Evaluating the impact of process improvements quantitatively using process modeling

The concept of continuous proces s improvement, although familiar in the manufacturing arena, is just starting to b e implemented in the software industry . Currently, software developmen t organizations have identified potentia l process improvements and are working to implement results from ISO 9000 audits an d Capability Maturity Model (C M M ) Software Process Assessments . Quantitative methods are needed t o prioritize process changes and to justify resources for proposed improvements base d on overall process performance . However, these methods have not yet been developed. This work uses software process modeling to predict the impact of potential proces s changes on quantitative performanc e measures before a substantial commitmen t of time and resources is made. The work addresses the performance measures o f development cost, source code quality, and project schedule. This analysis can be used to prioritize process improvements based on their potential performance impact and t o obtain management support for the proposed change . The proposed modeling approach i s applied to a realistic example process an d potential process change with assumed task durations and outcomes. This exampl e process and potential process change ar e analyzed using a stochastic simulatio n approach which was developed by Marc Kellner et al. at SEI using the STATEMATE modeling tool .