Report Says NASA Should Do Big Things And Just Send Them Money
NASA Should Continue its Large Strategic Missions to Maintain United States’ Global Leadership in Space
“NASA’s large strategic missions like the Hubble Space Telescope, the Curiosity rover on Mars, and the Terra Earth observation satellite are essential to maintaining the United States’ global leadership in space exploration and should continue to be a primary component of a balanced space science program that includes large, medium, and smaller missions, says a new report by the National Academies of Sciences, Engineering, and Medicine. However, controlling the costs of these large missions remains vital in order to preserve the overall stability of the program, the report finds.
In the past, concerns have been raised about NASA’s large missions as there has been a history of their costs exceeding original estimates, impacting the overall budget of the agency. The James Webb Space Telescope (JWST) is a recent example of a large mission that experienced substantial cost growth. Big jumps in costs for a mission can have an impact on the entire science program at NASA.
Although cost-evaluation and cost-management mechanisms developed at NASA over the past decade have proved to be effective, NASA should continue to use its various cost estimation and management tools to better assess and control the costs and risks of the missions and ensure they remain a viable option, the report says. The agency should also support the development of new estimation tools to perform robust cost estimates and risk assessment for future missions.”
Here is a thought. Instead of just throwing out and trying to sell a proverbial “box” of science, a magical thing (science), that most people don’t understand but “believe” that it is something good for them and a benefit to society, how about we actually come up with a long term, goal directed plan with short term attainable goals which, quantitatively, lead directly to the larger goal.
NASA provides that on a regular basis. Lots of planning. lots of lots of planning. so much planning it makes your head spin. But it never gets implemented or disseminated
Because the way they run their large manned space programs, they spend 10x the money that they should, straining the budget and dragging out the schedule. One need only look at the Ares I/V program and its follow-on the SLS program to see that NASA can spend tens of billions of dollars on such a program without a single flight (excepting of course Ares I-X which was a space shuttle SRB topped with a dummy SRB segment, a dummy 2nd stage, a dummy capsule, and etc.).
So, lets give NASA NO money and everything will be better?! I’m betting this crowd probably thinks the result would be the best space program ever!!
Because of opposite of wasteful is retarded?
How about just not wasteful? Can we try that? Just for a few years. We’ve had half a century of wasteful and often stupid, can we try not wasteful for just a little bit, see how it goes? Maybe we’ll like it.
Here’s my thoughts on the matter. Very large “winner take all” programs end up being run as “cost plus” contracts in the US with zero incentive to reduce costs. Competition has been proven to work with “commercial cargo”. If Dreamchaser starts cargo missions to ISS, we’ll have three different providers to choose from. That’s a huge motivator for contractors to increase efficiencies and reduce costs.
So, why not replace SLS with “commercial HLV”? Have ULA, Orbital ATK, SpaceX, and Blue Origin compete against each other during the initial proposal/development phase then down-select two HLV providers that are awarded a fixed length contract. When that contract is nearing completion, open up the competition again and repeat. This fosters not only competition, but innovation as well.
I don’t think a commercial HLV would work, except in Mr. Musk’s Mars colonization dreams. Once you get above the size of a Delta 4 Heavy or a Falcon Heavy, there just aren’t enough payloads. To really make a commercial (and competitive) launch vehicle environment, I think you need multiple launches per year. At the moment, there aren’t enough heavy lift customers out there.
I would say D4H, F9H, and New Glenn (Blue Origin) are HLVs. Anything larger just isn’t efficient. It’s not just a matter of cost. Without an adequate number of launches it isn’t possible to achieve an adequate reliability.
SLS is too expensive and flies to infrequently to support a manned Mars program (or likely even a manned lunar program). Cheaper, more frequent, access to HLV launches might go a long way to solving that chicken and egg problem.
A few other examples of NASA and Science program failures which had a “percieved” usage plan: X-38, Aerospike engine and the Superconducting Super Collider, and there are many more (inside and outside the space/science communities) that started and ended without being concluded and used up 100s of millions of dollars. All because the “plan” and long term question of use was not sufficient or deemed unnecessary. When are we going to start succinctly answer the question “WHY” in a sustainable and pragmatic manner.
OK. I think I see your point.
