Technological forecasting in corporate planning

Technological Forecasting in Corporate Plann i

T

HE

WIDESPREAD

INTRODUCTION

ng

l’:‘,“d’;;+

Drrectorate

OF

industrial technological forecasting, starting in the first half of the 1960’s in the United States, followed the breakthrough of corporate long-range planning with a phaseshift of approximately one decade. This relatively close succession of significant innovations in corporate management did not come about accidentally. Corporate long-range planning is essentially a strategic concept, dealing with alternative options, and so is technological forecasting. At the same time, corporate long-range planning adds a qualitative dimension to the predominantly quantitative framework of extrapolative planning. Technological forecasting, in attempting to bring potential future technologies into focus, aims primarily at targets which differ qualitatively from the present state. Thus, the development of technological forecasting as an integral element of corporate long-range planning was enforced by the evolution of corporate in areas marked by rapid planning technological change. 1. An Advanced Framework for Corporate Planning For the purposes of this paper, and particularly in the context of technological planning, three levels may be distinguished in an advanced scheme of corporate planning*. 1, Planning for policy-making (or normative planning) This type of planning defines This paper was prepared for a National Conference on Technological Forecasting, organised by the Ministry of Technology, The Times, and the University of Bradford in July 1968.

40

~on~u~t=nt~

for Screntrfic

Affarrs,

patterns of goals (functional objectives) which form the basis for policies. To make this a planning, not an intuitive, exercise, the feasibility and desirability of attaining specific goal patterns have to be tested by iteration in a feedback loop between the model of the present and various alternative models of the future. Ideally, the goal patterns are to be viewed as consistent anticipations (“possible futures”), i.e. intellectually constructed of discrete models futures, or substantial fractions thereof. Policies are tentatively selected on a comparative basis, but remain subject to continuous modification and refinement, or even exchange. 2. Planning for strategic decision-making (or entrepreneurial planning) At this level, alternative options for the attainment of specific goals, forming, in their totality, the “decision-agenda”, are conceived, followed through to their potential as viewed here, is essentially * Planning, “futures-creative” planning, in the version formulated by Hasan Ozbekhan of System Development Corporation as unfolding between the three levels of normative planning (the “ought”), strategic planning (the “can”), and operational planning (the “will”). The more traditional distinction between extraplanning and entrepreneurial polative (“change”) planning, which is widely applied in the business community today, focuses on the interaction between operational and strategic planning. However, it should be noted that operational planning for technofrequently objectives penetrates logical beyond the scope of extrapolative planning by incorporating qualitative change.

outcomes and assessed on a comparative basis. Technological planning here aims at objectives set in strategic terms, e.g. exploitation of resources, market dominance, market extension, competence in broad “integrated areas” viewed from a technological or market point of view, or need-oriented functions. Such strategic criteria give rise to different approaches to diversification. The essential point here is that technological planning at this level attempts to relate the strategic aims to a multitude of technological options. In particular, that type of strategic planning which orients itself towards the needs of society and the quality of life in the future (themes coming into the focus of corporate planning now), aims at functional objectives such as food supply, energy conversion and distribution, transportation, education, automation, etc. These objectives are explored on the basis of alternative technological options, and their combinations. Simultaneous strategies are sought which satisfy these multifaceted objectives in the framework of given policies. Feasibility trends, in particular trends of technological feasibility, are introduced at this level strictly as potentials which may become norms only in the context of a planned or intuitively selected policy. 3. Planning for tactical decision-making (or operatidnal planning) Planning at this level deals with one particular strategy only, delineating the sequence of action necessary for its implementation. In the framework

LONG

RANGE

PLANNING

of technological planning, such a strategy aims at a technological objective in terms of a specific product or technological system. The feasibility of effective operational planning depends here on a fairly clear view of the technological objective in question and of the principles to be employed for its realisation (good examples would be the recent phase of supersonic transport development, and the present phase of fast breeder reactor development). Sequential planning, assuming each planning step to result in a single outcome, dominates here in this narrow framework of a particular strategy. The additional assumption of linearity (the continuity of principles and forces effective in the past) is frequently made when technological objectives are to be approached on the basis of improved application of principles which had already underlain past realisations. It brings operational planning in the technological domain close to extrapolative planning which frequently dominates unimaginative corporate planning in the financial, marketing, and product development areas. This scheme for corporate non-linear planning, implying an integrated iterative process involving all three levels, provides a framework for the simulation, not the optimisation, of ways into the future which aim at qualitative as well as quantitative targets. The following implications are of particular importance for the incorporation of technological forecasting in the planning process: Planning is no more “point-to-point” planning, aiming at end-points which do not exist a priori, but assumes the character of continuous search and modification. It adopts the attitude of “inventing future”.

(Dennis Gabor) Planning aims at the rationalisation of the basis for action, not at the rationalisation of action itself. It inquires objectively into a wide spectrum of possibilities, pointing out the consequences for action to be taken in the present viewed in relation to the adoption of specific goals for the future. are decision-making Planning and separate exercises. The full planning process aims at simultaneous solutions in the context of functional objectives and of tentatively selected discrete anticipations (“possible futures”). Sequential problemsolving is restricted to those types of operational planning for which closed systems may be assumed (e.g. specific product developments).

