6.2 SOFTWARE OBSOLESCENCE MECHANISMS
The key enabler for proactive and strategic management of DMSMS obsolescence is the ability to forecast the obsolescence events for key hardware and software elements in systems.
Most long-term obsolescence forecasting for hardware (electronic parts) is based on the development of models for the part’s life cycle. Traditional methods of life cycle forecasting are ordinal scale–based approaches, in which the life cycle stage of the part is determined from a combination of technological and supply chain attributes, such as level of integration, minimum feature size, type of process, number of sources, and so on (Henke and Lai, 1997; Josias and Terpenny, 2004), and those available in several commercial databases. Market data has been used to forecast life cycle curves, from which obsolescence forecasts are obtained (Solomon et al., 2000; 2007), and data mining–based obsolescence forecasting approaches that leverage large commercial databases of electronic parts have also been developed (Sandborn et al., 2007; 2011). None of the methods developed for forecasting hardware obsolescence have been applied to software.
Forecasting software obsolescence will involve finding a metric that is directly related or highly correlated to the root cause of the software obsolesce. The methodology behind forecasting software obsolescence for each mode of obsolescence will be different; however, all methods in general will involve gathering metric data, finding ...