Faster-Better-Cheaper Anyone?
Lost in space? Nasa under pressure, BBC
“The most recent rover mission to Mars, Curiosity, landed successfully three years ago and has performed admirably. But the mission was around a billion over budget and three years late. These events were monitored by former Nasa scientist Keith Cowing in his blog Nasa Watch. “As upset as Nasa proclaims to be when these overruns happen, they just go off and do another one. It is an ongoing chronic issue with Nasa,” he told BBC News. “Nasa’s financial management system is still a mess. After doing Nasa Watch for 20 years it is almost like I have a key on my keyboard that I press and it says: ‘Nasa doesn’t understand what things cost’.”
So could it be time for Nasa to rethink the “faster, better, cheaper” plan? “Dan Goldin was prophetic,” says Mr Cowing. “But the way his idea was put into practice was flawed and inconsistent and insincere,” he says. “It’s like having the archetypical pictures of the little mammals running around as the dinosaurs are dying. There is always the seed of the next wave of doing things that emerges from the current way of doing things.”
This is easily at least a government-wide observation – the DoD has no idea what things cost since they have multiple, overlapping financial systems. Large corporations also routinely have over-runs, they just don’t have to tell anyone. But the Boeing 474 program is a good example, it almost killed Boeing.
“Boeing 474 program … almost killed Boeing”
Fortunately for them, when an intern suggested switching the numbers around to 747 – following the 7#7 nomenclature, it turned everything around and the company was saved just in time!! (and the intern was then let go for being a know-it-all)
ProfSWhiplash Yes, that intern was the only one to catch the nomenclature problem and it cost him his job! Now, with the “RS-71”, it was a mistake by the President that switched things.
let it be noted that NASA has good and bad days. There are multiple examples of small programs across all the centers performing well with small budgets and having great results.
Lunar Prospector orbiter comes to mind. The entire Morpheus program as a more recent example.
It is the big budget, multibillion dollar programs where this financial incompetence hurts. Curiosity, JWST are big sore examples. These are too big to fail, very risk averse, spending another couple of years and couple billion to get them just right comes with institutional inertia
I remember several significant failures with FBC Program in the past. It may work fine if you were sending dozen’s of probes a year, but they had the same low flight rate, with even lower reliability. If I remember right they had poor telemetry making failure analysis spotty at best. With the caveat that fraud and malfeasance has always been alive and well in America, generally you get what you pay for.
What would be the way to make it cheaper and faster? Just better financial management? More standardized probes like the Mariner Mk.II concept?
What would make it faster and cheaper? Contractor accountability. Contractor will bid on a project, under bid everyone else only to come up with thousand RFI’s that add up to lot more money, and NASA does nothing just hand over more cash because they can’t waste the time to recompete the project (at least three month process). All NASA has to do is apply standard business and legal practices – did you read this work order? did you bid this much? Nothing has changed on the project – deliver at the agreed cost – don’t deliver, your estimator, engineer, and ceo go to jail for fraud.
I think a different risk posture would really help. Currently, for class A missions, “failure is not an option.” Someone can, almost literally, walk in the door and tell a project manager, “I’ve just thought of a potential risk. Since no one can prove the risk is under one in a thousand, you must pay me to study it, and either show that risk is sufficiently low or come up with a way for you to spend even more money to mitigate it.” A manager is not expected to just say, “no, we’ll just take a chance.” Once a project gets big enough, this can become a cottage industry and really drive up cost. Cheaper, faster, better was predicated on the idea of accepting risk and more frequent failures.
The cost of JWST could have bought a nuclear powered submarine and an aircraft carrier. Much of that cost was risk mitigation/aversion as you described.
I can understand how it could be very difficult to accurately predict the cost of a mission before it is even built. But if your estimates, in retrospect, generally display a bias (say towards unrealistic frugality), this is a very easy problem to solve. It is just a matter of normalization. Perhaps a “normalization factor” is not considered an acceptable part of a detailed cost estimate, but it seems like if they were serious about delivering accurate estimates, they would include something like this.
I used to accept this argument- that the tech is advanced, cutting edge, blah blah blah. And I suppose it’s correct to some extent. But it sure doesn’t explain the doubling and tripling of some projects.
This is management pure and simple.
Perhaps I should have been more clear. Whatever the difficulties of estimating costs beforehand, the average cost estimate should be very close to the average final cost, because this is a simple error to address. If the average project comes out twice as expensive as the estimate, you multiply your future estimates by two. So called “normalization” is the easiest element of formulating a prediction, so if they aren’t getting that part right it’s a bad sign.
Years ago, a widely-used cost multiplier in the energy industry was 3.
This is done, and actually required, at the 30 to 40% level. Programs at the initial proposal stage (pre phase A) are required to carry a hefty budget reserve. In effect, normalizing their cost estimate by a factor of 1.4. Unfortunately, many institutions expect and even train managers to spend this reserve down to zero by the time of launch (end of phase D, actually.) So, this normalization becomes a self fulfilling prophecy.
