al.[18], the credibility of any one programme should never be diluted by the sheer volume of other concurrent (but less important) programmes. Initiative targets With reference to his own adaptation of Shingos methodology, Hay[19] states that:
Regardless of whether current equipment changeover takes 24 hours or 12 minutes, this methodology can reduce setup time of any piece of equipment by 75% without the need to buy a solution.

But is this always true? As changeover times come down further and further there must be a point at which money must be spent on design time and equipment/devices if further gains are to be found. The statement applies much more correctly to a situation where there has been no previous attempt to reduce changeover times, but even then, where existing equipment is highly complex and highly automated, the Hay approach may have only a limited impact.

The issue of targets is more complex than the arbitrary setting of a particular level of improvement. Ideally any target needs to be set to be challenging but attainable, given the capability of the team who are tackling it. Also, the target should be set in accordance with the level of benefit which the business will derive from it. If business needs dictate the targets to be significantly beyond the capabilities of the team, or if the time scale to meet the targets is too strict for a kaizen approach, the business has the option of buying the services of external specialists. For these reasons targets should integrate each of: The level of reduction that is sought. The time allocated to reach this level of reduction. The resource allocated to reach this level of reduction. Directive management style vs. participative management style Most authors are in favour of implementing incremental changeover improvement through a mechanism of team-based participation. Devolvement of authority is encouraged in the belief that those who operate process equipment are those who best understand how to improve it. There is also an implicit belief that those who have been so empowered have been, or will become, aligned to the goal of changeover improvement. The universal wisdom of this approach is questioned by Stanton[20]. Stanton sets out the opposing alternatives of participative management (as advocated above) and directive management and, by analysis of 450 articles, attempts to determine what factors and conditions make either the more appropriate for a given situation. The analysis concludes that it is by no means clear that participative management the route which Stanton observes management consultants favour normally will always give the greatest benefit in a shopfloor improvement programme. Skill and intelligence Two factors which Stanton raises are those of skill and intelligence. In terms of changeover it can be assumed that operators know the fine operational detail of their equipment, but it is not necessarily right to assume that they have the experience or ability to best develop more complex design-oriented solutions. For this the services of a specialist designer may be required. It is also possible, for example, that the team members as a whole, including the team facilitator, do not have the highest organizational and documentation skills. Training A lack of skill in any area can be addressed by training. Whether this always represents a better solution than utilizing an existing expert has been questioned by Beer et al. [18], who observe that there may be negative consequences of training which are not apparent at the outset:

Role of design in changeover performance 13

IJOPM 16,9

Training programs may target competence, but rarely do they change a companys patterns of coordination. Indeed, the excitement engendered in a good corporate training program frequently leads to increased frustration when employees get back on the job only to see their new skills go unused in an organisation in which nothing else has changed. People end up seeing training as a waste of time, which undermines whatever commitment to change a program may have roused in the first place.

14

This opinion has been verified by observation by the authors, where a number of improvement teams have been extensively trained and then not been able to make use of the training that they have received. Lack of exploitation A rapid changeover capability is of no value to a business until it is used. For this to happen the planning function must actively look to use the capability as fully as possible and integrate the current level of changeover performance into all batch schedules. Even when this occurs, faster changeover remains just a manufacturing capability. It is only of benefit to a business as a whole once all other business functions, most notably marketing, additionally seek to exploit it to the fullest extent as a tool to satisfy customer needs. The role of design in changeover improvement As discussed above, changeover improvement may be contemplated based on either participative or directive management. A similar distinction may be made at the implementation level between improvement by design or by the improvement of existing manual tasks and procedures. This section discusses the relationship between design and improved organization as different ways to attain faster, higher quality changeovers. The ways that design and organization may be used in unison are also considered, as are the advantages of placing an emphasis on either approach. The prevalence of design solutions Design is implicit whenever equipment is modified for the purpose of improving changeover performance. The role of design is inferred whenever devices to facilitate changeover are attached and used on existing equipment. On this basis, assuming that most initiatives get beyond pure organizational improvement, design solutions albeit mainly at a low level are relatively commonplace. This hypothesis is not supported by any in-depth analysis of the role of design for changeover. While examples abound, there is little guidance proffered, for example, as to why specific devices should be selected. At a higher level, there are no rules to enable design for changeover (vis--vis design for assembly), nor any rules to determine what design solution is best suited to any particular situation. The relative merits of design-based or organization-based solutions are not discussed; instead design is subsumed rather as an integral element of the team-based improvement process. It is an assertion of this paper that design should be viewed in an alternative light.