It is obvious that the extension of corporate planning beyond its original framework of extrapolative planning

SEPTEMBER,

1968

marks the transition from a deterministic to a non-deterministic planning model. This is the decisive aspect to be watched for the proper introduction of technological forecasting into a modern corporate planning concept. II. Non-Deterministic Technological Forecasting Technological forecasting, like corporate planning, originally developed along the lines of a deterministic approach to planning. It adopted the characteristic principles of linearity and sequentiality, to which the rather indiscriminate use of time-series relating to single technical parameters still bears testimony. In this primitive form, technological forecasting provided certain inputs to extrapolative planning, but did not interact with other elements of the planning process. It was, in its extreme interpretation, regarded as an objective source of truth about the future, external to planning and human action. These nowadays obsolete notions also appear to underlie the still widespread misunderstanding which believes in separating technological forecasting from planthe supposedly ning-forecasting as objective, planning as the subjective element. In reality, technological forecasting forms an integral part of the fill corporate planning process. For modern corporate planning, and in particular for technological forward planning, the purely extrapolative forecasting and planning approach yields nothing but feasibility indications for targets as well as for the pace at which they may be attained. The possibility of extrapolating technological capabilities, e.g. by means of envelope curve extrapolation (see Figure 1) permits the use of extrapolative tools for strategic long-range planning beyond existing and emerging technologies. It provides a means to “look further than one can see”, to apply Churchill’s paradox. Used wisely, this may become of particular value. In general, however, the evolution of corporate planning, from predominantly extrapolative to strategic and then to policy planning, implies an ever decreasing applicability of linear and sequential modes of forecasting. Not only will qualitative planning and the growing importance attached to value dynamics override the inertia inherent in linear developments. Simultaneous planning in the framework of large functional systems, and eventually of entire anticipations (“possible futures”) will also render the extrapokation of progress in single technological parameters of capabilities increasingly meaningless. All will depend on the interaction between a large number of

technological and non-technological parameters, and on the qualitative goals sought. Exploratory and normative technological forecasting Technologies are developed to match needs. The deterministic approach to planning inherently assumed that specific technologies corresponded to specific needs in an unambiguous way. All that needed to be done for technological forecasting in this framework of sequential problemsolving was to identify needs and technologies and match them. The additional assumption of linearity in extrapolative forecasting tacitly assumed that linear trends of technological parameters or capabilities automatically benefited goals of the future in the same way as they had benefited goals of the past. This imposition of pseudo-goals arising from the inertia of technological development, and not selected for their merits in a future context has been rightly condemned by Galbraithl. In contrast to this, advanced corporate planning in technological areas is now based on the recognition of the following facts : A technological development, in general, has a variety of possible outcomes. A problem or a set of problems, in general, have a variety of technological solutions. Technological developments form part of a system and have to satisfy simultaneous strategies with a systemwide scope. Exploratory forecasting, which starts from today’s level of knowledge and explores future feasibilities and probabilities, has to match normative forecasting, which implies the delineation of goals of the future and their translation into missions and tasks for scientific and technological development. Only the combination of both in a “bi-polar” view leads to technological forecasting fitting into a non-deterministic framework; exploratory forecasting alone is a deterministic concept. The inherent characteristics of technological development enhance the role of normative thinking in a significant way. Almost all technological innovations can be traced back to a clear need formulation. The enumeration of opportunities alone normally does not trigger action. Modern corporate management takes this into account by devoting increasing attention to the problem of translating corporate objectives all the way down the hierarchic ladder to stimulate the creativity of people at all levels. More recently, emphasis is being placed on linking corporate objectives to social goals, thus starting the normative

41

Environment Informotlon

level

Elements of total corporate environment (functions) Technol. missions Technal. options Implications for carp. objectives Direct market impl. Repercussions in total

market system “I

a,

ln m .2!

m

In

-5 2 0.0 OEZ cn=op .c Z.&l2

cl

+

a I__

I

I

.c

Feedback from real world z Production and oper. - E requirements; detailed Lfeconomic assessment Systems design and oper: char. _ Systems elements and developm. requ. (schedule, cost, time) %---Systems performonce Attainable technol. .L capabilities e 0 Basic potentials -_ and limitations $J

Science

& Technology

’ I

L-L

w----7r

-----5e I

t

Decision

for specific

technol.

dev.

FIGURE 1. TECHNOLOGICAL FORECASTING STAGES IN RELATION TO THE RDT&E PHASES OF A TYPICAL TECHNOLOGICAL INNOVATION. The formal scope of each technological forecasting stage is the forecasting of the outcome of the next RDT&E phase, the informal scope

(dotted arrows) extends further. Six forecasting stages and one evaluation stage may be distinguished. The thickness of the arrows indicates schematically the width of forecasting, narrowing down in the course of vertical technology transfer from the simul-

taneous consideration of a large number of alternatives to the decision for one specific development line, and subsequently widening again in considering alternatives of horizontal technology transfer (applications and service).

chain at the highest level. “The corporate leader who does not try to conduct his company so as to instill pride in his people is doomed these days” (Joseph Wilson). A complete technological forecasting exercise, on this basis, always constitutes an iterative process between exploratory and normative forecasting. This process, it should be noted, does not match rigid sets of technological opportunities with equally rigid sets of objectives, ordering them in such a way that “the right key fits the right hole”. It constitutes a feedback cycle in which both opportunities and objectives are treated as adaptive inputs. The search for new and more suitable opportunities goes hand in hand with the adaption of goals to technological feasibilities and higher probabilities. Technological forecasting in the framework of integrative planning includes feedback processes of even higher complexity.

the seeds for technological innovations over a period of not more than some 15 years. Moreover, it was considered superfluous to go beyond this time-distance, since good forecasting would guarantee a “surprise-free” period sufficiently long not to worry about the opportunities presented by new scientific discoveries. This limitation in outlook still applies to technological forecasting for the purposes of operational planning. Only at this lowest of the three planning levels, is it the task of technological forecasting to envisage concrete technologies and technological systems. Normally, one will keep the options open until the decision about a certain line of technological development has to be taken. The time-span of 5 to 10 years’ development generally encountered in complex development is frequently also adopted as the time-frame for technological forecasting in an operational planning context. The structuring of thinking in accordance with long-range and ultimate goals, which constitutes the essence of normative planning, also widens the time horizon of technoIogica1 forecasting. The goals for the future, which are usually very different from the goals applying at present, provide an outlook that may require a tech-

nological decisions-agenda quite distinct from one appearing satisfactory for a shorter time-frame. This is particularly easy to demonstrate by the example of the supply of vital resources.

