CMMI Model Selection and Scope

USA selected the CMMI Systems Engineering/Software Engineering version 1.02 model with a staged representation for the case study (version 1.1 was not yet available). This model version exhibited adequate shelf life and its use would result in only minor modifications during CMMI version 1.1 development. The use of version 1.02 would not generally negate the results of the case study.

The CMMI model selected most accurately reflected the core businesses within the company: software engineering. There was no need to review the draft IPPD or acquisition practices since these were not considered a part of the typical business profile of a USA project. USA selected the staged representation of the model because the case study participants were familiar with previous staged models (i.e., SW-CMM) and because the case study did not intend to isolate a subset of process areas for intense scrutiny (as the continuous representation would facilitate).

Since the CMMI model contains nearly 500 practices and the amount of time to conduct the case study was limited, only a subset of the CMMI practices was evaluated for each project. The case study's purpose was to sample the examination of process areas (PAs) suitable to selected projects in a planned and meaningful way to determine how well CMMI could be implemented at USA. As such, not all PAs at all maturity levels were investigated in total for each project. However, between the two projects, all of the practices were examined, but not all practices were examined for each project.

Different criteria were used for the two flight software projects to determine which CMMI practices to apply. Since the PASS project conducted SW-CMM self-assessments in 1996 and 1999 with results that mapped evidence to key practices, there was no perceived need to revisit the SW-CMM key practices at maturity levels 2 and 3, which are unchanged between SW-CMM and CMMI. The CMMI scope for PASS was based on significant differences from SW-CMM 1.1 to CMMI and high-maturity practices. In addition, all CMMI maturity level 4 and 5 PAs were examined.

The CMMI scope for the CAU project was based on its life-cycle progress (i.e., process areas and practices that can be reasonably applied to CAU at its current life-cycle stage). With the CAU project early in its life cycle, limited defined processes were operating. It was not feasible to assess CAU planned processes not yet established or operating.

Figure 7.1 identifies the process areas examined for each project. Each project considered all practices for the process areas examined. For example, all SPs and GPs for the Requirements Management PA were examined for CAU and all SPs and GPs for the Measurement and Analysis PA were examined for PASS.

Figure 7.1. CMMI Process Areas Examined by Project

The case study does not represent a comprehensive Standard CMMI Appraisal Method for Process Improvement (SCAMPI) appraisal of all PAs for all maturity levels for representative organizational projects. USA anticipates that future efforts will more broadly cover the PAs and practices across the company.