But how would we know anything about the aerospace [edit: aerospike], using our example, unless we started to build one? How do we know that a given approach is not going to be fruitful?
I know that question isn’t entirely applicable. Certainly the smart guys amongst us fashion experimental work with some chance of success- success being measured, not by the experiment in isolation, but by how far forward the field is advanced.
Yes, money is wasted. SLS has been mercilessly beat up here, and elsewhere, as a project with a knowable goal that is consuming laughable amounts of money. But the SLS program has one big room-sized elephant: SLS has comparable, namely Apollo, and F9. I wonder though how SLS would fare in an environment with no comparables? And isn’t that exactly how these flagship projects are assessed? Against a list of desired scientific returns, and a wild-ass guess on the cost of the hardware? and with nothing to compare?
(With time, of course, the comparables start to pile up).
However many of the goals of the SSC were achieved by CERN at lower cost with the Large Hadron Collider, and competition between Europe and US was avoided by very extensive collaboration. A lot of thought went into minimizing cost in the construction and operation of the LHC, even the efficient use of an existing accelerator tunnel. Cost increased from $2.9B in 1996 to $4.5B in 2007, an increase of about 54% over more than a decade, vs an increase from $4.4B to an estimated $11B for the SSC, an increase of 250%. LHC is now at 13 TeV, vs a planned 20 TeV for SSC, but LHC has considerably higher luminosity.
Cost growth was far greater in the Shuttle program, in part because of the lack of any flight experience with prototypes prior to selection of the production design, but the exact reasons remain obscure and more a matter of opinion than analysis, in part there was little attempt to predict or understand the reasons for the increasing cost.
I disagree with your statement that “Cost growth was far greater in the Shuttle program”. The management of that program were well aware of the technology challenges, particularly with the SSMEs and the TPS. Costs were estimated with a reserve specifically identified to take care of those technology challenges. The reserve was not provided, and that was the only cost growth. Given the advanced nature of the vehicle for the time, it was done on the planned budget. Even the schedule delay, about 2 years, was very reasonable given the challenges that were identified from the outset.
Planning, no matter how much, without sufficiently and explicitly answering the question of “WHY” for any project/program leaves them open to misdirection and waste as well as being unsustainable. Historically the word/concept of “science” has been used as a misguided and magical (therefore unquestioned as to its worth) proxy for answering the question “WHY” and this needs to change.
Perhaps I’m missing the thrust of your argument, or even the tone— but isn’t it the case that the “Why” is simply to learn stuff?
There’s a school of thought that science ought to be goal-oriented (likely the same noobs who think learning how to factor a quadratic is a waste of time).
But science by definition shouldn’t be goal oriented. Sure, we sent Cassini all decked out with all sorts of cool sensors and instruments, ostensibly to precisely measure X, or Y, or Z; but the real reason was simply to learn more about Saturn. Similarly the search for a “cure” for cancer—at least one kind, as a start. But this involves basic research into cell structure, into energy transfer in cells, into how cells differentiate, into a whole litany of research projects that might or might not move the ball forward when measured against some sort of “goal”.
I could be over-reacting. But I’ve gotten very sensitive to the whole dumbing down that we see around us.
In a way, NASA science is goal oriented, and many scientists who propose missions are instruments aren’t completely honest.
Proposal selections are very heavily goal-oriented (here is an important scientific question, here are the measurements required to answer that question, here is a proposal for a spacecraft and instruments which will make those measurements and answer that important question.) Reviews of proposals often use words like “closure” (is selected, will the question be “closed”, i.e. will we be done) and, informally, anything like just looking to see what is out there is criticized as a “fishing expedition.”
Of course, most scientists do recognize the value of just looking to see what’s out there. They write proposals with all the formal flow-down from specific questions to measurement requirements, which NASA expects and requires. But, at the same time, they also make sure the instruments and measurements can do quite a bit of “just looking.” That isn’t completely honest, but it’s one reason why many missions make so many serendipitous discoveries.
Would it be correct, then, that the criteria and the “managers” aren’t really assessing the proposed mission?
I’m not sure I’d say that. It’s more like the scientists are being a little tricky, and designing instruments which can do X as well as Y. If Y is what the criteria and the managers call for, and adding the ability to do X is a marginal cost increase (or possibly improving the mission’s ability to do Y), then they are getting what the payed for. And the scientists have built in an opportunity for serendipitous discoveries.
