WFIRST Report Released
WFIRST Independent External Technical/Management/Cost Review (WIETR), NASA
“This report responds to the questions asked in the Terms of Reference (TOR) that established the WIETR and includes recommendations and options for NASA to consider. This report is input to NASA in support of its formulation of the WFIRST implementation plan so that the mission is both 1) well understood in terms of scope and required resources (cost, funding profile, schedule) and 2) executable. The WIETR recognizes the scientific importance and timeliness of WFIRST. The objectives of this ambitious mission are driven by the goal of answering profound questions about the Universe beyond our solar system and planet Earth. This ambition comes with challenges that must be recognized and addressed – these are the focus of this report.”
– NASA Decides To Reduce Cost/Complexity of WFIRST, earlier post
The report’s findings are mostly a list of options, with the pros and cons associated with each. For example, is the coronagraph a scientific instrument or a engineering test of the technology? The report didn’t really say, it just said treating it as both, in inconsistent ways, was a problem. But the last finding sounds pretty harsh to me: “The NASA HQ-to-Program governance structure is dysfunctional, and should be corrected for clarity in roles, accountability, and authority.” True or not, that’s pretty strong language for this sort of report.
I follow the reasoning, but this part is puzzling to me:
“The In-house acquisition strategy for the spacecraft also carries risks because of the temptation for continued
improvements; while recognizing that workforce capacity coverage and workforce development at GSFC is an
unstated objective reinforced by the ASM decisions.”
If the acquisition is in-house, wouldn’t GSFC have both control over acquisition costs and responsibility for sticking to the agreed budget? Isn’t this better than an external procurement where the contractor is pretty much expected to lock in the program and then do everything they can to inflate the cost? Unless, of course, we believe that NASA can ensure adherence to cost, schedule, _and_ performance for an external procurement by simply making it a requirement in the contract?
If it’s internal, and being used for “workforce development”, the risk is high that “requirements creep” will become rampant. My opinion, of course, but keeping it all internal means letting changes creep in without much external review. Unfortunately, changes quite often mean both budget increases and schedule slippage.
Combine that with the report’s concern over how this project is being managed and you’ve got a recipe for budgetary and schedule disaster.
I agree this pressure exists. Nowadays there is little institutional funding, most of the money is allocated to “programs” so the actual researchers, engineers and technicians, often support contractors, can lose their jobs if they can’t point to a “requirement” that only they can meet, even if they have years of experience and skills that cannot be replaced.
But simply outsourcing the project doesn’t by itself solve the problem because the contractor has stockholders to satisfy and is driven primarily to increase profits, i.e. costs.
Someone has to have direct responsibility for both accomplishing the mission and controlling the cost. If we look at successful aerospace development programs of the past there is usually a strong leader who combines strategic vision, technical understanding, and budget authority.
Outsourcing the project probably wouldn’t help, but outsourcing parts of it can. I’m not sure about the current structure of WFIRST, but past missions have split work between several institutions. Most of the requirements are “owned” by the project manager and/or the project scientist.
For example, the managers could be at a NASA center, and an aerospace company is subcontracted to build the spacecraft (but not the instruments or telescope optics or handling operations.) In that case, the manager isn’t as likely to care about supporting the workforce at the aerospace company, and more concerned about risk to the project’s budget and schedule. There may be some concern about not alienating the subcontractors (since you many want to work with them on a future project), but that’s a less significant concern.
I’ve even heard some planetary scientists say the best way to keep a project within budget is to have a Principal Investigator and Project Manager at institutions with only a small fraction of the project’s budget. Some of the more successful, recent missions (e.g. Juno or MAVEN) have been done that way, with only a couple instruments and some science operations at the PI or PM institutes.
From watching many of these projects over the years I have observed that the actual workaday scientists have little actual project authority, relegated instead to a ‘program code’, as it has bee described.
Not necessarily. The project scientist or principal investigator is a scientist (or scientist who has been promoted into management.) Less senior scientists are often the ones writing and justifying the requirements. And the various advisory groups (MEPAG, OPAG, etc.) give scientists some opportunity to say things like, “involve the entire community; don’t make it a single-NASA-center project.” I’ll admit that advice may or may not be listened to, but it’s better than nothing.
Agreed, which is why I prefer fixed costs contracts, whenever possible. Unfortunately, for “bleeding edge technology” programs, that’s not always possible.
There’s a subtlety here which has to do with a mismatch between the center’s personell management goals and the mission-based funding model. The funding is for specific personnel to work on specific aspects of specific projects. They aren’t really supposed to work on other stuff, otherwise known as “cost-sharing”. But the reality is that the center maintains expertise in individuals whom they want to fund long-term (i.e. “keep around”), and who are often sort of passed around between projects as needed. But again, the managerial and funding model aren’t really designed around this idea. The result is that any given project may actually be funding a lot of marginally related work on other projects which is very hard to track. I think the red flag they are raising – a lot of the funding, particularly with such a large budget so early in the game, is probably being used to cover other center funding deficiencies. Hell, I’m not guessing that. I actually know that is what is going on. I’ve seen this game for decades. Everyone knows about it and looks the other way. Since this hacked up system has more or less worked, everyone just sort of accepts that this is how it is.