The forecasting

time-frame

In the framework of linear planning, technological forecasting focused primarily on the recognition of future technological realisations in concrete, descriptive terms. The point might be made that the scientific princi$es known today, with relatively few exceptions, constituted

42

If, for the year 2,000 the world population forecast reaches 6 billion people, they could still be fed on the basis of agricultural food production. However, it appears that this level will hardly represent the maximum (we cannot influence the linearity of this development effectively enough), which may perhaps reach 12 to 15 billion in the middle of the next century. If the goal is to feed this number, even for a temporary period of a few decades only (before a decrease in world population can be effected), the decision-agenda of agricultural food production technology is not sufficient. The non-agricultural food production technologies which are required after the year 2,000 have to be developed and introduced simultaneously before that date, with the pushing of agricultural productivity by all available means. The longer outlook yields a dramatically different structure of the World Food Problem and the technological solutions to be prepared. It is not surprising to find that those branches of industry which, so far, have pioneered the application of trategic and normative planning in their corporate thinking, practice technological forecasting

LONG

RANGE

PLANNING

in ~1 time-frame which extends up to 50 years, and sometimes beyond.* These are branches which are conscious of the social implications of the technologies they develop, and therefore try to anticipate social goals of the future. Technical companies of an innovating type, such as the aerospace, electronics, and chemical industries, usually apply technological forecasting to a time-distance of 10 to 20 years, whereas consumer-type industry is satisfied with 5 to 10 years. These indications refer much more to the emphasis placed on strategic and normative planning than on a difference in “visibility” of the technological developments and their outcomes. In general, operational technological plans, integrated in the formal corporate long-range plan, hardly ever go beyond a period of 10 years. When then is the limit for normative planning? In many contexts, it appears to be set by the dynamics of the global system of human society. The disequilibria and dynamic instabilities developing in this system at present, namely the “gaps” between the developed and the less developed countries, the unfinished trend towards world unification, and the intertia of the population explosion, can possibly be brought under control not before 100 or 150 years have elapsed. If, in normative a new situation of global planning, stability is envisaged it is a time-frame of approximately this length which is implicitly applied. On the other hand, if the aim is just vaguely formulated in terms of an improvement over the present situation, the goal of this improvement is not clear. It does not appear, therefore, as utopian if a complete corporate planning framework includes technological forecasting which aims at goals 100 or 150 years in the future-although it will certainly not be able to specify by what means these goals will be attained. But it will attempt to verify whether the technological decisionagenda resulting from the application of existing or imminent technology is sufficient to reach those distant goals. If this is forecasting will not so, technological stimulate the search for new technologies and for new types of fundamental research.

III. The Scope of Technological Forecasting in Corporate Planning As stated above, the task of technological forecasting is not to predict technological realisations of the future. The inputs to strategic and normative planning do not generally include forecasts in terms of concrete technological systems and their design elements. But it must also be noted that the scope of technological forecasting in the various stages of corporate planning * Examples munications

SEPTEMBER,

are the petroleum industries.

1968

and

the

com-

differs widely. This gives rise to quite different approaches, depending on the planning stage, which are taken simultaneously in a complete framework of corporate planning. It is useful to discuss the scope of technological forecasting as viewed against the background of two different “crosssections” of the corporate planning process : following the generation of a typical technological innovation through its various RDT&E stages7 from prediscovery to application and service engineering; and discussing foretechnological casting at the three levels of an advanced corporate planning framework, as outlined in Section I: normative, strategic, and tactical (operational) planning. The unfolding scope qf’forecasting visibility Figure 1 attempts to follow a technological innovation through its various stages identified by the RDT&E phases:: pre-discovery, discovery, creation, substantiation, development, advanced engineering, application and service engineering. This succession of RDT&E phases provides a much clearer picture of technological development than the widespread use of the ill-defined notions of fundamental research, applied research, and development-which, in addition, do not form an unambiguous succession of the actual work phases, but overlap and supplement each other particularly in the decisive stages of creation and substantiation. It can be readily seen in Figure 1 that exploratory forecasting, starting from the known science and technology base, and normative forecasting, starting from the environment, cannot be matched directly and in explicit terms until relatively late in succession of RDT&E phases. Although some tentative forecasts, made between the creation and the substantiation phases, may attempt to anticipate possible systems designs in descriptive terms and possible market impacts, these forecasts may still undergo considerable modification on the basis of the results obtained in the substantiation phase. It is only before the beginning of full-scale development work that such a forecast enters formal planning and leads to a decision for one particular development line. Only then can needs and RDT&E is the commonly used abbreviation for “research, development, testing, and engineering” comprising all scientific-technical aspects of the technological innovation process (whereas the traditional notion of R & D i.e. research and development only, is more restricted). These stages were adapted from a scheme proposed by the Stanford Research Institute.

opportunities be matched in a complete forecasting cycle. Figure 1 tries to express this by using forecasting arrows of different thickness, indicating that, as long as exploratory and normative forecasting do not clearly “see” each other, a number of alternative possibilities had to be taken into account simultaneously. After “vertical” development for scientific principles up to technical realisations turns to “horizontal” development adapting the new product or system to various applications and services, forecasting again various simulincludes alternatives taneously. The forecasting arrows are depicted as pointing in both directions, thereby indicating that all forecasting is a continuous feedback process. Norms derived from the environment, and technological feasibilities derived from the known science and technology base, are continuously modified and updated against these everchanging backgrounds. Figure 1 also indicates the structure of one of the most important tasks of technological forecasting namely to improve the coupling between successive RDT&E phases. By anticipating the possible results of each work phase, technological forecasting stimulates and guides work to be undertaken in the successive phase. Not fewer than six different coupling functions can be read from Figure 1, of particular importance is the coupling between the discovery, creation, substantiation, and development phases-in vaguer terms the famous coupling between science and technology, which is blamed today for Europe’s lagging behind in some technological development areas in spite of the sound scientific basis available. * One would think that, ideally, there ought to be a technological forecasting exercise at each of the six “coupling joints”. This is indeed the goal towards which the detailed structure of technological forecasting is developing today in advanced thinking corporations. The first two forecasting stages may be handled in an informal way where no formal policy planning has yet been established. Of decisive importance for successful product development are the forecasting stages between creation and substantiation, and between substantiation and development-in other words during the transition period from strategic to operational planning. It is important to note that, as exploratory and normative forecasting approach each other, the forecasting scope does not shift from level to level, but expands to include more and more information levels. If the first forecasting stages may be compared to visualising vague mirror images of only the most basic formulations of

