Planning of research and development

Planning of research and development

Planning of Research Development and Alan Pearson Director, R & D Unit, Manchester Business School Alan Pearson is Director of the R & D Research ...

740KB Sizes 0 Downloads 72 Views

Planning of Research Development

and

Alan Pearson Director, R & D Unit, Manchester Business

School

Alan Pearson is Director of the R & D Research Unit at the Manchester Business School. The paper sets out to describe some of the most important work which is currently being undertaken in the area of R d: D and which is particularly relevant to the planning function. It is a slightly modified version of a talk given to the Manchester Se%y$; of the Long Range Planning

‘6

NE OF THE

PRIMARY

PREREQUISITES

0 for effective discharge of the research director’s duties is that he acquire and maintain a clear view of the longrange objectives of the enterprise as a whole and of the research and development activity as a part of that whole.” This quotation from Seiler’sl book, repeated by Gee2 in a recent article is a view which is being increasingly heard and also increasingly listened to. There will inevitably be some conflict between the objectives of an organization and those of different sections within it. The R & D functional area is no exception to this, and in fact is often thought to be the one where most differences are likely to be found, particularly in such fundamental matters as time scales and attitudes to risk. However, many organizations are developing methods which assist them in linking R & D more effectively with the business, and one such approach is discussed in this paper. Other sections deal with more specific, but not necessarily less important aspects of planning in R & D and reference is made throughout to current work in these areas, the majority of which is of recent origin in a field

56

where increasing attention is being paid to the planning function. HOW

MUCH

PLANNING?

Research and development is not a homogeneous activity. Its nature varies from organization to organization, its characteristics in terms of timescales, degree of uncertainty and of urgency cover a wide span. This inevitably means that the methods and procedures which are successfully adopted by one organization may not be appropriate in exactly the same form in another. Planning for research can be very different from planning for development. It may be possible to have some of the former .undertaken outside the organization but because of pressure on timescales it is unlikely that any significant amount of the latter can be subcontracted unless extremely efficient mechanisms exist for ensuring close contact between the groups involved. The decision to introduce planning into the R & D functional area can be taken overnight, in most situations we have studied, the implementation, if it is to be successful, takes a lot longer. The problem facing many organizations at the moment is that they do not appear to have the time to wait. Most observers would admit that in the past R & D departments have had somewhat favourable treatment in the way they have organized their affairs. Fears of suppressing creativity, the promotion of myths that scientists are somewhat unique beings (with norms which are different from their contemporaries who chose to be managers in the commercially oriented parts of the business) and myths that the profitability of an organization is closely linked, via technology, with expenditure on basic research, have all helped to encourage the attitude that R & D should be treated differently to other functional areas.

And perhaps it should be, but not for the reasons given above. R & D is different because of the particular types of skills involved and the type of problems encountered in an area with a relatively high degree of uncertainty. But it is not different in the sense that it should not be judged by similar standards as other parts of the business and in particular by its ability to earn an acceptable return on the resources on which it lays claim. The question as to how the return on investment should be measured is a difficult one, but failure to make an attempt will inevitably lead to a reduction in the funds available to the activity, in the long if not short term. R & D must put its own house in order if it is not to be dictated to from outside, and it must do it relatively quickly-a very difficult task in an area where so much emphasis has always been placed on the longer term. In studying the way in which organizations have adopted, or are adapting to, the changing circumstances, we have seen both successes and failures. Unfortunately, due to the complexity of the operations, it has not always been easy to identify the reasons for the particular outcome. Suffice to say that the problem of planning in R & D has been approached in a variety of ways, and the more formal methods and procedures which have been introduced have had a mixed degree of success. The total time scale over which any real effort has been specifically allocated to the problem of assessing the return from R & D is less than a decade in most organizations, and in many it is considerably less than this.3 In most cases very little evaluation of the effects of the new approaches has been attempted, and hence few clear conclusions can be drawn about their value to an organization. It would therefore appear to be valuable to draw together some of the work which is

LONG

RANGE

PLANNING

currently going on and to refer the reader to more specific examples which have been reported in the literature. In attempting to do this, it is proposed to consider planning under a number of separate, although not necessarily independent, headings and to point out the information which is currently available which can assist decision makers who are involved in the R & D functional area. AGREEING