When NASA history continues to repeat itself, despite pronouncements that: ‘we must learn from our past failures so as not to repeat them”, you can rest assured folks have no clue what is going on that gives them the failures in the first place. They think they do, but not really.
My own experience is limited to construction where it is hard enough to think of everything and where occasionally the snake finds me, turning the project upside down.
But here’s where costs are controlled: lots of eyes. After I complete a design project I prepare a cost opinion. Then a contractor does the same. Frequently the client also has in-house eyes. And most of the time there are subs that are watching the entire thing ready to add considerable experience. In the end we are all answerable to the person paying the bills.
So although I don’t miss cost estimates any more very often (after nearly 30 years), I know I am part of a team.
How is Webb, for example, different? As far as I can see that’s the point of BBCs comments. By now NASA should have in place eyes on the various parts and the ability to call bullshit at the appropriate moment.
“By now NASA should have in place eyes on the various parts and the ability to call bullshit at the appropriate moment.”
These projects are more about shoveling the proper amount of pork to the correct Congressional districts. Financial managers embedded in these types of programs have no incentive to call bullsh!t since this will lead to re-assignment to some lesser project, and outside observers rarely have enough insight into the program to know what the lies and games really are.
When you can’t figure out why some ugly government train-wreck project continues to get funded year after year, just follow the money.
That’s interesting. The NASA missions which stay on budget (more or less) have something in common with your personal example. Discovery and New Frontiers projects tend to have distributed management. The PI is in one place (often a university), the project manager in another (often a NASA center) and most of the work is contracted out to an industrial partner. The real budget disasters typically have most of the management under one roof. Ironically, that’s sometimes justified by saying a project is too big or important to trust anyone else to do right. Of course, there are other issues, but I’d be amazed if a problem of this magnitude had a signal cause.
I have a hunch that the budgetarily successful projects under that system are not developing new technologies for the project — or at least not many of them.
Those cost-capped missions are designed to be high in technology heritage, taking full advantage of developement costs that are sunk already. In simpler terms, the spacecraft has a lot more off-the-shelf in it.
I don’t that’s the case. Discovery has included the first sample returns from beyond Earth orbit (note that the moon does orbit the Earth), the first use of electric propulsion on a scientific mission, the first Mercury orbiter (with thermal requirements which really drove the hardware requirements), etc. Juno is the first mission to Jupiter using solar power, and that wasn’t exactly off the shelf solar arrays. Technology development and doing new things is a common explanation for the high cost of flagship missions. But the more I think about it, the less convinced I become.
I think the relevant question is how much new technology those missions required and how much it cost to make it work. Off-the-shelf may be too loose a term, but I believe those missions benefitted from technologies that were straightforward advances on know-how already in hand.
Luck, too, plays a role. For example, MSL was going to use a dry-lubrication system on its “joints.” Well, after spending a good bit of money, that didn’t pan out. The EDL system, especially the skycrane, needed more work and testing than originally planned. And so on.
By the nature of the cost-capped programs they’re proposed under, Discovery- and New Frontiers-class missions are not meant to involve substantial new technology expenses and testing. Everybody knows if the overruns get too big — CHOP!
Flagship missions, such as MSL, are approached differently.
Irrelevant aside:
I would call it the Earth-Moon orbit. Our combined center of mass (barycenter) is what circles the Sun (its about a third of the way up from the center of the Earth..
https://www.youtube.com/wat…
Pluto and Charon have a barycenter in the space between them as can be seen in this New Horizons video:
https://upload.wikimedia.or…
When NASA missions are proposed for funding, the Technology Readiness Level (TRL) of significant elements must be stated. In some cases, program managers will demand that certain key items be reported at a higher TRL than they actually are, hoping that they can make up the difference in the time to launch.
JPL (NASA) bit off more than it could chew with the Curiosity rover (MSL). There were many new, untested technologies on that craft, so it’s not so much that MSL overran the budget, but that its budget was overly optimistic.
That’s the real NASA problem. Everyone approaches these things in hopeful mode. And because this problem is not confined to just NASA, but includes aerospace in general, it’s my conviction that this is characteristic of all organizations doing tech development on the public nickel.
I see no way of doing it differently, so long as public money is involved.
That’s an excellent phrase: “hopeful mode.” Having worked on these types of projects, I, too, have seen this again and again to the point that no one ever believes delivery dates.
I’m not sure that necessarily applies only to public money, though:
http://www.universetoday.co…
(Not picking on Musk, just that he seems to be as hopeful as the rest of us).
hey, I forgot about FH. Did a quick search to find this:
“SpaceX has slowed development of the Falcon Heavy rocket, a mega-booster made of three Falcon 9 booster cores strapped together, as the company recovers from a launch failure last month, delaying the new rocket’s inaugural flight until early 2016.”
http://spaceflightnow.com/2…
Regarding arguments of SaturnV, SLS, N1, etc. will FH be sustainable for many flights? Multiple flight rates per month? (actual numbers vs. predictions). I wonder if Moon will be a target of missions (i.e. everyone except Wingo and Spudis loves to talk about Mars).