43

needs and potentialities, the succeeding stages build more and more firmly on this basis. Technological forecasting, if practised correctly in the framework of advanced corporate planning, is emphasising “/zigA coupling” within the iterative forecasting cycle itself, meaning that gradually each information level is matched with all other levels. For example, forecast systems design elements are matched not only against direct market implications (economic demand, price, etc.), which would constitute “low coupling” but also against the norms derived on the level of technological missions, at the level of the elements of total corporate environment, as well as against attainable technological capabilities and basic potentials and limitations, and so forth. Technological forecasting at policy-planning level The scope of technological forecasting is different for each of the three planning levels corresponding to the advanced corporate planning framework. Instead of attempting to formulate a generally applicable and necessarily vague definition, it is more useful to distinguish between technological forecasting as applied to planning at these three levels. The scope of technological forecasting at policy-planning level is the clarification of scientific-technological elements determining the future boundary conditions for corporate development. In more concrete terms, technological forecasting at this level focused on basic scientific-technological potentialities and limitations as well as on their conceivable ultimate outcomes in a large systems context. Non-linear planning at policy level deals, at one and the same time with problems for which the inherent freedom of choice is very narrow and others for which it is very wide. The tools which serve the purposes of technological forecasting at this level range from quasi-linear extrapolation to the almost uninhibited use of intuitive imagination. A small degree of freedom is encountered when dealing with large systems of the corporate environment, which may range from market systems and industrial sectors to political and societal systems of national and even global dimensions. Large dynamic systems, especially if the aspects of social dynamics dominate, imply a relatively high inertia of macroprocesses. This becomes visible, for example, in demographic, political, and macroeconomic trends. Of inertia significance is the particular inherent in institutions of all kinds. Industry is itself such an institution and corporate development is restricted by the feasible degree of internal and external

44

flexibility, including vested interests, qualitative and quantitative aspects of manpower, production capacity, marketing, and so forth. A large degree of freedom is encountered in the adaptation of corporate policiesincluding their scientific-technological elements-to the environment. The multitude of technological solutions contributing to a specific goal, and the multitude of possible outcomes of each technological development in a large systems context, guarantee the possibility of flexible choice, including, in particular, non-linear changes in technological development. Industry is emerging today as the “planner for society” and taking over many of the traditional government functions in this area. It is thus assuming an active and decisive role in “shaping the future”. As a matter of fact, it is the adoption of this basic attitude which, more than anything else, stimulates the introduction and application of the advanced corporate planning framework, and of elaborate technological forecasting, in American industry* today. The idea of studying, in the framework of independent national and international institutions, functional systems as parts of complete anticipations (intellectually constructed models of alternative possible futures) is therefore of the greatest interest for industry. In the United States, the National Industrial Conference Board has recognised this and joined the efforts to set up an “Institute of the Future”. Such an institute would be the logical extension and of the system of “early detection reporting of environmental trends”, started by NICB in 1966. The pattern of industrial activities is becoming increasingly function oriented, i.e. oriented towards specific broad-need areas defined in functional terms, such as transportation, communication, education, power generation and diffusion, etc. The large function-oriented corporations, leading the active “shaping of the future” as prime contractors for the complex technological systems developing around this task,t will not only derive their corporate policies from such common anticipations, but will also play an active role in con* European industry, in general, is lagging behind in the adoption of such advanced thinking. Notable exceptions are, for example, the big oil companies. t It is estimated that, within ten years, industrial production in the Western part of the world will increase by 65 per cent, but that the number of industrial companies will decrease by 15 per cent. Three-quarters of the industrial potential of the Western world will then be in the hands of 300 corporations. Fewer than 700 corporations will handle more than half the world’s total business.

tributing to the conception and comparative evaluation of alternative anticipations. Forecasting and planning at policy level face particularly difficult problems since, to a considerable extent, they deal with value dynamics. There are no infallible ways to judge the future merits of certain goals, or of outcomes of technological development. Nevertheless improved knowledge of the mutual consequence of the adoption of specific goals and of decisions for action in the present can be achieved through the simulation of planning alternatives in a feed-back cycle. To a certain extent, this will already make it possible to distinguish between “good” and “bad” alternatives. A special problem for technological forecasting for the purposes of industrial policy-planning concerns future changes of inter-industry structures. Apart from the acquisition policies of financial holding companies and diversification on the basis of special skills, there are at least three effects which are directly related to technological development : Diversification through “organic” forward and backward integration to form integral functional chains (exemplified by the Esso/ Nestle enterprise for the development of Single Cell Protein on petroleum basis-the functional chain here covers the whole range from petroleum prospection to finished food stuffs). The “invasion” of one industrial sector by another on the basis of technological innovation (exemplified by the invasion of the textile by the chemical sector, the chemical and pharmaceutical sectors by the petroleum companies, etc.). The “blurring” of sectoral boundaries due to technological developments pushed to their extremes (exemplified by the vanishing distinction between electronic components and systems in the era of integrated circuits, and between airframe and propulsion in the era of supersonic and future hypersonic aircraft).

Figure 1 graphically shows the central problem of technological forecasting in the early RDT&E phases, when policy planning is still isolated, without the benefits of elements resulting from strategic and tactical planning. It establishes, over a wide gap, a dialogue between functional targets on the environment side, and basic potentialities and limitations on the science and technology side. Thereby, technological forecasting stimulates in a significant way fundamental research, at the same time providing broad guidelines for it. A number of advanced-thinking industrial companies already attempt to align their fundamental research with the questions raised by technological forecasting.

