.

ACAR 2012

Technology Capability Assessment & Technology Forecasting

AIL ACAR

KADIR HAS UNIVERSITY 2012

Technology Capability Assessment & Technology Forecasting

The word technology has a wider connotation and refers to the collection of production possibilities, techniques, methods and processes by which resources are actually transformed by humans to meet their wants. Ferr (1988) has defined technology as practical implementations of intelligence. However, Gendron (1977) has provided a more comprehensive definition: A technology is any systematized practical knowledge, based on experimentation and/or scientific theory, which is embodied in productive skills, organization, or machinery.

The role of technology in fostering economic growth of nations and enhancing their industrial competitiveness has been widely recognized, through its domineering influence over industrial productivity. Further, technology has emerged as the most important resource that contributes directly to socio-economic development. Hence, technology is viewed from various perspectives: as an engine for economic development, as a strategic resource, and as a competitive weapon. This necessitates effective management of technology - at both national and firm levels. Technology Management (TM), which inter alia aims at planning and developing the technological capabilities of an organization or a nation, has now occupied the centre stage of decision-making.

Gaynor (1989) has provided the following description of TM: Managing technology is a method of operation that leverages human resources, technology and other business assets by optimizing the relationships between the technology functions of the business enterprise. It is the process of integrating science, engineering and managing with research, development and manufacturing in order to meet the operational goals of the

business unit effectively, efficiently and economically. It includes managing the totality of the technology operations from concept through commercialization. TM embraces several interconnected issues such as: technology policy; technological forecasting and assessment; technology strategy; technology transfer; technology-induced as well as market-oriented Research and Development (R&D); process technology and product technology and their continuing improvement; human resource management in terms of innovative capabilities, flexibility and contribution; and technology project management. Gaynor (1989) has provided the following description of TM: Managing technology is a method of operation that leverages human resources, technology and other business assets by optimizing the relationships between the technology functions of the business enterprise. It is the process of integrating science, engineering and managing with research, development and manufacturing in order to meet the operational goals of the business unit effectively, efficiently and economically. It includes managing the totality of the technology operations from concept through commercialization.

Technological Change

Theories And Measurement

Technological change has been broadly defined as: the process by which economies change over time in respect of the products and services they produce and the processes used to produce them. More specifically, it has been termed as:

Alteration in physical processes, materials, machinery or equipment, which has impact on the way work is performed or on the efficiency or effectiveness of the enterprise. Technological change may involve a change in the output, raw materials, work organization or management techniques - but in all cases it affects the relationship between labor, capital and other factors of production. While the policies to stimulate technological progress and productivity growth - at both national and firm levels - must be formulated in a broad socio-economic context, their focus ought to be on the internal dynamics of technological change. It has been suggested that the knowledge pertaining to technological change in the less developed countries would be a crucial input to the understanding of the phenomenon in industrialized nations. In this light, the following paragraphs provide a selective survey of theories and measurement of technological change (Saren, 1991; Sharif, 1986; Stoneman, 1983; Saviotti, 1985).

Measurement Of Technological Change

Economic Indices
Arithmetic indices are derived based on price variations in capital and labor in relation to the industrial output. Technological change is measured as the weighted average of the change in factor prices, holding inputs constant. Solow derived a geometric index based on the premise that technological change is equal to the change in output not accounted for by the changes in labor and capital.

Patents
Patent statistics have been used as indicators of technological change. They have also been used to analyze the diffusion of technology across firms or industries or countries. Patent studies are also concerned with the analysis of the innovation process itself in order to assess and evaluate the output of research activity.

Rate of improvement of technology

A figure of merit for each functional capability of a technology is to be identified. For a product it could be travel speed per unit time (transport vehicles), or lumens per watt (lighting fixtures), or instruction execution rate per second (computers). For a process, it could be the efficiency of fuel utilization or reduction in waste generation. S-shaped growth curves are formulated to form a system of curves depicting advancement of technologies.

Rate of Substitution of Technology

It is determined on the basis of relative changes in the market shares of two technologies or two sets of technologies. Cumulative pattern of gain in market share by a technology exhibits S-shaped growth.

Rate of Diffusion of Technology

It represents the cumulative number of adopters of a new product, material or process. This follows an S-shaped curve.

Use of Technology Forecasts

The purpose of any type of forecasting and the proper role of the forecaster is to assist the contemporary decision-makers in the choosing of policies and making of plans that are most promising. In the National Context: Developing technological competencies so as to meet global competition and international trading imperatives. Planning for creation of sustainable comparative advantages in select technological thrust areas. Planning for the well-being of citizens with the aid of technological innovations.