1

Company

Brands

OBJECTIVES

The amount of planning which is needed, and perhaps more important, the amount of control which should be exercised depends a good deal upon the degree to which the aims of the R & D department match the needs of the company. Some differences are bound to occur, and indeed are probably healthy, but if they become too large the organization as a whole is likely to suffer. The way in which R & D is linked to the company differs from organization to organization. A method which has been adopted makes use of relevance trees. The aim is to break down the objectives of the two groups in a hierarchical wayeach level in the tree being made up of a series of tasks or subtasks which are directly linked to one or more tasks or subtasks in the level above and below. In a recent article Hubert’ describes an example in which four levels of breakdown are used for the company and three levels for the R & D department (see Figures 1 and 2). The author goes on to argue that “taking these two relevance trees, i.e. two main tasks (company policy and R & D policy), they can be linked together into one tree at the level of approaches (see Figure 3). The picture here is that of the policy of the company reading from left to right, and the policy of R & D reading from right to left, resulting in a common set of approaches which represent the day to day strategy of R & D.” The use of this method of analysis in a number of organizations has led to the restructuring of parts of the R & D programme and in some cases to a reappraisal of the objectives of the organization. The net result is an improved relationship between the different functional areas and an increased prospect of effectively using the outcome of research at the earliest possible time. The method itself can also indicate where particular strengths and weaknesses lie and point the direction in which changes should be made. In particular it can indicate areas of activity in which R & D could be more effective and where more effort, if successful, would lead to projects which would be rapidly taken up and exploited.

MARCH, 1972

1 Policy

Figure 1. Relevance

Tree for the Company.

Policy

R&D

I

1 C

Background

I

Figure 2. Relevance

Tree for R & D.

Policy

Company

Policy

Brands

Briefs

Approach

Figure 3. Joint Relevance

Background

Tree.

R&D

IDEA GENERATION This is probably one of the most talked about, less well documented, and least understood subjects. Many people still consider, feel might be a better word, that ideas are spontaneously generated. that ‘good’ research men are full of them and that these ideas will inevitably lead to advances in technology and hence in economic payoff to the organization. However, hindsight analysis cf major innovations frequently fails to identify inputs from the more way-out ‘basic’ area of research, more often than not advances coming from sheer hard slog, and stepwise improvement by missionoriented scientists and engineers.6,6 In fact the present movement towards more rigorous evaluation and planning systems has worried managers in R & D to such an extent that they have fought for an element of free time, often stated to be of the order of 10 per cent, for people to work on their own ideas without pressure from outside. Such work is often confined within fields in which the company is potentially interested (it would probably be so anyway because of the skills of the people it employs), but is nevertheless free to be used as an individual chooses. Unfortunately, although admittedly difEcult to measure, it appears that people are not able, or willing, to make full use of this spare time, i.e. ideas do not seem to be forthcoming at the required level, and this phenomenon is not specific to the U.K. New projects frequently arise as adaptations and modifications to existing technology. However much he may dislike this, the research man seems relatively incapable of changing it, not perhaps because he is unable to do so, but because he is now conditioned to regard anything very innovative as potentially dangerous. This is not an unreasonable fear given that innovation will usually carry with it a large measure of uncertainty-sometimes with big payoffs, more likely with considerable loss, both financially and in prestige terms. It is nottherefore surprising that studies of innovation have so often revealed the presence of ‘key’ individuals (sometimes know as ‘product champions’), people who have the ability, both resource and reputationwise, to carry a project through the difficult pat&es it will inevitably meet.7 Suggestion schemes have been introduced in some organizations in an attempt to encourage ideas, and aids such as brainstorming and technological forecasting are being used. The value of all these is open to considerable doubt. Many people maintain that little has arisen in their organization using any of these

58

methods which was not known before. However, they will admit that anything that has arisen has been given enhanced visibility and this alone has often encouraged an idea to be taken up, perhaps after lying dormant within the organization for many years. If such techniques do nothing other than this they are providing a worthwhile service. EVALUATION In the majority of situations, increased visibility at an early stage in research is useful. It encourages comment, which can improve the effectiveness of the research effort, by injecting ideas from other research workers and by indicating directions in which the research might develop in order to fit in with the objectives of the organization. The general tendency now is to encourage all researchers to complete a proposal form for any idea on which they would like to work. Some organizations make an exception in the case of proposals which require only small amounts of resources, e.g. a simple idea or a feasibility study, but others make it mandatory for all proposals-the view being taken that any idea is worth knowing about at as early a stage as possible, as suggestions and encouragement from other people may well lead to more useful results. ,Much of the emphasis in the early attempts to introduce evaluation into R & D was oriented towards attaining numbers for ranking projects, and much has been *written on this topic.* This approach was encouraged by the preoccupation at that time of most people, both academic and industrial, with the concept of discounted cash flow. The idea that producing numbers, which indicated that the DCF, return would be greater than some specified value, legitimised a project, was then much in fashion. Recent events suggest that this is no longer so. Our ability to produce meaningful estimates of costs in future time periods is very much suspect with a degree of optimism nearly always showing through. On the benefit side we are even less sure of our ability to obtain realistic figures for the likely outcomes of successful research and in certain areas the relative worth of particular in noise, outcomes, e.g. a reduction higher safety standards, are increasingly becoming the subject of value judgments. The important thing that can be said about evaluation is that, used properly, it forces attention on the work which is contemplated. It encourages discussion among a variety of people from different functional areas, it highlights potential problems and suggests areas to which attention must be paid if the maximum