At least on the planetary science side of things, the “WHY” is explained, at length, and everything (in theory) flows down from four or five straight-forward questions or goals, all the way down the details of the instruments on planetary spacecraft.
Unfortunately, all of that is in documents written by committees, and which are so mind-numbingly dull that they get virtually no press coverage and are almost completely unknown to, at a guess, 99% of the people in the country.
A lot of management is not simply managers managing, and in the case of science missions is not managing science. You really have to know what it is you are trying to manage. Frequently you have to either have experience to know what to look for or what to look out for.
ISS provides some great examples. Remember, Dan Goldin said he wanted NO experienced managers in the program. They filled the program with some zeros. Some are now ‘managing’.
There was an ISS manager, an astronaut, who felt that all payloads needed to be multi-fault tolerant. He did not want astronauts on ISS missions dealing with failed science hardware, or even worse, he did not want the astronaut to be bored, with no science to do. Its a nice thought, except this increases significantly the cost to the experiment developer. Most experiment developers have almost no resources to start with, particularly on ISS. So this astronaut almost single handedly ran off many of the payloads and experiments by trying to institute requirements that were too costly. BTW, simultaneously the ISS Program refused to pay for testing of critical life support hardware. That hardware broke down almost as soon as astronauts started using it, living on ISS. Which forced the program to consider abandoning the station during its first expedition.
Another example, the ISS manager who was responsible for setting up experiment integration had no experience in experiment integration, and she tried to set it up so that the experimenters would need to deal with each of the organizations in the ISS Program independently. She wrote an elaborate book trying to define the functions, responsibilities and products of each ISS organization. The ISS is composed of a series of organizations all of which do various aspects of integration and of operations. There are often overlaps and gaps. Often the organizations fight over which has or does not have different responsibilities. The book did not make much sense, because the organizations did not make much sense and besides that the individual was not a competent communicator or author and the text was in gibberish rather than English. 15 years later the same individual was put in charge of ‘fixing’ the payload integration process for ISS when they figured out that they were getting few payloads to fly on ISS. She started out a meeting of people interested in this fix by asking “who knew that PIs or experiment developers were trying to get their payloads integrated on a schedule, like in less than 3 years?”
Anyone who had worked payload integration on any previous program knew this; apparently it was only the managers in ISS who had no clue.
Sometimes it helps to know who you are working with, what the goals and constraints of the respective organizations or individuals are.
The same managers were responsible within ISS for taking the money intended to generate and integrate science, in the hope or expectation that someone else would pay for science. So they were simultaneously responsible for eliminating the resources to develop the science while also imposing more difficult requirements for integrating science.
Managers managing? Sometimes they ought to put people into positions who actually have some idea of what they need to be doing.
Its now 20 years later. 2/3 of the lifespan of ISS has been used. Science is often not proceeding at all thanks to the efforts of “managers” like these.
Good job NASA! [NOT]
Planning. and managing! Lots of managers, all managing!
Not sure why the snarky title but its true – to do big science takes big budgets – and these unique, highly sophisticated scientific programs like Hubble, Curiosity and Cassini have been incredibly impactful to science and to the US’ role in the world. The priorities are also carefully vetted through the decadal science plans and while some programs (e.g. JWST) had management challenges that led to cost over runs, the scientific benefits cannot be achieved by commercial companies.
Yes, but as great as JWST may end up being, It should have been considered for cancellation if proper rules were followed when the costs ballooned. If there is never any consequences then it emboldens the behavior. JWST got to a point of no return. This would all have been fine if the cost estimate had been accurate in the beginning and all parties had agreed that it was worth 9 Billion vs 4X2 Billion + astrophysics missions. Just one persons view. HSF should be held to a similar standard but I guess there it is harder because they keep changing the plan so when did/does the clock start?
“If there is never any consequences then it emboldens the behavior”
This assumes, wrongly, that accurate costing for these one-off missions is possible, and that, deep into a project, anything can be done about it.
I’m not sure who is to be “emboldened”; perhaps a deeper understanding of the process wold be helpful.
True but low balling missions over and over and over just to get them started in the hopes they will not be canceled once the real prices come out has been a problem and is well documented by insiders that have did legacy projects.