One starting point in this situation was the “Full Cost Initiative”, maybe the idea of Sean O’Keefe?
https://www.nasa.gov/pdf/19… The idea was all money would be spend on specific “missions” that management could easily track, making all decisions at the top. In reality it’s not easy to have any plans for the future without the opportunity to execute small-scale investigations and tests, and even complete small-scale projects like the ones that are often featured in NASA tech reports may not reach the scale where they would compete for funding in one of the handful of major programs, and of course the funding varies considerably and makes it hard to retain experienced personnel. Unlike private industry there isn’t even a way to retain funding from one year, when there may be more money than needed, to the next, when there may not be enough.
Ideally line organizations could have a reasonable level of discretionary funding to maintain capabilities and develop new proposals, but that doesn’t seem to be considered in the full cost accounting plan, so the system is hacked up and works to some extent, but not as well as it might.
Full cost accounting isn’t as one-sided as you suggest. Without it, NASA centers have an unfair advantage in competing for proposal funding. They can claim artificially low costs because salaries of civil servants and facilities are billed to the center’s operating budget and not to the proposal. And, for missions competing for scares resources like DSN antenna time, not accounting for which mission is consuming more (and making them pay for it) doesn’t encourage prudent or reasonable requests.
But you are correct that having everything billed to projects isn’t a great idea. You do also need additional funding for all the things that aren’t (or shouldn’t be) directly tied to big projects. Direct non-project funding to NASA centers does that, as does overhead charged by NASA contractors. But that isn’t necessarily enough, and large projects do tend to attract extraneous work since they are a convenient way to cover those costs.
Of course there’s an apple-to-oranges issue. In particular, the two primary players here are GSFC and JPL. But JPL is actually an FFRDC, and doesn’t have civil servants. All of the funding is mission based, particularly at IPAC, which lives entirely hand-to-mouth on mission support funds.
JPL is certainly in a fairly unique position, but what I’ve seen is very different from what you describe. JPL does get NASA funding for things like the Deep Space Network, which isn’t tied to or billed to individual missions.
JPL also charges overhead on the NASA projects and missions they are involved in. Actually, quite a bit more overhead that academic institutions would change (although not more than a private company would.)
In addition, JPL is the only place (other than NASA centers and, in one and a half of one case, JHU/APL) to be the lead institution for directed planetary missions. Which means they get overhead off major programs without having to compete for selection.
It might not go to all parts of JPL, but that adds up to a significant and fairly secure source of money for internal research and development, proposal development and the sort of workforce development and retention that was the original topic of this exchange.
I don’t want to get into any more specifics – frankly it would be easy to guess who I am at that point. But your description sure doesn’t match my direct experience at all with JPL and CIT over the course of 30 years, and specifically the part that deals with WFIRST. The overheads were absolutely never available for anything like employee retention. The overheads pay for things like physical infrastructure – keeping the lights on and the plumbing working. I have watched generations of scientists and engineers lost as projects shut down and nothing existed to replace them.
I have to admit, I’m seeing this from about twenty years as a JPL contractor rather than an employee. And one involved in planetary science rather than astrophysics. So I may have a different perspective.
For employee retention, which I guess was the original subject, I think I agree with you. I can’t think of examples of JPL overhead going into that (although, as an external contractor, I may not have been in a position to notice.) I can think of people who had a dry spell in their funding and JPL management finding a way for them to contribute to a well-funded project. I also remember some discussion about how necessary those contributions were.
But in other respects, I think JPL is in an intermediate position when it come to non-project funding. The lab does get non-competed funding for both assets like the DSN and directed missions. They do charge overhead on that. So the lab has a source of overhead funding without the need to write proposals and hope they get selected. That isn’t the case for industry or academic institutions. In that respect, JPL is more like a NASA center.
And JPL definitely uses that overhead. Their Team X and more recent A Team groups are all about taking an idea for a mission and developing it from the back-of-the-envelope level to something solid enough to propose. That work at JPL is better funded (and organized) than similar support at academic institutions. And I assume it’s funded off overhead, since directly billing proposal development work to NASA projects is illegal.
All that makes me believe JPL has better sources of non-project funding than institutions which are outside NASA contractors. But perhaps not as much as a NASA center would have. If JPL management doesn’t spend it on things like workforce retention and related things, that’s their decision.
Maybe the right answer is to use the other 2.4m telescope body/mirror for a separate coronagraph and star shield mission.
That one was even less complete than this one was. It doesn’t have all the support structure. You wouldn’t save anything by using it.
That’s close to one option suggested by the report. They didn’t suggest using the other 2.4m telescope (and pointed out that the WFIRST one wasn’t really “free” given the modifications required.) But they did list dropping the coronagraph in favor of flying it on a separate mission. Another option they listed is making the coronagraph a very clearly defined (and limited) engineering test, with specific plans to use it as a serious science instrument on a future mission.
Or: to keep the folks busy and make them smarter.
At first, I felt that old, anti-government bias boiling up, rezoning that if the work isn’t there just let people go.
Then I realized this is just how I run my own operation.