Levels of design involvement It is useful to categorize where design changes can be made. In terms of the existing process equipment, design changes may be applied to : Ancillary equipment (e.g. delivery vehicles, lifting equipment). Fixtures. Tooling. Machine system (e.g. individual machines, guards, conveyors, spatial relationships). It is important to remember that product design changes may also be made. For example, Sepehri[21] notes significant improvement by this route at HarleyDavidson. In general, if it is possible to do so, any appropriate minor design change to the product is liable to offer a low-cost solution to changeover improvement. Design and SMED Shingo[1] shows an extensive number of case studies to support the SMED philosophy. A breakdown of ideas taken from this work, as shown in Figure 3, illustrates how frequently Shingo reports the application of design-based solutions. Design features to a greater or lesser extent in 64 per cent of all noted improvements. For the remaining 36 per cent of cases time reduction is by improvement to the organization of the changeover; for example, colour coding of change parts, parallel working or the ready provision of all necessary tools. Assessing Shingos work as design and methodology It is interesting to postulate that Shingos approach to changeover improvement was in fact two-pronged. The methodology of internal/external time has been well articulated and publicized. However, as shown by Figure 3, design is also extensively used. It may be argued that design is a component of the streamlining process, but it is perhaps more useful to consider design separately. Design has the effect of
Tooling 8 per cent Product 3 per cent Organizational aspects 36 per cent Ancillary equipment 7 per cent Fixture 2 per cent

Role of design in changeover performance 15

Machine system 44 per cent

Figure 3. The proportion of design-based solutions used by Shingo

IJOPM 16,9

changing the manufacturing system which previously existed, and thereby imposing change on existing changeover procedures. Superfluous changeover activity can be designed out of existence. Assessing design and organization in practice a case study The authors undertook an extensive on-site study in collaboration with an existing shopfloor team engaged on a changeover improvement initiative. The study was for a period of four man-weeks at a business engaged in the manufacture of food packaging products. The purpose of the study was not to act as consultants, but rather to understand the thinking and procedures of the team and, if necessary, to act as a catalyst to develop and progress particularly good ideas. The attachment sought to record and foster the development of the teams own ideas. There was an explicit policy not to promote or impose external solutions. In total 26 improvement ideas were identified and investigated. These were analysed at the conclusion of the attachment, and a summary of the resources required to implement the proposed changes was made, as shown in Table I. There is a considerable degree of overlap between the resource needed for each idea. The exception to this is for organizational changes (for example, ensuring that scrap was removed before the changeover started). A further breakdown was made of the cost of implementing the improvement ideas, the results of which are shown in Table II.
Resource needed Components to be made or modified Drawings required Design input requireda Parts of equipment to be purchased Organizational changes Number of occurrences 19 16 13 10 5

16

Table I. Different types of resource required for 26 improvement ideas

Note: a Design input required refers to non-trivial design where measurement, reference to drawings or involvement of professional design or process engineers is required

Cost 0-100 101-200 201-500 501-2,000 > 2,001

Number of occurrences 14 5 4 1 2

Table II. Cost breakdown for 26 improvement ideas

In practice, because of management restrictions to changeover improvement activities, it is unlikely that all of the proposed modifications will be instigated. However, by conducting a rigorous analysis of the current changeover procedures and determining how these procedures will be changed, it has been possible to estimate the likely impact that these improvements will have. This is shown in Figure 4. Two points stand out from Figure 4: (1) A number of machines were not tackled. Working on the principle that focus should be given to the bottleneck machine, the fact that no ideas were generated for machine three needs to be explained. There are in fact two reasons for this. First, because only four or five personnel are engaged on the changeover, most personnel have to tackle two or more machines consecutively. With the current distribution of work this means that machine 3 is not in fact a bottleneck machine at the existing level of performance. A more significant point is that machine 3 is compact and highly complex. The team members were not able to generate any improvement ideas for this machine over the period that the study was being conducted. (2) The changeover time for the tracking has been reduced from 120 minutes to an estimated time of one minute. The site has many tens of
Sweeps Tracking Machine 1 Machine 2 Machine 3 Machine 4 Machine 5 Machine 6 Machine 7 Machine 8 Machine 9 0 20 40 60 80 100 120 140 160 180 200

Role of design in changeover performance 17

Set-up period (minutes) Key Time before improvement Estimated time after improvement