And that is part of the problem. You get ONE SHOT at a mission in your entire life. Are you going to low ball the costs to get it started if this will be your one and only mission?
Actually, the “ONE SHOT” business has other effects than lowballing the estimated cost. If it’s the only opportunity in your lifetime, you’d hate to have it fail because of a hardware problem. You’d hate to waste that opportunity by flying fewer or less capable instruments than you might. And operationally, you get to the point where a couple dozen scientists will spend half an hour arguing about 25 seconds of observing time (I’m not joking, I was on the teleconference and timed it.) All that drives up cost.
In terms of cost, I could argue quire convincingly that accepting higher risk is a good idea. Consider the Discovery program. The demonstrated success rate is 9 out of 11, with one partial success thrown in. They are developed as Class A/B missions, when it comes to reliability and risk posture. 9.5 out of 11 is close to the 90% success rate reported in a National Academy study for Class A/B missions in general. Missions developed according to Class C/D standards have a 80% success rate, according to that same study. But let’s be extremely pessimistic and say planetary missions developed to Class C/D standards would only have a 60% success rate. It’s generally agreed that the difference between Class A/B and C/D is about a factor of two or three in cost (although I would really like more real work and analysis to firm that number up.)
So, let’s do the math. The Discovery program is at 9.5 successes for 11 attempts. If had been done with more risk-accepting standards, the same funding could have supported over 22 missions. Perhaps 40% of them would have failed. But that’s still 13 successes, and 13 is greater than 9.5.
Unfortunately, that logic is all about the _average_ and the success of the entire program. When every mission is a once in a lifetime opportunity for someone (or lots of people), the averages and the program don’t motivate people too strongly. It’s not about taking a risk to free up money for some other, unrelated Discovery mission involving other people. It’s about _your_ mission and _your_ team. And that makes projects risk averse.
Yes.
Voyager started as the planetary grand tour that was supposed to fly by Jupiter, Saturn, Uranus and Neptune. It was identified as an expensive mission. Nixon said he would not support it. So NASA went back and suggested a lower cost mission to just Jupiter and Saturn. NASA always had in mind that they could do the grand tour with the lower cost program, and they did.
That’s a serious charge, Vlad. But it’s a view that appears to be the default position of those who so deeply distrust government.
From time to tim I am asked for a Proposal that is based on an RFP. Sometimes, there is a ‘hole’ in the proposed methodology that will affect the price.
What to do?
Often in the past I’d include the needed work in my proposal, which would inflate the bottom line, and I’d not get the project because 1) Price or 2) Failure to follow instructions.
So, often and in the absence of clarification, I’d leave it out, but include a small piece of text about additional work. meaning additional money.
Other times, I’m invited to write a Proposal not based on an RFP but by my own assessment of what’s needed to do the project, having full control over the Scope of Services. I these cases, knowing WHAT to do is part of the work.
And from time to time, despite >35 years’ doing this work, there will be unforeseen elements that should have been included in my Proposal; I’m the “expert”, after all.
Now these unforeseens can be two types: stuff that I should have known, and are therefore usually responsible for including in the final work; or other unforeseens that I couldn’t possible have known about without the benefit of having started the project. An example might be additional public hearings triggered by some outside event.
How much of this is applicable to the world of high-tech and high-stakes Big Science? Quite a lot I imagine, with the additional monkey wrench being the unknown difficulties of one-off instruments, plus issues coordinating all of the instruments, propulsion, communication and others stuff on a unique platform, plus coordination across the planet, and naturally all of the uncertainties that rockets bring to the table.
So, yea, I can see how even the very smart people can screw up the budget estimates early in a project; and based on my own experience, I’d rather not assign disingenuous motivation or mendacious behavior without direct evidence.
As far as I can tell with my own (very) limited observation, those science types who are proposing and designing Flagship Class missions- they are the guys in the white hats.
Well as people that have worked at NASA and have posted those “charges” I guess we have to take it will a grain of salt.
There are a few good points about the NASA selection process, at least in these respects. For something big, like a mission, a draft version of the AO usually comes out and is open for comments. So if it left out something important, people have a chance to point this out and have it corrected (which keeps competitors from ignoring it and giving an unrealistic cost estimate.)