LONG

RANGE

PLANNING

For normative forecasting at policy planning level a special formal approach is use-the Environmental into coming Information System (EIS) comprising data input and storage techniques as well as the processing of data to link them to corporate objectives. f Technological forecasting at strategic planning level The core of technological forecasting is situated at this level. It is here that the nature and impact of future technological options become discernible in their profiles, that the decisions leading to the adoption of specific technological development lines are prepared, and that imaginative forecasting is able to influence the course of social and economic macroprocesses most effectively. The scope of technological forecasting at strategic planning level is the recognition and comparative evaluation of alternative technological options, or in other words, the preparation of the technological decisionagenda. The focus is not on descriptive forecasts in terms of systems design, e.g. how a new type of machine may look or work, but on assessing feasible systems performance in the light of attainable technological capabilities, and on relating technological options to functional missions. Nevertheless, as indicated in Figure 3, forecasting attempts to penetrate, to some extent, into areas which are to be clarified by operational planning. The most important task for strategic planning is the elucidation of causeleflect relationships over long periods of time. The human mind, intuitively, relates causes and effects which succeed each other within relatively short-time spans. This implies the danger that incorrect or artificial relationships are established which tend to blur the issues of corporate long-range planning and replace them by a mosaic of short-range relationships. The Industrial DynamicsS approach permits this problem to be structured in the correct way and implicitly underlies most of the simulation models, in particular computer models. Scenario-writing, in an industrial context, may be regarded as a special technique derived from Industrial Dynamics (or the study of dynamic systems behaviour in general). Both the input and the output of such simulation studies include important elements of technological fore: The development

of Environmental Information Systems is pioneered by General Electric’s TEMPO Center for Advanced Studies. Such systems also play a decisive role in environmental studies with a national scope, such as those produced by Lockheed (the classified “Mirages” series) and McDonnell-Douglas.

S: Jay W. Forrester, Industrial Dynamics, Cambridge, Mass., 1961; and a large number of publications following this original treatise.

SEPTEMBER,

1968

casting, in the form of technological capabilities and performance characteristics in the input, and in the form of outcomes in the context of large systems in the output. The large systems aspect and the search for simultaneous strategies satisfying the boundary conditions and a multitude of systems requirements, constitute another basic task for strategic planning and forecasting. Strategies can, of course, aim at different systems, such as those formed by resources**, special skillstt, markets, financial compounds, and functional areas. It is most important to note that planning with a qualitative scope in a societal framework-the scope emerging now for advanced corporate planning in industryhas to be based on functions. Forecasting at strategic level evaluates technological missions and alternative options in the framework of broad functions, i.e. societal need categories. A third important aspect of forecasting at strategic level is the deliberate and unbiased introduction of non-linear planning alternatives. All forecasting techniques applicable at strategic level incorporate this element of non-linearity, and some of them are designed with this particular aim in mind, for example, envelope curves, morphological research, and contextual mapping. All of them, however, have some inherent limitations which give rise to “high-order” linearities, which also have to be recognised and carefully avoided in sophisticated strategic planning. It is significant that the “low coupling” approaches in the forecasting cycle, e.g. techniques relating development requirements to market impact are increasingly making use of strategic aspects. If resource allocation and priority ranking procedures on the basis of simple operations research formulae aim at maximising returns on investment (or “present net values” in a discounted cash flow framework), their results are over-ruled by priorities derived from simple decision theory formulae which place emphasis on the strategic impact of certain development. This means that strategic thinking increasingly dominates also those RDT&E phases at which operational planning is starting. The best example of a resource-oriented strategy may be found today in the petroleum industry which exploited the basic resource petroleum to the fullest extent by adding to the latter’s scope every new utilization petroleum found through technological innovation; lubrication, illumination, energy for transportation, combustion and power generation (fuel oils), petrochemistry, pharmac&ticals, food production. To a certain extent, the chemical industry such a discipline-or skill oriented sector.

is

Technological forecasting at tactical planning level If normative and strategic planning are still somewhat aloof and keep sufficient distance from reality to maintain a capability of flexibility choosing between alternative options-attempting to “ride” the future reality, not be “ridden” by it-tactical or operational planning is the level at which these optional concepts have to harden into reality in a process that involves the adaptation of both the plan and the factors of reality to each other. In technological forward planning, it is the results obtained at this level which enable the planner to make the decision for and objectives specific technological developments. The scope of technological forecasting at tactical planning level is the probabilistic assessment of future technology transfer. Both vertical and horizontal technology transfer are considered here, implying that the forecast penetrates beyond the design and performance characteristics of a specific technological system to its utilisation in the context of different applications and services, repercussions in the market system and implications for developments of social, technological or other nature. Tactical planning may also include a contigency approach, e.g. the various network planning techniques, which may be considerably enriched by technological forecasts related to critical “obstacles” and decision points. A particular aspect of technological forecasting at this level concerns its contributions of extrapolative planning, or linear planning in general. Only in this narrow framework can deterministic planning be put to use. Nevertheless, the skilful application of strong normative forecasting, in combinations with the purposeful assessment of absolute or practical limits, already implies the possibility of gaining decisive advantages over the competitors. One may distinguish between two basic corporate attitudes : “Leading the main-stream” by aiming at a steeper gradient of performance over time. This is the attitude of aggressive and alert techmcal corporations, aiming at high-risk profit as they dominate the market today in the United States. “Drifting with the main-stream” by aiming at closely following the forecast curve, especially in industrial areas characterised by a large number of equally capable competitors::. This attitude is taken by $j: Since trend extrapolations usually refer to pace-setting feasibilities under the assumption of certain inherent dynamics, being “on line” in mono- or oligopolistic systelns of industrial innovation may well imply leadership in any case.

45

Anticipations Normative technologica forecasting

I

Functions

Product lines are pushed to the extreme; Demand is created artificially for the sake of specific scientific and technological developments, or to maintain certain skills and capabilities. In general, uncontrolled development occurs where a deterministic attitude is taken, in particular the attitude of linear and sequential planning. In order to overcome the restrictions of linearity and sequentiality, planning has to be integrated vertically (policy strategy tactics) as well as hori~ontaIl,v (system wide).

