Probabilistic technological forecasts using precursor events

The author describes the use of the method of maximum entropy for generating a probability distribution for the time lag between demonstration of a device and its market introduction. Historical data are used to generate a probability distribution for the lag times between a specific type of precursor event and a specific type of event to be forecast. The probability distribution can be used to estimate the risks of either planning to commercialize it and possibly allowing a competitor to commercialize it first. Examples from the aerospace and automotive industries are used to illustrate the technique. In particular, attention is given to the lag times between commercialization of several aluminum alloys and the first flight of the first production aircraft to incorporate those alloys.<>