Second, they try not to specify measurement techniques or instruments in a mission AO, just science goals questions to be solved. So, if someone has a clever way to get the same results for a lower price, that’s fine. It can be dangerous, since a reviewer could say he wasn’t convinced the new idea would work. It can also cause some hard feelings. If people want a mission which will do W, X, Y and Z, they may work very hard to get a decadal survey to say doing Y and Z are extremely important. They may count on the fact that, although they can’t convince the whole community about W and X, the most obvious way to learn Y and Z will get them W and X as well. Then some smart guy comes along and finds an easier way to do Y and Z all by themselves. The Juno microwave radiometers are an example of this; that are a much easier way to get volatile composition at sizable depths in Jupiter’s atmosphere, than the physical entry probes descending on parachutes (which is what the decadal survey mentioned.)
Finally, although they aren’t supposed to, reviewers do compare proposals, or at least the contents of one will influence their review of another. So if a competing proposal is likely to ignore something and give an unrealistic cost estimate, you can make a point in _your_ proposal of how much including that something is. That may make the reviewers noticing that something is missing from those competing proposals.
another great example of the fragility involved when extrapolating personal experience.
I’m extremely cynical about cost modeling of scientific spacecraft and missions. But NASA does use these models to make decisions, and sometimes very bad decisions. The models are taken seriously and used as input into decision making. They build in a 30% per mission cost inflation (which isn’t intentional and most people don’t realize) and penalize novel, cost-saving ideas (the models extrapolate from past missions, so a new idea for saving money gets flagged as a proposal with an unrealistic budget.)
Well, how else? The question I (crudely) asked above remains: how else except by learning from past experiences? And if not the models – which I knew nothing about but figure are quite detailed – if not models, then what?
Wouldn’t continual improvement in the models eventually yield something more accurate?
I can’t see any realistic alternatives to estimating cost based on past experience. It’s just that the current situation, of largely one-off missions, makes this extremely difficult to do accurately, and it’s prone to a huge amount of subjective biases, if not outright manipulation.
Since the missions (at least scientific ones) are largely unique, how do you decide which past missions are or are not similar? One common standard is the number of instruments on the spacecraft. But even a simple count is a matter of opinion. To use Cassini as an example, the spacecraft officially has 12 instruments. But in reality, most of those instruments contain multiple sensors. If one of those “instrument” hadn’t been provided by the same team, it might easily have been considered two or three instruments. I think I could reasonably say Cassini has 23 instruments, although this could easily be debated. The CAPS instrument consisted of three sensors, and the same functionality on the MAVEN spacecraft is officially three instruments (SWEA, SWIA and STATIC.)
So, as the input to a cost model, what number of instruments do you use? If I want to, I can say my mission concept will be very cheap, because I’m claiming it only has one instrument. Admittedly, that one instrument is actually a suite of a dozen different sensors. If someone wants to claim my concept’s budget is unrealistic, he can take the same cost model and type in “12” instead of “1” and get a very different budget estimate.
In addition, the use of past mission costs has some unintended consequences. Innovative new ideas to reduce cost and improvements in the state of the art aren’t considered. They can’t be, since the past missions in the data base didn’t have them.
The practice of including and always spending a 30% budget reserve introduces a huge, artificial inflation in mission costs. If someone develops a Discovery mission, and the models say it will cost $340 million, they are expected to tack on an extra 30%, to deal with unexpected problems. They also end up spending all of that reserve. You can alway chase down potential problems, and many institutions actually train their managers to spend reserves down to zero by the time of launch. So the $340 million mission ends up costing $450 million. That’s ok, since the Discovery cost cap is $450 million. But the next time some proposes a similar mission, the data base and the cost models say the last mission cost $450 million, so proposing the new one for $340 million is unrealistic. Experience shows it should be $450 million, and with a 30% reserve, that’s $585 million.
Some, if not all, of that isn’t an inherent problem with the models. It’s a problem with how they are used. That’s really what I’m cynical about.
Modeling cost of a unique planetary mission is never easy as much of the cost is in development. On the human side spacecraft and launch vehicles remain the same for years, but there is little attempt to accurately model cost.
Not sure I have your point: are you saying that modeling the space craft design process isn’t practiced? Or the cost of the booster (this can’t be right)?