possible benefits are to be obtained. In order to do this use is often made of simple checklists; occasionally these are weighted, although many companies have dropped this latter approach because it too often ends up by producing a single figure or weighted index. This can then be overemphasized with a consequent reduction in the attention that is paid to the body of the checklists in which most of the important information is contained, e.g. information about the likely availability of test facilities when required and of finance for future stages if the research is successful. The form of the evaluation procedure varies considerably between organizations, ranging from a request for a simple statement of the objectives and an estimate of expenditure to a request for a complete statement of resource requirements and cash flows. The key part of the operation is the discussion which is opened up as a consequence of the explicit statement of objectives and of research plans. These indicate to all concerned the likely direction of the work and hence enable consideration to be given to the possible ways of exploiting the results of the research if they are successful. In an increasing number of organizations this process is being taken a stage further. Research projects and programmes (with certain exceptions) have now to be agreed with product divisions or marketing groups before work is commenced and the total costs of the project are paid by the division concerned. In this sense R & D is now becoming a service function in which its objectives are to do ‘good’ research which it can demonstrate is of use to the organization which it serves. The future consequences of this movement to 3 type of ‘contract’ research are difficult to assess. The immediate benefits would appear to come from closer links with the users and hence a more rapid exploitation of new technology; the longer term consequences are that there may be no longer term. The end result will depend upon the degree of mutual trust which exists between the different functional groups and the amount of confidence which each places on the others’ ability to take a wide and 3 long view. In the past there has been much conflict in this area, possibly because of the different types and levels of training of the people who have been engaged in the different activities. This situation is now changing, and should change even more rapidly in the future, 3s an increasing number of science graduates find it impossible to obtain employment within their own functional area and therefore necessary to involve themselves with ‘non-research’ activities.

LONG

RANGE

PLANNING

An additional influence on the system is also coming from the growth of the corporate planning function itself, which can help to mediate where the interests of the different functional areas clash. We are now hearing more and more discussions about the possibility that a market research group should be responsible to the R Pr D director, indeed, one or two companies already have this, if only on a small scale. Such a development is likely to bring closer integration to the organization. The more difficult problems to be overcome are mainly concerned with new areas of technology, and areas of overlapping technology where more than one division of a company has an interest. These are precisely the areas which could become the growth spots of the next decade. Some of the current trends are likely to make it difficult for such areas to be developed and to grow until they are viable in their own right. This has been recognized and a number of companies are paying considerable attention to this at the present time.

+ : 0 ._ ‘c

.E In

May June

n

45”line. which is struck by the event progress line when the event is completed

MayJuneJuly

Aug Dates

RESOURCE

1972

of

Ott

NovDec

Jan

reviews

ALLOCATION

In briefly discussing evaluation mention was made of the difficulty of obtaining reliable estimates of the costs of projects in the R & D field. In practice it has been found that there is a degree of optimism about both costs and time and that the degree of optimism is more pronounced with the latter than the former. This has been pointed out by many authors, one of the most recent and also most relevant to the U.K. scene8 showed that cost overruns on average were of the order of 46 per cent and time overruns 204 per cent. It is also interesting that the accuracy of estimates in this case did not improve with time. This is particularly unfortunate because the benefits to be obtained from a project so often depend upon the time at which the product or process is brought to the market. The reasons for the overrun have been studied and various methods have been suggested to reduce them to an acceptable level. The well known problem of changing objectives, or, if you like, shooting at a moving target, is, of course, well documented and many organizations have produced beneficial results by concentrating more on getting their objectives right at the outset and making as few modifications to them as possible during the progress. This is not always possible. and sometimes it may not be desirable, but where it can be done it can be a major factor in reducing cost and time overruns. The introduction of planning techniques, e.g. bar charts, networks, slip charts

MARCH,

Sep

Figure 4. Example of ‘Slip-Chart’.