Technologies (products, processes, el -C

Simulation-not

Resources (material and non-material 1

FIGURE 2. TARGETS FOR PLANNING AND FORECASTING. Each technology can, in general, be realised by different combinations of resources, and has a variety of outcomes; each functional objective has a variety of technological solutions. “Commonalities” are indicated schematically. The three layers of the figure correspond to and policy planning. tactical, strategic, Technological forecasting implies simulation runs through every possible combination of unbroken vertical lines.

Exploratory technological forecasting

corporations aiming at relatively low-risk profit, and is partly at the roots of present European complaints about a “technological gap” between the United States and Europe. A number of techniques for economic analysis, the more sophisticated ones applying discounted cash flow calculations and optimisation by simple operations research, provides the “bonding” between technological and market forecasting. IV. Integrative Planning and Technological Forecasting Integrative planning is planning cutting through two or more of the following dimensions : social goals and values, institutions and social structures, technology and society, population and material resources, physical environment. Corporate planning within the framework outlined above is inherently integrative planning and technological forecasting contributing to it has to deal explicitly with all the above dimensions. Technology has become the most important factor in social change and substitution. The uncontrolled development of technology leads to metastable solutions and disequilibria, instabilities in the dynamic system of society, waste, and uncontrolled or cancerous change. Within the corporate framework, the uncontrolled development of technology may mean, in particular: Technological feasibilities, when recognised, are turned into reality; is understood as “ought”; “can” Market demand is followed “blindly”;

46

analytical

solution

Forrester has shown that the simplest systems of real interest to thp manager begin at fifth order and that most of the actual systems encountered in corporate management may be situated between tenth and hundredth order. At the same time, corporate systems have to admit positive feedback, a high degree of nonlinearity, and multiplicity of loops. The same development may thus be expected to happen as in engineering when it reached a degree of complexity that outgrew conventional mathematics, namely the application of analogue and digital computers to replace analytical solutions of complex problems by simulation. “Rather than naively oversimplifying the real world to fit within the limited power of mathematics, the engineer turned to the simulation of large number of special cases. From these special cases he can generalise, within limited regions, on the character of the system under study”. The aim of corporate planning is therefore to determine “good” strategies by selecting them from as broad a basis of alternatives as possible. Technological forecasting, especially at the strategic planning level, is called upon to enrich this basis for selection. At the same time, the systematic evaluation of a large number of alternatives-implying the iteration between environment and the science and technology base by means of multiple simulation of the consequences inherent in the selection of any one technological option-underlines the importance of modern icformation technology for the tasks of technological forecasting. One may expect that technological forecasting will be incorporated into future complex management it7formation systems. It should also be noted that, in a nondeterministic framework, technological forecasting forms an integral element of planning, not an “external” contribution to it. The feedback cycle between the exploratory and the normative approach in nondeterministic forecasting actually implies a

LONG

RANGE

PLANNING

feedback relationship between forecasting and the design of concrete plans. To the extent that a technological forecast influences the design of the plan, this plan also influences technological forecasts. This, of course, is irreconcilable with the older notion of technological forecasting as some source of absolute truth. It is of the utmost importance to understand that both technological forecasts and plans are adaptive to each other, and to make use of this flexibility in corporate planning. Runningthrough the simulation of a number of pre-conceived alternatives will normally lead to second and further simulation runs “testing” modified and adapted alternatives in technological forecasting and plan design.

The vertical ,forecasting

integration

of

technological

Figure 2 depicts in a schematic way the scope of technological planning over the three levels of corporate planning*: It may be readily seen that one specific technological realisation (a produce or technological system) corresponds to a multitude of outcomes as well as, in general, to a multitude of resources to be employed for its realisation. The systems aspects, however, can be represented only in an unsatisfactory way by looking at as a specific “commonalities”, such, technological realisation benefiting more than one functional target, etc. Before deciding on the development of a specific technological realisation, one should be able to simulate any combination of links betu-een the resources and the anticipation level-not just for one technological realisation (comparing the alternatives relating to resources and to the different outcomes a specific technological realisation may have in the framework of different functional systems), but also for all other technological realisations which are related to the same functions. Only in this way can the technological decision agenda become really informative as to the alternative decisions possible. It is clear that such “complete” simulation runs can be made only for technological developments which have reached advanced RDT&E stages. One will therefore have to add “incompletely filled pages” to the decision-agenda, which may nevertheless influence the decisions considerably. Another difficulty is possible because functions may not be clearly definable * A fourth level, for future management

systems, may develop in the form of “planning for attitudes”; the solutions would then no longer be purely technological, but would include “social engineering”.

SEPTEMBER,

1968

FIGURE 3. INTEGRATIVE TECHNOLOGICAL PLANNING IN A CORPORATE FRAMEWORK. The three main feedback loops correspond to policy, strategic, and tactical planning; Planning at every the dotted lines indicate the possibility and desirability of “high coupling”. level has a system-wide scope; anticipations, functions, and technological systems, respectively. The characteristic inputs of technological forecasting are indicated for every level; on the environment side, they are enriched by inputs from non-technological forecasting.

Science and technology

Environment

Bos~c

potentloIs

ond limltotions

Antlclpotion

Function01 technolog missions

-

r

ii;-Function01 technolog.

_

feasibilities

Function01 nological cision

techde-

- ogendo

I/j

Corporate strategies

‘II ‘II I[LY II

Design ond developm. -

Ii

chorocteristics

11

I--

Technological objectives

_

0

Ii1

Lk_---Y

_

Non-moterlal resources (science and technology, skills, etc.)

Resources

--.I

*I&J

Moteriol (manpower, moteriols,

”

resources f inonce, etc.)

Market implications

over longer periods of time. Developments maturing in a farther time-distance might have to tie in with functions differing somewhat from those aimed at by developments which can be expected to impinge on these functions earlier. Especially in areas of very complex interaction between rapidly moving technologies, “mobile” functions may complicate forecasting and planning. The future function of the “home”, and of urban development, in general, will depend to a large extent on the development of such functions as automation (people not going to work, but interacting from their homes by means of a computer console, etc.), communications (chi!dren and students learning at home via satellite communication, teleshopping etc.), and transportation.