I think SLS or Orion may be an example. I’m not involved, and I’ve only seen the media reports. But I keep seeing GAO or other oversight organizations, reporting how the cost of those projects isn’t really based on any rigorous analysis. Or that it doesn’t include everything.
I think the cost modeling for unmanned missions isn’t reliable (or, to be polite, far more precise than accurate.) But it’s at least an attempt to come up with a realistic number based on a standard methodology. If the SLS or Orion budgets are as dubious as the reports imply, I can’t see how those projects could have gotten through the sort of TMCO (Technical, Management, Cost and Other) review required for most unmanned missions.
“to be polite, far more precise than accurate”
Reminds me of a lecture I heard (back in high school!) about sigfigs…
When the clock starts ticking is also a problem for unmanned missions. In the case of MSL/Curiosity, I’ve gotten into arguments over how badly they overspent. I’d call it 450%, since the original JPL number, which they used to sell it to the decadal survey was $600 million. Others in the field insist that it’s much less, and the baseline should be the estimated cost at the start of phase B development (i.e. at the end of a reasonably rigorous concept study.)
Despite being a great fan (and employee) of the Cassini project, I think the last planetary decadal survey go it right. There needs to be a balance between large and small missions. Realistically, Cassini cost about $5 billion. (Don’t pull the $2.6 billion number off the web. That’s from an estimate at the time of orbital insertion in 2004, and doesn’t include inflation, the European contributions, extended mission operations costs and a bunch of other things. $5 billion is my best guess at the inflation-adjusted FY17 number.) There are very definitely things Cassini has done that a smaller mission could never have done. But $5 billion is also the cost of the entire Discovery program (again, inflation adjusted.) There are some things which require a big mission and a big budget, but there are also lots of things that are best done with many, small, inexpensive missions.
Well, Dr. C, if you are in an expansive mood…what do you think SX could have done if assigned the science goals of Cassini?
An unfair question to be sure. But a question many will ask. They will ask it in many different contexts as the entire PTCBN costs are finally and widely compared to F9 and FH. Many will extrapolate the costs of these programs and wonder if there are other, similar disparities to be found.
Folks will recognize that there is an incremental cost assigned to NASA, a cost that covers management, perhaps
over” management; it covers failure avoidance, and the cost of contract management issues as well.
My own curiosity would be about the ratio between cost and return. If the $5B could be reduced with no reduction in return, what would the number be?
And how much for a Neptune or Uranus orbiter?
That isn’t an unfair question, and in a way, I’m actually paid to think about it. The Cassini project does have some funding after the spacecraft is gone, and much of that is to compile lists of lessons learned to benefit future missions. But I’d rephrase the question a little.
First, I’ll assume $5 billion is available, and think it terms of maximizing the scientific return (not doing the same thing for less.) Second, SpaceX isn’t a scientific research company, and Mr. Musk has a thing about Mars, not Saturn. So I’ll assume you mean a company with a similar philosophy and backed by a billionaire with Saturn fixation. Finally, I’ll skip most of the real lessons learned from Cassini: There are lots of improvements we could make, some from improvements in the state of the art (Cassini had some awkward work-arounds for things that couldn’t be done efficiently with 1990s technology), some from fairly low-level technical details, and some even from things like the sociology of how scientists, engineers and instrument teams interact with each other. I think you’re asking about more of a high-level, philosophical way of doing things differently.
So I’d go for a mission with the same capabilities as Cassini, but a shorter design life (or a higher risk of failing before the end of the prime mission.) By building a spacecraft that would definitely work for seven years of cruise and a four year prime mission, they ended up with one that could operate for an additional nine years of extended mission (and actually, could have kept going for a few more, if it weren’t out of fuel.) Were a similar mission designed to definitely last for one year at Saturn, and a fifty-fifty chance of lasting four years, I think it would cost half as much and you could probably shave off enough mass for a shorter cruise to Saturn.
What would that accomplish? Well, it would give you most of Cassini’s prime mission discoveries. Cassini’s extended mission was largely proposed as, in effect, a whole new mission to follow up on the prime mission’s discoveries. Since I said I’d assume the same total budget was available, and that the funding was from someone with a Saturn fixation, I’d use the money saved by intentionally building a shorter-life spacecraft to actually fly a follow up mission.