etc., has been successful in a number of organizations, but usually less because of the formal method itself and more because of the way in which the organization has accepted and adapted to the tool at its disposal.10 A number of organizations have developed management information systems, some incorporating formal planning procedures, others not, which are aimed at giving advance warning of likely problems.“~r2 A simple diagram often known as a slip chart can be of great value for following the progress of a project. The diagrams take a number of forms, one such is shown in Figure 4.13 Examination of the delays which occur in research projects leads one to conclude that these are often caused by conflicting claims on scarce resources, many of which could have been anticipated before the projects were authorised. The net result of ignoring this is that progress meetings are often concerned almost entirely in discussion over priorities, projects whose priorities are moved downwards having their rate of return reduced by a mere stroke of the pen when this possibility could have been foreseen at the beginning, and perhaps the project need never have been started. A more careful appraisal of the overall programme at the outset and an analysis of how a new project would fit into it is now being given particular attention.

A number of people have commented upon the fact that it may be better to undertake a number of projects and to progress them faster. Indeed, it has been demonstrated, on DCF grounds, that this can be a more profitable mode of operation, particularly under conditions of cettainty.14 However, the problem has been brought more to the forefront and more systematically analyzed by a number of groups who are considering the problem as one in portfolio analysis.lK The result of this has been the development of mathematical models, which in general have taken the form of a linear program. This requires that the resources of the laboratory be broken down into different categories, e.g. chemists, physicists, engineers, technicians etc., and that all projects are specified in terms of the resources they require in each of these categories in future time periods. The payoff from each project is usually expressed in net present value terms and the objective to maximize the expected net present value from the portfolio subject to meeting the constraints on the availability of resources. Projects can be specified in the form of a number of alternative and mutually exclusive versions, e.g. slow, medium, fast, and the model used to indicate which projects should be accepted and also at what rate they should be progressed. The model therefore assists in the planning

59

of projects as well as the selection, and this introduces a very beneficial overlap between these two stages which are often treated separately. Such a model raises many problems, many of which have been discussed elsewhere. Of particular importance are the question of what value to put on to resources, in particular manpower, in future time periods. In the short term these might be considered to be hxed and hence should not be included as a cost in the selection process because they must be paid whatever projects are selected. However, adopting this approach in the longer term can have undesirable results. In particular it will tend to perpetuate the existing system which may not be desirable. It is a generally accepted fact that one of the major determinants of research projects, and indeed of entry into most areas of activity, is the type and level of existing skills. Taken to extreme this policy is a no change one, and almost certain to lead to disastrous consequences, particularly in rapidly changing environments. Some provision therefore needs to be made in any portfolio selection model for the change in relative values of different categories of resources in future time periods. This can only be satisfactorily done in conjunction with an appraisal of the future opportunities which are likely to be available to the organization and also of the manpower policies which it is able, or willing, to operate-in particular policies on recruitment, retaining, redeployment, and redundancy. Attitudes to all of these are changing dramatically at the present time, and more than ever it is necessary for the selection and planning of research to be seen in the context of an overall corporate plan and in particular of a manpower policy for the organization. MANPOWER

ambitions is being undertaken by more and more companies on a regular basis. The net result should be an increased awareness by scientists of their own strengths and weaknesses and of the opportunities open to them both inside and outside the R 8c D department. In addition to this the mathematical programming models are being extended to take in the manpower planning aspect,‘” and it looks as though there are good possibilities for development of these into useful tools of management. If maximum use is to be made of this increased ability to model the manpower side of a research laboratory there must be more active cooperation between the different functional areas and a more coherent manpower planning policy for the organization as a whole.

.

PLANNING

People are an important resource of any organization. In R & D they are the vital component. On average they form over half the total cost of the department’s and although in absolute terms this is not a very high proportion of the total expenditure of most companies, they are involved in work which if successful will require follow-up funds which will be orders of magnitude higher. They also form a body of people, whose skills and/or potential skills, are second to none, and as such can be of great value on other parts of the organization. But are they generally thought of as such-and if SO, how are their skills used to advantage? Some organizations have paid great attention to the people in R & D, many of whom are noti holding down very

60

responsible positions in other functional areas and some are now on the boards of major organizations. But this process has tended to be a fairly informal one, and except in very isolated cases there has, until recently, been little attempt to coordinate the recruitment and manpower to the R & D department with that of the organization as a whole. This deficiency is now being rectified, if only slowly. Some companies are now setting up a system of assessment of ‘peoples’ capabilities which will enable the laboratory in the future to recognize its strengths and its weaknesses, and, equally important, enable the individual scientist to see where he fits in the organization and what possibilities are open to him. 17 Such systems are likely to increase, particularly as discussion of individual scientists’ achievements and

Distance

Figure 5. Probability

(feet)

of Communication as a Function of the Distance Separating Pairs of People.