In the context of Business Firms: Establishing technical parameters and performance standards for new products and processes. Augmenting new product development efforts as well as improvement of existing products. Enabling better timing for new technology introduction and facilitate take-to-market strategy formulation. Aiding prioritization of research programmes and identification of techno-scientific skills required for the same. Identifying major opportunities and challenges in technological environment and offering guidance for technological planning. One of the main aspects of Technology Forecasting is its communication aspect. Technology Forecasting (TF) initiates and fosters the communication between various communities such as: Science and science (inter-disciplinary fields) Science and Technology Industry and politics
6

 Technology and public administration Technology and the general public.

Technology forecasts can be a short, medium or a long-term exercise. Short term forecasts are of usually a year or less, might typically deal with a single technology. Medium-term forecasts might cover a 2-10 year period. Long-term forecasts cover 10-20 years - a time horizon long enough for totally new technologies to emerge. Longer the time frame, tougher it is to predict what is in stock for the future pertaining to technology changes. Sometimes, forecasts misfire because of a fascination of ones own technology and also due to the enthusiasm it generates among market analysts and magazine writers. A classic example is the bold forecast that NASA made regarding the huge sun-reflecting satellites to illuminate night-shrouded areas of the earth which they predicted would be possible by the mid-1970s. Such forecasts with faulty timelines can be totally worthless to corporate planners, and worse they can turn into money pits. This brings into focus the need for employing accurate technology forecasting methods. The most appropriate choice of forecasting method depends on: What is being attempted to forecast Rate of technological and market change Availability and accuracy of information The planning horizon The resources available for forecasting

Classification Of Technology Forecasting Methods

The technological forecasting methods can be classified as exploratory and normative forecasting methods. Exploratory technological forecasting starts from todays assured basis of knowledge and is oriented towards the future, while normative technological forecasting first assesses future goals, needs, desires, mission, etc., and works backwards to the present.. Explorative forecasting is more focused on predicting how a new technology will evolve on a pre determined curve, which is S-shaped, while the normative forecasting attempts to be more pro-active. With so many choices available, each having distinct characteristics, a calculated choice has to be made while choosing a
7

forecasting method. Some of the factors that influence the choice of the forecasting methodology can depend on the time frame for the forecast, how much precision is needed in the prediction and the purpose of the forecast

Figure 1.1 Technological Forecasting Methods

Technology Life Cycle (TLC)

The nature of a technology and its implications to firms have been observed to undergo perceptible changes over its life span. Although it is correct to view a TLC as a continuum, it is more convenient to consider it as a series of discrete stages for the purpose of better understanding and analysis. A technology typically evolves through the following stages in its life cycle. Figures below broadly portray the various aspects that characterize these stages.

Figure 1.2 Industry and Market Structures over aTLC

Figure 1.3 Characteristics of Technology over a TLC

Cutting-edge: This stage refers to the birth of a new technology. The primary focus of entrepreneur(s), here, is on its Research and Development (R&D) to a demonstrable form. Nevertheless, the need for financial support for R&D and for testing prototypes is also a pressing concem at this stage. Although basic research forms a major part of the effort in this stage, firms seldom do it without a specific application in mind. However, the scope of such an application may not be clearly known to the firm and the related knowledge may still be abstract. Hence, the target market, and the feasibility and viability of R&D at this stage are also uncertain. State-of-the-Art: This stage in a TLC begins with the deployment of cutting-edge technology to solve customers problems. The customer base state-of-the-art technology is usually small, but sophisticated. Hence, technical specialists are, perhaps, ideal for marketing the technology. Some of the functions, such as engineering, manufacturing, finance and administration may have begun formally while others may still be embryonic at this stage. Many other important functions are carried out with the help of hired consultants or agencies. The market witnesses a high rate of product innovations and as a result a great product diversity can be observed. The production process will be usually nonstandard. Hence, the state-of-the-art technology can respond easily to the varied market requirements, but only ineffectively and inefficiently. Advanced: The gradual standardization of technological characteristics leads to rapid market expansion. The market largely consists of less sophisticated customers who seek all the benefits of state-of-the-art technology, but have no desire to develop the technical expertise needed to understand the technology. This market calls for professional marketing. The potential profits attract a large number of competitors and hence, the technology is no longer of the hi-tech variety available with only a few suppliers. A market shakeout, segmentation, and further standardization are bound to occur. A firms survival in the market beyond this stage indicates its relatively efficient operations. At this stage, a part of the production system is likely to be automated, and division of labour is more pronounced. However, there is a possibility that the rapidly increasing demand may cause manufacturing and marketing capacity crises which will call for a major transformation in the organization structure. Parallelly, financial and administrative crises are also likely. Mainstream / Mature: As technology enters this stage, the scope for further product innovations reduces and process technology becomes the target of innovative effort of firms. However, gradually, even the production process gets so well integrated that the possibility of
10