I find predictions of flight rates similar to when mass transit folks talk about the number of riders a new system will have. Both end up being way inflated.
The thing about FH is that it is only somewhat a heavy lifter. If full re-usability is demonstrated FH is designed to be the standard workhorse for heavier payloads.
Right now the F9 (even the enhanced F9) is pushing the limits for heavy payloads to GTO with re-usability.
Yes, the FH can loft massive Bigelow space stations, but it also can cheaply launch comm sats (if re-usable).
On the SpX website is the fixed price: “$90M for up to 6,400 kg to GTO”
An important difference is the fact that the Falcon Heavy will have quite a bit in common with the Falcon 9. Saturn did and SLS will need high fixed costs just to keep the production line open. Without high flight rates, that becomes very inefficient. With most of FH coming off the same line as F9, a low flight rate would still be viable.
It’s worth noting that it looks like, because of its lifting capability, Delta-IVH will outlive its single-core equivalent. Add on top of that Falcon-9’s crew launch variant, I think that FH and F9 both have a good many years ahead of them, assuming that SpaceX can avoid frequent failures.
“Since when has optimism been a crime?”
“I’m going to ding you for pessimism!”
Those are both quotes from Dilbert but they well capture the essence of this faith-based project management mindset. I have no problem with encouraging your staff to believe in the project and want to push for success but when that starts being at the expense of reality then you’ve taken it too far.
“The basis of optimism is sheer terror.”
– Oscar Wilde
JWST management at Goddard (Maryland) overshot their budget in such an egregious manner because they could. Senator Mikulski (Maryland) was willing to use her considerable political clout to keep the project alive and funded, regardless of the constant parade of technical mis-steps and fumbles and re-directions. This was an open secret on the program. Annual exercises to come up with realistic cost projections were considered a time-wasting joke.
Now, with Mikulski on her way out, such a big-ticket fiasco is very unlikely to be repeated at Goddard and survive Congressional scrutiny.
This is incredibly misleading:
“Instead, Nasa seems to have lurched from bargain basement space missions back to gargantuan projects where costs have spun out of control – with the exception of the low-cost Discovery class and medium-cost New Frontiers class missions (of which New Horizons is one).”
Huh? Except for the handful of Mars missions (and, really, only one in particular that was a huge overrun), NASA’s planetary program is dominated by Discovery and New Frontiers missions. They are not the exception. Not only that, the Discovery missions exemplify FBC.
All I see in the article is Curiosity this and JWST that. Those are two missions (one of which is not even in the planetary program) out of many successful ones.
As for ESA, how’s that Bepi-Colombo mission coming again? The one that is now at $1.6 billion which is 50% higher than originally estimated?
It comes down to one word: “incentive.”
What incentive do NASA managers have to tell Congress the truth about how much a project is going to cost?
It becomes a clash between actually wanting the project to succeed against a fear that, if they tell the decision-makers the truth, either the project will be cancelled or they will be fired in favour of someone dumber and less ethical who will tell the lies that their superiors want to hear.
In most R&D agencies (NIH, DARPA, NIST, parts of DOE and FAA, etc) Congress does not micromanage to anywhere near the extent they do with NASA. It doesn’t make much difference what NASA tells Congress because even with the best information Congress is unlikely to make a wise decision. It’s more important for NASA to improve its internal decisionmaking. If NASA has a success (i.e. commercial cargo and crew) Congress will be glad to take credit for it even if they fought against it.
When it is about pride and honor one can’t be cheap. NASA is the pride and honor of this country.
John Adley Maybe it was, back in the 1960s.
That’s certainly true- that is, it’s true amongst those not more deeply aware of NASAs struggles and of the crazy quilt of space policy.
More informed people know differently.
You know. Like all of us smart guys here 🙂
Generally in the private aerospace industry there is one manager with responsibility for both technical and financial aspects of a program. like Kelly Johnson, Ed Hienemann, or Elon Musk. If there is a problem, the manager can’t blame someone else. The basic concept of BFC was good, control and reduce cost, and SpaceX has done the same. Take away that goal, and you can pretty well guarantee you will not achieve it.
Unfortunately, the days of Johnson and Heinemann are long gone. Look at the F35 or any other big acquisition. Industry managers of large programs don’t get those slots by rocking the boat. They’re groomed and promoted into them with each program a step to the next rung on the corporate ladder. A failing program can still generate profit. A failed program can make money through termination costs. There is no reward for careful stewardship and plenty of penalties for plain speaking.
Musk is different but following the recent launch failure and short sheeting of commercial crew it remains to be seen what he and SpaceX will be able to do.