Assessing massive maintenance processes: an empirical study

We present an empirical study from the experience of a major. international software enterprise in conducting massive adaptive maintenance projects with a close deadline. The adopted process entails the decomposition of the application portfolio into loosely coupled work-packets that can be independently and incrementally worked out by teams distributed on different sites. The study analyzes the correlation between maintenance size and productivity metrics of a large Y2K project. The resulting models allows to estimate the costs of a project conducted according to the adopted massive maintenance process and distribute them among the different phases.