any major changes in product or process technological elements will be limited and product differentiation becomes difficult. Hence, the strategic thrust shifts to efficient and economic production for minimizing costs. This may be achieved by huge capital investments or shifting the production base to the Third world. When the production costs also cease to offer comparative advantage, the competition will shift to customer service. This may help the firm to maintain and improve upon growth momentum and market position. In this stage, the firm will be better managed by a professional management team than entrepreneur/owner. The organization will have grown in size and will call for more formal structures, communications, and systems. Decline: During this stage, the scope for further improvement of the technology diminishes rapidly. The increasing marginal cost of its improvement, coupled with the functional superiority of the next generation technology, results in the older technology giving way to the new one gradually. However, it may not be easy for the new technology to wipe out the older technology one and pervade the market. The overlap between the above successive stages of a TLC makes it a continuous process rather than a set of discrete stages (Mohan Babu and Ganesh, 1997).

11

TECHNOLOGY FORECAST METHODS

Families of Technology Forecast Methods There are hundreds of TF Methods, which can be fit into 9 families [2]: Expert Opinion, Trend Analysis, Monitoring & Intelligence, Modeling & Simulation, Scenarios, Statistical, Descriptive, Creativity, and Valuing/Decision/Economics Methods. Lists methods in each family as: 1) Expert Opinion Delphi (iterative survey) Focus Groups [panels, workshops] Interviews Participatory Techniques 2) Trend Analysis Trend Extrapolation [Growth Curve Fitting] Trend Impact Analysis Precursor Analysis Long Wave Analysis 3) Monitoring and Intelligence Methods Monitoring [environmental scanning, technology watch] Bibliometrics [research profiling; patent analysis, text mining] 4) Statistical Methods Correlation Analysis Demographics Cross Impact Analysis Risk Analysis Bibliometrics [research profiling; patent analysis, text mining] 5) Modeling and Simulation Agent Modeling Cross Impact Analysis Sustainability Analysis [life cycle analysis] Causal Models
12

 Diffusion Modeling Complex Adaptive System Modeling (CAS) [Chaos] Technological Substitution Scenario-simulation [gaming; interactive scenarios] Economic base modeling [input-output analysis] Technology Assessment 6) Scenarios Scenarios [scenarios with consistency checks; scenario management] Scenario-simulation [gaming; interactive scenarios] Field Anomaly Relaxation Method [FAR] 7) Valuing/Decision/Economics Methods Relevance Trees [futures wheel] Action [options] Analysis Cost-benefit analysis Decision analysis [utility analyses] Economic base modeling [input-output analysis] 8) Descriptive and Matrices Methods Analogies Backcasting Checklist for Impact Identification Innovation System Modeling Institutional Analysis Mitigation Analysis Morphological Analysis Roadmapping [product-technology roadmapping] Social Impact Assessment Multiple perspectives assessment Organizational analysis Requirements Analysis [needs analysis]

13

9) Creativity Brainstorming [brainwriting; nominal group process (NGP)] Creativity Workshops [future workshops] TRIZ Vision Generation Science Fiction Analysis We will briefly review some of the most popular methods in each category.

EXPERT OPINION Expert Opinion methods include forecasting or understanding technological development via intensive consultation with subject matter experts. The most popular method in this family is the Delphi Method. This method combines expert opinions concerning the likelihood of realizing the proposed technology as well as expert opinions concerning the expected development time into a single position. In Delphi, a sequence of individual interrogations is followed by information and opinion feedback derived from analyzing the initial response data. This feedback, which includes the reasoning and/or justification behind each individual expert's forecast, allows the other experts to revise their forecast in light of the new information. A single acceptable forecast is typically agreed upon after several rounds of this process . Delphi, being the most widely used technique, has been subjected to scrutiny by many authors.

TREND ANALYSIS Trend analysis involves prediction via the continuation of quantitative historical data into the future. Trend analysis is a broad term that encompasses economic forecasting models and techniques such as regression, exponential smoothing and Box-Jenkins' ARINA model and growth curve fitting . A technology usually has a life cycle composed of several distinct stages. The stages typically include an initial adoption stage, a growth stage, a maturity stage and a declining final stage. Growth curve forecasting is based on the parameter estimation of a technology's life cycle curve. The growth curve forecasting method is helpful in estimating the upper limit of the level of technology growth or decline at each stage of the life cycle. This method of forecasting is also helpful in predicting when the technology will reach a particular life cycle stage.
14

One type of growth curve forecasting method is the Fisher-Pry Analysis. It is a mathematical technique used to project the rate of market adoption of technically superior new technologies and, when appropriate, to project the loss of market share by old technologies . The technique is based on the fact that the adoption of such new technologies normally follows a pattern known by mathematicians as the "Logistic Curve." This adoption pattern is defined by two parameters. One of these parameters determines the time at which adoption begins, and the other determines the rate at which adoption will occur. These parameters can be determined from early adoption data, and the resulting pattern can be used to project the time at which market takeover will reach any given level. Results produced by this technique are highly quantitative. The technique is used to make forecasts such as how the installed base of telecommunications equipment will change over time, how rapidly a new chemical production process will be adopted, and the rate at which digital measuring devices will replace analog devices in petroleum refineries, etc.