Figure 4. Breakdown of time improvement per machine

IJOPM 16,9

18

metres of roof-mounted trackwork which need to be adjusted between different product sizes. This is a highly labour-intensive task. There has been a proposal to automate this at a cost of approximately 20,000 by the use of programmable servo motors. Effectively this would eliminate this task and, additionally, improve its quality by accurately and repeatably achieving pre-set positions. The thrust for organization-based improvement of changeover If design truly influences changeover performance to a significant extent, the question has to be raised why design solutions have not been promoted more vigorously previously? The answer perhaps lies in the fact that many of those who have published on changeover are management consultants (Hay Rath & Strong, Mather Hal Mather Inc., Oishi JIT Automation Inc., Claunch ProTech, Johansen and McGuire Manufacturing Excellence Action Coalition, Steudel and Desruelle H.J. Steudel & Associates Inc.). Others report results of changeover improvement either during or shortly after involvement with consultants (Jones Productivity Europe, Lee Rath & Strong). It is only natural that firms of management consultants who take up the gauntlet of changeover performance improvement should emphasize those aspects which most closely match their own existing specialist expertise. There is a strong bias towards aspects of organization rather than aspects of design. Changeover improvement beyond management consultants Once aware of the possibilities of using design, it becomes easier to evaluate the true extent to which design-based solutions to changeover problems are exploited elsewhere. Noaker[22] makes reference to other approaches to press setup reduction, none of which uses SMED, and shows design techniques applied to faster changeover by reference to a number of case studies. The three instances below are used to further illustrate where and how design already has a commanding influence: (1) FMSs. (2) Machine tool accessories. (3) Design consultancies. FMSs. FMSs represent a special case of a highly automated system whose stated objective is flexibility (although this may be for a relatively small range of products). It is inconceivable that an operator could fully come to terms with the complex hardware and control logic on which these systems rely, although it may be that the operator can identify where improvement may be sought. Machine tool accessories. Poling[23], writing as a machine tool accessory manufacturer, identifies the need for single-digit-changeover (SMED). However, unlike the management consultants noted above, there is discussion only of solutions by design. A review of quick-change toolholding systems by

Mason[24] supports this emphasis on design. As another example, modular fixturing systems can also be identified as having been explicitly designed for changeover. Design consultants. Globally there are a number of design consultants who will take conventional tooling and substantially redesign it for improved handling, location and clamping. These are tasks which will draw on expertise and experience beyond what a local improvement team could offer. The organization-design spectrum While it is easy to contrast organizational improvement and design-based improvement as polar extremes, the reality is that, in most cases, each solution will adopt an element of both. It is therefore reasonable to consider a solution for any specific task to be positioned on a spectrum between pure organizational improvement and pure improvement by design. Take a simple improvement idea that of replacing screws securing an adjustable guide rail with quick release devices. The idea is to reduce the total effort associated with this aspect of the changeover. The use of a device moves the idea from being one of pure organizational improvement, yet does not in itself represent a pure design solution because essentially the same manual task of releasing and moving the rail is retained; it is just made easier and quicker to accomplish. The solution would therefore lie on some point on the spectrum, as shown in Figure 5. The concept that all changeover improvement ideas lie at some point between pure organization and pure design prompts thought as to what relative advantages are derived as implementation occurs at different points between the two extremes. In Figure 6, a number of variables associated with changeover are mapped on to the spectrum.
Range for the majority of programmes? Text example?

Role of design in changeover performance 19

100 per cent organization

100 per cent design

Figure 5. The organization-design spectrum

Other factors There are other factors which a business might additionally take into account as it evaluates how changeover improvement should be tackled. These factors will include the motivation of potential members of the team and, as discussed

IJOPM 16,9

Organization (better existing practices)

Design (instigating new practices)

20

Time to achieve improvement Effort required

Usually quickly achieved Usually lowmedium Usually lowmedium Constant attention required 70 per cent

Often lengthy evolution Usually medium-high Usually medium-high Relatively easily maintained ~ 100 per cent

Cost Figure 6. Mapping attributes of changeover on to the organization-design spectrum Maintaining SUR gains Potential time savings

previously, aptitude. Other less tangible influences such as equipment not invented here may need to be considered. The business may have an agenda to change the workplace culture, for which changeover improvement may be seen as an ideal vehicle. In the same vein, team-based activities may be preferred for changeover improvement as part of a TPM initiative. The advantages of design Design confers two major advantages over incremental kaizen activity: (1) Conceivably design can be used to fully automate a changeover and render it simply an exercise of pressing a button on a console. A secondary effect of this is that design also deskills the operation and makes skilled manpower available for other tasks. (2) Second, design has the effect of changing the manufacturing system. Good design means that better things will be done, rather than simply doing existing things better. Because better things will be done, design will impact on the quality of the changeover as well as the speed of the changeover. Sustaining improvement Doing existing things better requires constant attention to prevent procedures and work practices lapsing back to their original state. By this token designbased solutions impose lasting change. As procedures are made simpler by design, so they become inherently easier to sustain.