Figure 3 attempts complete

FIGURE 4. THE ORGANISATION OF TECHNOLOGICAL FORECASTING WITHIN THE CORPORATE STRUCTURE. Each forecasting stage involves the interaction between different levels. The synthesis is always made at the highest level involved. (This schematic representation does not show the full feedback taking place at each forecasting stage.)

Forecasting

to give a scheme of the

technological

stages

Research

groups

II X

lab’s

0

0

Divisional level Management (staff groups and div. lab’s) Project

cycle,

Evaluation

Board level Board officers Corporate level Horizontal staff

forecasting

applied here simultaneously to all domains of corporate planning with technological implications. It shows the feedback cycles between environment and the science and technology base at each corporate planning level, but also the triple feedback loop between technological forecasting on the three corporate planning levels. Of particular importance, although less well-developed as yet in terms of techniques and formal approaches. is the

engineering

groups

RDTand E phases

Forecosting Objectives

Environment

Science

ond

tee hnology X

Synthesis

0

Inputs

Corporate environment

Technol. missions

Technol. options

Basic pot.

Attain. technol. copob.

Systems perform.

ond limit

Impl.for Corp. obj. Systems el.ond dev requ.

Morket imp1 icot. Systems design

Reperc. in morket

Feedback

Prod. ond oper. requ.

Feedbock

and inputs

LONG RANGE PLANNING

“high coupling” between non-adjacent levels of objectives, e.g. between the functional decision-agenda and the basic scientific and technological resources. A few companies, however, already pay particular attention to such “high coupling”. A merely tactical approach to technological forecasting and planning would correspond to simple (linear) product line development. The combination of the tactical with the strategic approach lends to a flexible approach in the framework of a specific function which is assumed rigid. But only the addition of the policy level permits the rational setting of corporate objectives which tie in with national and societal objectives in a flexible framework of “mobile” functions. If we want to control the outcome of a technology, we have to work back from an anticipated outcome. This is also, if we do an ideally good job, the only way to plan for consistent and reasonably stable futures on the level of the corporation as well as of society. This is what we understand under “shaping the future”. Forecasting and planning, in the context of “shaping the future”, provides the insight into the necessities and consequences of alternative decisions. They do not anticipate or favour a bias in the decisions. The horizontal forecasting

integration

of technological

The necessity of system-wide planning is recognised to an increasing extent especially in areas where technology and society are in close relationship. Single track technological solutions, arrived at by sequential planning, tend to become metastable in large systems and deteriorate. The nuisance effects of certain technologies exemplified by water and air pollution, and unbearable noise levels-are examples affecting everybody today. System-wide planning is of particular importance where different technological solutions are not only possible but simultaneously necessary, for example in urban transportation systems. The possibilities of non-linear planning are here restricted by the systems boundaries and by the inertia inherent in the various dynamic elements of the system. We cannot change instantly either the direction or the momentum of a dynamic systemonly the curvature of its movement. It is to be expected that a new economic attitude will develop towards such technological solutions as satisfy systems requirements rather than merely sequential requirements. The development of a cost/ efictiveness approach for the application in corporate development seems therefore of particular value.

SEPTEMBER,

1968

The dimensions of systems in which the outcomes of future technologies sought to be studied are generally at least four-fold : social, economic, political, and technological. Again, the complexity of the systems-like the corporate systems mentioned above-is far too high to expect the feasibility of analytical solutions. Simulation may be attempted either by “scenario-writing” (or its simpler version “iteration through synopsis”) or by computer simulation. The latter approach, still in its infancy, appears to progress at a fast pace and will probably lead to the incorporation of such models in future management information systems. V. The Organisation of Technological Forecasting in Industry In an inspiring article, Forrester considers the future form of management as “enterprise engineering” and points out that of the possible contributions which technical engineering could make to it, the following stand out : (1) The concept of designing a system. (2) The principles of feedback control. between policy (3) The clear distinction making and decision making. com(4) The low cost of electronic munication and logic. of simulation for (5) The substitution analytical solutions. Technological forecasting may be considered as a tool of enterprise engineering involving, ideally, all these principles. The feedback character of technological forecasting, the need to employ simulation, and the implications of computer use, have all been brought out clearly in the preceding Sections. In this last Section the remaining points (1) and (3) will be considered. The basic distinction between policy and decisions can be drawn most clearly in Forrester’s own words: “Policies are those rules by which the input information streams are converted into decisions to control activity. The decisions are the continuously generated results of applying the policy to the input data flows. The difference between policy and decisions can be illustrated in engineering terms: policy is the transfer function (or the computer program), whereas the decision stream corresponds to the moment-bymoment values of the resulting signal that has been processed through the device described in the transfer function:” In analogy to technical engineering, Forrester envisages two kinds of engineers also for “enterprise engineering”: the stationary engineer, who is an active component of the operating system; and the systems engineer, working at the next higher level of abstraction in the systems

hierarchy than that of the stationary engineer and dealing explicitly with what the stationary engineer treats intuitively. In management, the systems engineer “discharges his responsibility by designing consistent and mutually enhancing relationships between social structure, information channels, and managing policy.” An ambitious attempt to realise this vision may be found in the Systems Engineering concept implemented at the Bell Telephone Laboratories. There, the systems engineers, senior people with an outstanding record of earlier scientific and technical achievement, also carry out the bulk of technological forecasting at both strategic and tactical level. Planning may now be understood as the task of designing a system (a corporate policy), flexibly changing sub-systems (corporate strategies) and the ways to operate them (operational plans). Technological forecasting furnishes much of the material to build the system. Management today is mainly the stationary engineer’s affair. The manager is a component in the operating system. This ought to remain that way, because there is a well-defined split of management tasks at the strategic and policy levels where a large number of alternatives have to be simulated and evaluated in an impartial way: the planner prepares the basis for decision-making as objectively as possible the manager makes a subjective decision-but, with good planning, in knowledge of the consequences. The planner designs alternative systems configuration, the manager operates one of them. At strategic and policy levels, the planner and the decision-maker should never be identical, or else one or other of their

functions will invariably suffer, the tedious and complex planning effort being more likely to deteriorate first. At tactical level, once the choice for one particular strategy has been made, planning and decision-making can be put in the same hands. The prerequisite for this is the

formulation of a clear policy, or, in other words, the design of the overall system which is to be operated. Technological forecasting, as part of planning, depends on decentralised initiative and central synthesis. This implies that each of the six forecasting stages (plus one feedback evaluation stage), as outlined in Figure 1, is carried out by means of interaction between different layers of the corporate structure. Figure 4 attempts to outline a possible scheme for the various forecasting stages along the RDT&E phases of a technological innovation. The decisive role of horizontal, corporate level staff groups in synthesizing technological forecasts is brought out clearly; for not fewer than three or four out of the six

