1. It's smaller, sort of.
Version 1.3 of the CMMI has 356 Practices in Maturity Level Three. These included all of the Generic and Specific practices across eighteen Process Areas.
Version 2.0 eliminates the Generic Practices altogether (whoo hooo?), combines Requirements Management (REQM) and Requirements Development (RD) into one Practice Area, and mashes up Validation and Verification. Great! But wait....there is a NEW Practice Area called "Implementation Infrastructure" (II) that takes the place of the Generic Practices. II must be applied to all Processes in your process set, effectively ,multiplying the, across many Practice Areas. They've also added Causal Analysis and Resolution (CAR) and Governance (GOV) to ML3. So...smaller? Yes, but not as much as you might think.
2. Practices are no longer very Specific....or Generic
After spending a lot of time learning about "Specific," "Generic," and "sub-practices," for v1.3, we all thought we had the language mastered. But Noooo! Now there are just "practices" inside of the newly-introduced "Practice Groups." The "sub practices" are completely gone, with just some implementation advice at the bottom of the page.
3. It's not your father's architecture
The CMMI v1.3 architecture was, roughly: Process Category------>Process Are------>Specific/Generic Goal------->Specific/Generic Practice------->Informative Material.
The CMMI V2.0 architecture is Category------>Capability Area------>Practice Area------->Practice Group-------->Practice--------->Explanatory Materials
Think of categories as "SuperProcess Categories," and think of Capability areas as "Process Categories" but better,
4. You won't have SCAMPI to kick around anymore
SCAMPI A, B, and C are gone!
New we have an "Evaluation Appraisal" (like a B or C). and a Benchmark Appraisal (like a SCAMPI A). We still have the After Action Re-Appraisal, and there is a new "Sustainment" appraisal, which can give you an extension for two extra years if your organization hasn't change significantly.
5. They're picking the sample
In v1.3, the project sample for an appraisal is negotiated between the sponsor and the Lead Appraiser. In V2.0, the CMMI Institute will randomly generate your sample NO MORE than 60 days prior to your Benchmark appraisal. This means that ALL projects need to be performing at MLx - not just a pre-selected few that get polished up for when the appraiser comes onsite!
BONUS: You'll be speaking a new language!
Specific Practice, Generic Practices, and Sub-Practices are now "Practices."
Process Areas are now "Practice Areas."
SCAMPI is now "Benchmark."
Categories is now "Capability Areas."
SCAMPI B/C is now "Evaluation."
Like this blog? Forward to your nearest engineering or software exec!
Jeff Dalton is a Certified SCAMPI Lead Appraiser, Certified CMMI Instructor, author and consultant with years of real-world experience with the CMMI in all types of organizations. Jeff has taught thousands of students in CMMI training classes and has received an aggregate satisfaction score of 4.97 out of 5 from his students.
Visit www.broadswordsolutions.com for more information about engineering strategy, performance innovation, software process improvement and running a successful CMMI program.
Awesome summary. Broadsword is the way to go if you want the bottom line on CMMI.
Post a Comment