MONITORING AND INTELLIGENCE METHODS

Monitoring and its variations, Environmental Scanning and Technology Watch, are suitable for making one aware of changes on the horizon that could impact the penetration or acceptance of the technologies in the marketplace. Resource availability is one of the scoping issues associated with these methods since a number of the scanning approaches require the use of experts. Expert panels are created to look out for changes on the horizon that could be important to implement or accomplish plans. Experts are also tracked in a scan the scanners manner. TF analysts identify the experts in a field and keep track of those individuals by making occasional contact with them, observing them at conferences or searching the Internet for insights they may have posted.

STATISTICAL METHODS In the Statistical Methods family, the most popular methods are Correlation Analysis and Bibliometrics. Correlation analysis forecasts the development patterns of a new technology when the development patterns of the new technology are similar to those of existing technologies. Use of this method presupposes that data regarding the development patterns of the existing technologies are available.
15

Many new or potential tools, currently used in future studies, have resulted from advances in information technology and information science. Among them, stand out scientometrics and bibliometrics, tools used traditionally by the information science experts to measure scientific productivity and to identify science and technology networks.

MODELING AND SIMULATION A model is a simplified representation of the structure dynamics of some part of the real world. Models can exhibit future behavior of complex systems simply by isolating important system aspects from unessential detail. Modeling requires a good understanding of interactions between these forecasts and the underlying variables or determinants. One example in this family is Agent Modeling. An agent model involves the creation of computer generated agents that populate a computer screen, and interact with one another according to a set of behavioral rules [24]. The agents may be of different species; that is, they may have different attributes and may be assigned different rules. Their interaction over time is usually simulated by successive plays of the rules as the evolving attributes and spatial positions of the agents are computed. The spaces in the environment in which the agents are placed may also contain rules.

SCENARIOS Scenario writing proposes different conceptions of future technology. Each conception of the characteristics of the future technology is based on a well-defined set of assumptions. A scenario represents alternative characteristics of the future technology, with each alternative being based on certain assumptions and conditions. The forecaster evaluates the validity of the assumptions. The results of this evaluation are used to determine the scenario most likely to occur .

VALUING/DECISION/ECONOMICS METHODS The most popular method in this category is the relevance tree approach. This is a normative approach to TF. The goals and objectives of a proposed technology are broken down into lower level goals and objectives in a tree-like format. In this way, the hierarchical structure of the

16

technological development is identified. The probabilities of achieving the goals and objectives at the various levels of technological development must be estimated. The probabilities can then be used to forecast the likelihood of achieving the stated goals and objectives of the proposed technology.

DESCRIPTIVE AND MATRICES METHODS A growing activity in this category is technology roadmapping, which projects major technological elements of product design and manufacturing together with strategies for reaching desirable milestones efficiently. Roadmaps typically run several technology or product generations (e.g., 2 to 10 years) ahead. In its broadest context, a science and technology roadmap provides a consensus view or vision of the future science and technology landscape available to decision makers. Thus, the predictive element emphasized in early TF is supplemented with a normative element, that is, however, narrower, more targeted, and more directly actionable than is the normative element implicit in TA. In the past, the institutional champions for roadmapping were mainly military industrial organizations; more recently, they have been other large corporations and industry associations.

HOW TO EVALUATE THE QUALITY OF TF METHODS? Evaluation of TF methods is quite challenging. Evaluation should establish, as far as possible, how far an activity has achievedor how far it appears to be achievingits intended outcomes. Yet, here is no general-purpose toolkit for evaluating TF studies influence and outcomes. A key challenge is establishing where a TF process begins and ends. Also, determining the extent to which an activity would have taken place without the intervention of the TF is problematic.

CHOOSING A FORECASTING METHOD A large number of methods have evolved for TF, but the quality of forecasts greatly depends on proper selection and application of appropriate methods. The application demands that the technique used need to be time-, space- and technology-specific. Yet, there is little research done on matching the TF methods techniques to a particular technology. One such study comes from, Levary and Han, who have considered three basic factors,

17

namely the extent of data availability, the degree of data validity and degree of similarity between proposed technology and existing technologies. Each factor has been categorized into cases as small/low, medium/moderate, large/high and their combinations. According to [6], given a small amount of low or medium validity data, and no similarity between proposed technology and existing technologies, a reasonable choice is a method based on information obtained from a panel of experts.

18