LONG

RANGE

PLANNING

COMMUNICATIONS

SUMMARY

Finally, an area in which considerable interest has been shown is that of communications. Considerable work has been undertaken in the U.S.A.10 and also in the U.K.20 which has demonstrated that the performance of a research laboratory is influenced by the communication channels which its members make use of. It has also been shown that these communication channels are affected by such diverse factors as the structure of the organization and of research teams, by the physical layout of the laboratory, and by the roles played by certain key individuals. In the American studies it was shown that the frequency of communication between individuals within an organization fell off as their distance apart increased. This result is not entirely surprising, but the rapidity with which this fall-off occurs is something which is worrying many research managers (see Figure 5). The evidence also suggests that improvements in communications do not occur through the use of telephone facilities-contact by this medium is in fact highly correlated with face to face contact. However, more recent research is revealing that organizational structure has a considerable influence on communication patterns, and that some of the undesirable effects of separation can be overcome by manipulating other variables. This information, now well documented, has led to a number of changes being made in the design of laboratory buildings and in the organization of work groups and more attention is likely to be paid to these factors in the future.

There have been considerable changes in the management of R JL D over the last decade, the results of which are difficult to assess. An attempt has been made to indicate the major changes which are occurring and in particular those on which information is available in the literature. The main conclusions are that R & D is becoming more closely involved with overall company policy and increasing attention is being paid to the return which is obtained from the resources it consumes. The net result of this is a concentration upon integration, which if effectively carried out can reduce the time lag before use and hence before profit generation. Integration can also improve job satisfaction for many people and open up careerpossibilities which were not previously obvious. m

MARCH,

1972

AND

CONCLUSIONS

REFERENCES

(I) (2)

(3)

(4)

(5)

(6)

R. E. Weiler, improving the Effectiveness of Research and Development, McGraw-Hill, New York (IQ@). R. E. Gee, How often do research objectives match corporate goals?, Research Management (November 1970). A. W. Pearson and A. S. Topalian, Project evaluation in research and development, Management Decision (Autumn 1969). J. M. Hubert, R & D and the company’s requirements, R 6 D Management, Vol. 1, No. 1 (October 1970). Office of the Director of Defense Research and Engineering, Project Hindsight, Final Report (Washington DC 2lXOl-1989). M. Gibbons, J. R. Greer, F. R. Jevons, J. Langrish and D. S. Watkins, The value curiosity-oriented research, Nature, Lond. 225 (14 March 1970).

(7) J. Langrish, M. Gibbons, W. Evans and F. R. Jevons, Wealth from Know/edge, Macmillan (1971). (8) M. J. Cetron, J. Martin0 and L. Roepcke, The selection of R 6. D program content -survey of quantitative methods, /EEE Transactions on Engineering Management (March IQ67). (9) J. M. Allen and K. P. Norris, Project estimates and outcomes in electricity generation research, Journal of Management Studies, Vol. 7, No. 3 (October 1970). (10) D. G. Marquis, Ways of organizing projects. Innovation, Technology Communications Inc., No. 7 (lQ69). (11) R. P. Hardingham, A simple model approach to multi-project monitoring, R & D Management, Vol. 1, No. 1 (October 1970). (12) S. F. Rainbow, The use and development of a research project planning and progressing system, R & D Management, Vol. 1, No2 (February 1971). (13) Taken from a paper given by S. S. Teal to the O.R. Society R & D Study Group meeting on Q March 1971. (14) F. 6. Hellyer, A yardstick for profitable research, New Scientist (26 January lQ67). (15) D. C. Bell and A. W. Read, The application of a research project selection method R d D Management, Vol. 1, No. 1 (October 1970). (16) T. W. McRae, Financial control of research and development activity, Abacus, Vol. 4, No. 2 (December 1968). (17) P. L. Wright, J. R. Wathen and A. G. Baker, The development of an experience inventory: An aid to the effective use of technical manpower, R d D Management, Vol. 1, No. 3 (June 1971). (I 3) A. E. Gear, J. S. Gillespie, A. G. Lockett and A.W.Pearson,ManpowerModelling:Astudy in research and development. Chapter in Manpower and Management Science (Edited by D. J. Bartholomew and A. R. Smith), Enolish Universities Press (19711. networks in (19) T. IJ.- Allen, Communic&bn R & D Laboratories, R & D Management Vol. 1, No. 1 (October 1970). (20) R. D. Whitley and P. A. Frost, Communication patterns in a research laboratory, R d D Management, Vol. 1, No. 2 (February 1971).

81