Conclusions There is a consensus in manufacturing industry that changeover difficulties should be tackled by workplace improvement teams. This approach has merit, but it is wrong to fully dismiss the option of substantive design-based solutions to changeover problems. If a balanced improvement initiative is to be undertaken it is important that a business appreciates the strengths and weaknesses of each approach. It must be recognized that management consultants who offer changeover training packages and advise on changeover implementation methodologies will be emphasizing their own areas of expertise. These are far more likely to reside in organizational issues rather than in design. Changeover improvement has been found to be difficult to sustain in a number of instances. A variety of reasons have been identified why failures occur. Many of these are to do with weaknesses in the way that an improvement programme is structured and run. Also, there is strong evidence that management fails to quantify the true impact of sustained changeover improvement to its business. In some instances improvement teams have been observed struggling to identify improvement ideas for compact and highly complex equipment. A business must be ever-conscious of the difficulties of sustaining more rigorous working practices as a means to achieve improved changeover for there is always a natural tendency to revert back to the less efficient practices which previously existed. By using design there is usually opportunity to eliminate some changeover procedures and deskill others, and thereby gain improvement which is far easier to sustain.
References 1. Shingo, S., A Revolution in Manufacturing: The SMED System, Productivity Press, Cambridge, MA, 1985. 2. Hay, E.J., Any machine setup time can be reduced by 75%, Industrial Engineering, Vol. 19 No. 8, 1987, pp. 62-7. 3. Sekine, K. and Arai, K., Kaizen for Quick Changeover, Productivity Press, Cambridge, MA, 1992. 4. Lee, D.L., Set-Up Time Reduction: Making JIT Work, Proceedings of 2nd International Conference on JIT Manufacturing, IFS Publications, Bedford, 1987. 5. Jones, J., Achieving Quick Changeover, IMechE Total Productive Maintenance Workshops, Stratford-upon-Avon, 1994. 6. Slack, N., The Manufacturing Advantage: Achieving Competitive Manufacturing Operations, Mercury, London, 1991. 7. Schonberger, R., Japanese Manufacturing Techniques, Free Press, New York, 1982. 8. Pegler, H.C. and Kochhar, A.K., Rule based approach to just-in-time manufacturing, Computer-Integrated Manufacturing Systems, Vol. 3 No. 1, 1990, pp. 11-18. 9. Johansen, P. and McGuire, K.J., A lesson in SMED with Shigeo Shingo, Industrial Engineering, Vol. 18 No. 10, 1986, pp. 26-33. 10. Hall, R.W., Zero Inventories, Dow Jones-Irwin, Homewood, IL, 1983. 11. Schonberger, R., World Class Manufacturing Casebook, Free Press, New York, 1987.

Role of design in changeover performance 21

IJOPM 16,9

22

12. Oishi, O., Total Quick Mold Change, Proceedings of 48th Annual Technical Conference of Society of Plastics Engineers, Dallas, 1990. 13. McIntosh, R., Culley, S.J., Gest, G. and Mileham, A.R., Achieving and Optimising Flexible Manufacturing Performance on Existing Manufacturing Systems, Proceedings of IEE Factory 2000, University of York, 1994. 14. Olhager, J., Manufacturing flexibility and profitability, International Journal of Production Economics, Vol. 30 No. 1, 1993, pp. 67-78. 15. McIntosh, R., Gest, G. and Owen, G., The Relative Merits of Set-Up Reduction and Surplus Capacity to Attain Manufacturing Flexibility, University of Bath, School of Mechanical Engineering, Document ref. 065/94/RIM, 1994. 16. Mapes, J., The effect of capacity limitations on safety stock, International Journal of Operations & Production Management, Vol. 13 No. 10, 1992, pp. 26-33. 17. Noaker, P.M., Turning out faster setups, Manufacturing Engineering, Vol. 107 No. 1, 1991, pp. 43-46. 18. Beer, M., Eisenstat, R.A. and Spector, B., Why change programs dont produce change, Harvard Business Review, Vol. 68 No. 6, 1990, pp. 158-66. 19. Hay, E.J., Driving down downtime, Manufacturing Engineering, Vol. 103 No. 3, 1989, pp. 41-4. 20. Stanton, E.S., Employee participation: a critical evaluation and suggestions for management practice, SAM Advanced Management Journal, Vol. 58 No. 4, 1993, pp. 18-23. 21. Sepehri, M., Manufacturing revitalisation at Harley-Davidson Motor Co., Industrial Engineer, 1987. 22. Noaker, P., Pressed to reduce set up?, Manufacturing Engineering, Vol. 107 No. 3, 1991, pp. 45-9. 23. Poling, D., The key to single-digit-changeover, Modern Machine Shop, Vol. 60 No. 6, 1987, pp. 78-86. 24. Mason, F., A turn to setup time reduction, American Machinist, February 1992, pp. 53-7.

You might also like