49

forecasting stages, the synthesis is best entrusted to such staff groups. This is the principal reason for the preference to be encountered for corporate level technological forecasting groups, either alone or in combination with permanent groups in the research laboratories and in the operating divisions, as far as companies carrying out strategic corporate planning are concerned.* The responsibility for synthesizing technological forecasting at the three different planning levels may thus be shared as follows : Technological forecasting for policymaking at the Board level, by special Officers to the Board (the solution of an American electronics company), or by Policy Committees. Technological forecasting for strategic decision-making at the corporate level, e.g. by horizontal staff groups (usually incorporating mixed scientifictechnical and marketing expertise), working for the President or a Senior Vice President in charge of corporate development. Technological forecasting for tactical decision-making at divisional management level, also preferably by staff groups. The complete organisntional framework of technological forecasting may thus also be viewed as unfolding by the interaction of synthesis at three different levels of the corporation. But the involvement of creative people at all hierarchic levels, and the stimulation of decentralised initiative, are decisive for the technological forecasting input into planning. In advancedthinking companies, top management considered as one of its principal tasks the creation of self-motivation at all hierarchic levels by translating the implications of policy-planning all the way down to the tactical forecasting and planning level. A particularly effective way of stimulating creativity has been found in comprehensive joint forecasting “rounds” involving management as well as key technical peop1e.t

The flexibility required by advanced corporate planning cannot be attained in the administrative and operating structure. Therefore, a typical and good solution is to superimpose a flexible “innovation emphasis structure” over the more rigid administrative and operating structure.

Corporate

I i

I

I I Strategies !‘I I I I

I

I

Unilever reports excellent results from such a joint effort involving all its research laborawith several successive tories and taking, refinement and modification steps, a full year. Similar satisfaction has been expressed by a big American electro-technical company.

50

i

I

I I I

Tactical Action Programs

EMPHASIS FIGURE 5. “INNOVATION STRUCTURE” (example from an American electronics company). This structure, which is superimposed over the administrative structure, resembles a relevance tree and enforces normative thinking at all hierarchic levels.

An American electronics company established an “innovation emphasis structure” as shown in Figure 5, with annual updating of IO-year strategies. To make this scheme work with only a very small central synthesis group for strategic planning, the principle was adopted of, eventually, having 1,000 “general managers” (denoting people at different hierarchic levels, but with equally free access to pertinent information as the company president) throughout the company. Board-

I

- - -;

Chief Executive Oficer --’

]--Policy

Planning

I

I

I

Corporate Development

Operations (“Present”)

&s An OECD survey (E. Jantsch, Technological Forecasting in Perspective, Paris 1967) found such corporate level staff groups with the task of synthesizing or carrying out technological forecasts in 20 out of 23 American, and in 26 out of 39 European companies considered. However, in all but one American, and in 16 European companies, they were supplemented by forecasting groups in the research laboratories and the operating divisions.

Pbjectives

Coraegic Level R & D Tactical

is established, as shown schematically in Figure 6. Such a scheme asks for a particularly strong Chief Executive Officer: maintaining the balance between his two genera!s, one of whom commands the troops, and the other the general staff. In such a scheme, the capacities for forecasting and planning, corresponding to the three levels-policies, strategies, tacticsare clearly separated also in the corporate structure; at the same time, the separation of planning and decision-making is emphasised. Industry today is on its way to make Forrester’s vision3 come true: “To deal with the system of which engineering is a part and thereby to make engineering more effective within that system, the engineer must understand the system components and structure. He needs an insight into the nature of the corporation, the tasks of top management, human motivation, the relationship of the individual to the organization, and the psychology and processes of innovation and change. The engineer could become a ‘change agent’ to precipitate improvements in our social systems . . . He would try to clarify the enduring goals and objectives f3r his organization and the people within it . . . He would give more attention to the surrounding social system as a whole rather than as an array of isolated parts . . . He would consider the transient and steady state behavior of his organization from the same system viewpoint that he approaches complex physical systems. He would strive to perfect models of social processes that permit simulation studies leading to a better understanding of organization, information links, and policy. And he would bring to actual human organizations the courage to experiment with promising new approaches based on a foundation of design.” Technological forecasting is a means to achieve that foundation of design.

: Characteristically,

in the two American companies which first introduced such a basic corporate structure, this was the President who later moved up to become Chairman of the Board.

Planning

Planning

I I I I 1 1-1 FIGURE 6. BRINGING INNOVATION INTO SHARP FOCUS THROUGH A HIGH-LEVEL SPLIT BETWEEN THE “PRESENT” AND THE “FUTURE”. (Examples from two American companies.)

References The New Industrial State, Boston 1967. However, the current importance of goals imposed upon society by the “technostructure”, has certainly been exaggerated by Galbraith. c-3 Jay W. Forrester, Common Foundations Under!ying Engineering and Management, /EEE Spectrum, September 1964; and: The

Technological innovation can best be focused when high-level split between the present and the future, or between Operations and Corporate Development,

The Academy of Management Meeting, 30 December 1964, Chicago, IllinoisCommon Foundations (3) Jay W. Forrester, Underlying Engineering and Management, /E&i! Spectrum, September, 1964.

Divisional

R& D

(1) John K. Galbraith,

Structure

Underlying

Management

LONG

RANGE

Processes.

PLANNING