Technology forecasting

Technology forecasting is a technique to forecast the future characteristics of useful technological machines, procedures or techniques

Primarily, a technological forecast deals with the characteristics of technology, such as levels of technical performance, like speed of a military aircraft, the power in watts of a particular future engine, the accuracy or precision of a measuring instrument etc. The forecast does not have to state how these characteristics will be achieved.
Secondly, technological forecasting usually deals with only useful machines, procedures or techniques. It excludes services or techniques intended for luxury or amusement.
Methods of technology forecasting

Commonly adopted methods of technology forecasting include the Delphi method, , growth curves and extrapolation.

The Delphi method relies on a panel of experts. The expert answers questionnaires in two or more rounds and after each round a facilitator provides the summary of the answer along with the reason to the participants. . Thus, participants are encouraged to revise their earlier answers in light of the replies of other members of the group. It is believed that during this process the range of the answers will decrease and the group will converge towards the "correct" answer. Based on a pre defined stop criteria the process is stopped and the mean or median scores of the final rounds determines the result.

A growth curve is a process where values for the measured property like population, body weight etc are plotted on a grapg as a function of time.

Extrapolation is the process of constructing new data points outside a discrete set of known data points. Its results are often less meaningful than interpolation, and are subject to greater uncertainty.

Reasons for combining forecasts
The reason forecasts go wrong because they ignore related fields. A given technical approach may have been superseded by another technical approach which the forecaster might have ignored. .
Inconsistency between forecasts is another important reason for failure. Because of these problems, it is often necessary to combine forecasts of different technologies. Therefore it is better to combine forecasts obtained from different methods rather than concentrating on a single method. If this is done, the strengths of one method may help compensate for the weaknesses of another.
Trend curve and growth curves
The growth curves and a trend curve are used in combination to allow the forecaster to draw some conclusions about the future growth of a technology which might not be possible, were either method used alone.
A single method like growth curve might not be enough to say anything about the time at which a given technical approach is likely to be supplanted by a successor approach.
Trend curve alone cannot say anything about the ability of a specific technical approach to meet the projected trend, or about the need to look for a successor approach. Thus the need for combining forecasts.
Identification of consistent deviations
We assume that the scatter data points about a trend curve are due to random influences which are neither controllable nor measurable. However, consistent deviations may represent something other than just random influences
The opportunity to apply an analogy comes where such consistent deviations are identified. The typical events which bring about deviations from a trend are wars and depressions. Thus the usefulness of combining analogies with a trend forecast is to predict deviations from the trend deviations which are associated with or caused by external events or influences.
As with other uses of analogy, it is important to determine the extent to which the analogy between the event used as the basis for the forecast, and the historical model event, satisfies the criteria for a valid analogy.
Forecasts of different technologies
We can get different forecasts from a single technology but combining these forecasts as less important than combing forecasts derived from different technologies because of the fact that technologies may interact or be interrelated in some fashion. Another reason for this is that of consistency in an overall picture or scenario. “One of the simplest examples of interacting trends is the projection to absurdity, i.e. simply projecting the given data indefinitely without getting any specific result. For instance, if one simply projects recent rates of growth of world population, one arrives at some fantastic conclusions about the density of population in a particular place by various dates in the next millennium.”(wikipedia)

Some other trends which can confidently be expected to not continue indefinitely are:
Annual production of scientific papers.
Number of automobiles per capita.
Kilowatt hours of electricity generated annually.
Another instance of interacting trends was in the case of the number of scientists in the U.S. growing faster than the overall population. Since 1940s through the 1960s, science as an activity in the United States grew exponentially. The number of dollars spent on R&D was growing faster than the GNP (in the 1960s).
If projected indefinitely, these two curves would give the result that eventually every person in the U.S. would be working as a scientist and the entire GNP would be devoted to R&D alone, which are however absurd conclusions. Thus it is clear that the scientific discipline of technology forecasting is not mere trend extrapolation but also involves combining forecasts
Uses
The method has considerable use in the manufacturing industry.


References

http://en.wikipedia.org/wiki/Technology_forecasting

. Klopfenstein, Bruce K. "Forecasting consumer adoption of information technology and services - Lessons from home video forecasting". Journal of the American Society for Information Science 1989 Jan;40(1):17-26.
Martino, Joseph (January 1983). Technological Forecasting for Decision Making, 2nd edition, North-Holland Makridakis, Spyros; Steven C. Wheelwright, Rob J. Hyndman (December 1998). Forecasting: Methods and Applications, 3rd edition, John Wiley