I cannot stress enough how key it is as a NASA team that we strike the right balance between programmatic and institutional authority and accountability over program content. Our senior agency leadership has devoted considerable attention to this over the last two and a half years, and as Constellation continues to mature we are able to test those relationships at ever finer levels of detail. This is a healthy thing, and it is being done with great care by all involved. I'm immensely proud to be part of this agency at this juncture in time, in that we all - together - have a chance to fundamentally improve upon the lessons of the past and create the healthiest possible environment within which to execute Constellation's challenging mission...
Mike G has encouraged me to share the following with the team, in hopes that this helps to clarify his position on how 'governance' explicitly relates to Constellation and the institutions executing our plan...
It is NASA's job to figure out a way to "support" ESMD and Cx. Not the other way around. A "successful" independent review board, or review process generally, is one that serves the Program Manager. Not the reverse.
The institution devises and imposes standards and processes solely for the purpose of aiding programs to be successful, not as a matter of doctrinaire adherence to a standard of "technical purity". There is no such thing as a "successful" independent review board, or a "successful" NASA, without a successful Cx. If 7120.5D does not serve Cx well, meaning help produce a good outcome, it is the document that will be revised, not Cx.
Cx is not accountable to review committees, reviewers, OCE, IPAO, or PA&E. Cx is accountable to the AA/ESMD, and thence to Code A. All other institutional functions and staff offices are tools -- MY tools -- to be used as an aid in producing a good outcome. The primary function of any review is to allow us, all of us, to detect problems early and correct them while it is easiest to do so.
Reviewers do not manage the program. Information derived via review is properly furnished to program management, up to the Mission Director, for decisions. Program management is not accountable to reviewers for their decisions in regard to the advice and recommendations which are provided by reviewers. They are accountable to me.
Let me review the bidding from the top level perspective, addressing the classic variables of program management -- cost, performance, and schedule. At the outset of any program, one can determine the relative priorities of these parameters to be in any desired order, depending on policy, which is set above the PM.
For Cx, budget is prescribed from above -- believe me, I'd alter it if I could -- so that is not a free variable. It is prescribed.
The architecture has been chosen and the technical content to satisfy that architecture has been generally scoped out. I don't wish to alter it, other than on the margins and in response to things that we learn as we go. (One such example was upgrading the 5-seg SRB early, rather than upgrading later for Ares 5.) So that "variable" is now "fixed".
The remaining free parameter is schedule. We will finish when we finish, given the budget which has been allocated and the job we are trying to do.
If the nation's leadership allocates more money, we will finish sooner, and contrariwise. More skillful (and luckier) implementation will allow us to finish sooner, and contrariwise. But we will spend our money going in the right direction until we are done.
We are planning, scheduling, and budgeting in all that we do in Cx to 65% confidence.
That means our formal schedule for IOC for Orion/Ares 1 is March 2015, per our best estimates to date, validated independently by Aerospace Corp. This is an estimate, of course, but the best we have.
Cx is to be judged against that date, and the intermediate milestones which support it. All external reporting will be against the 65% confidence dates and budgets.
Internally, Jeff will (and must) hold his project and element managers to a higher standard; meaning that he and they should work to a more aggressive schedule, such as (for example) a 50% confidence budget/schedule. If they slip, then they do, but it would be foolhardy for them not to try to deliver as early as possible.
The difference between 65% confidence and (say) 50% confidence is Jeff's reserve. In turn, Skip Hatfield, Steve Cook, Tip Talone, etc., should be working their programs to (say) 40% confidence budgets/schedules. The difference is their reserve. Time is money. All should be striving to deliver as early and as cheaply as possible. If everyone does their best, then 65% of the time Cx will deliver, overall, on schedule. (Of course, we only get to run the experiment once!). That's how it works.
Mike