In reality, Cassini has done a great job of following up on its own discoveries. But I don’t think there is a single scientist involved who doesn’t wish he could change something about the instruments. I don’t mean new technology (although that is a major advantage of a new, follow on mission), just capabilities we didn’t build in because we never knew we’d want them.
The Huygens II probe could be targeted to land in one of the lakes or seas on Titan (now that we know where they are.) One of the instruments I’m involved with, the ion and neutral mass spectrometer, could be replaced with one could measure negatively charged ions and ions with a mass over 100 AMU. We never expected to see any of that, but another instrument (CAPS, the plasma analyzer) happened to be able to detect both negative ions and ions going up to thousands of AMU in mass (arguably not single, large molecules, but very small aerosols or condensates.) But those measurements are pretty limited, and real mass spectra would be real nice. I think there are similar examples for every instrument on Cassini.
So that’s what I’d do, given the choice. Fly a less reliable and shorter-lived mission, and use the savings to fly a second mission with instruments tailored to follow up on the first mission’s discoveries. I thin that’s also consistent with the SpaceX or “new space” philosophy of not trying to be perfect from the start, but having an evolving system that incorporates the experience you gain along the way.
Reading your response (and thank you) I was thinking of LISA and LISA Pathfinder. Precursors make a lot of sense. Social surveys, for instance, often do quick, initial studies that could establish some sort of upper and lower bound on the expected results, or ask open-ended questions that inform further study.
I also wondered, as you were building the case for balancing longevity with reliability, expecting to reap a financial reward: how does one build a “less reliable space craft”? What features would be compromised?
Maybe you are thinking of, say, the MILSPEC for nuts and bolts holding the thing together; perhaps these march to a different and less demanding spec. Or batteries not quite as “hardened”? But no, these examples don’t really make sense when failure is binary.
Some things are gradual. For example, memory (volatile or nonvolatile) and its degradation with radiation exposure. Permanent damage tends to take out a single memory location or a block of memory at a time. Flight software can easily be reprogrammed to avoid the damaged locations (actually, you just update a don’t-use table.) So I can say, “I’ll use the cheap parts, a third of them will fry before orbital insertion plus one years, and I’ll be down to only one third working by insertion plus four years.” That differs from the common practice of using much higher radiation hardened parts and assuring they will last four years (and, thereby, also assuring most of them have a fair chance of lasting much longer.)
But for truly critical systems, it is all or nothing, and everything is done on a statistical basis. If the requirement is to make sure the spacecraft survives to the end of its prime mission, they really mean that the risk of it dying is a small number (e.g. 0.5%) and then doing all the testing, built-in redundancy, certification, reviews and paperwork necessary to achieve that.
All (or much) of that scales with the design lifetime. I may have a part which is definitely (and previously proven) to survive the radiation environment for one year, but would have to be retested and prequalified if I need it to survive for four years. Or I may need to open and close some valves in the attitude control thruster system about once a week. Do they have to be build, tested and qualified for 50 cycles or 200? Sometimes the savings is just in the reduced testing and certification (which is significant but not a factor of two), but if it’s the difference between using a pre-existing part or design, versus custom parts or new designs, that can be a huge savings.
For unmanned missions, I think a reasonable allocation of funding is met by a mission and a quantity of science that do make good use of the funds. Then based on performance over the last 50 years, the unmanned programs always seem to deliver more than what they promise.
In human space flight it seems to be the opposite. They always want a lot more money than they can reasonably expect. When they are allocated a nominal budget, they always seem to deliver less, at a slower pace. Nominal budgets are pretty well known. NASA was told in 1966-67 and again in 1972 about what to expect (less than 1/2 of 1% of the Fed). As the prominent NASA programs have delivered less than expected, the budget has declined some but not significantly. It definitely has not gone up except in a few instances. And as far as NASA absolutely needing much more, Space-X and even Boeing’s commercial effort have shown much can be done with a lot less.
I think NASA’s human space program has now been pretty thoroughly discredited. Shuttle was terminated without ever having been improved upon significantly. ISS is in a holding mode. Construction was finished long ago. Few significant improvements are even discussed. That world-breaking science they talked about for decades is not happening-in fact very little science even going on anymore except for some micro-sats and of course some life sciences to prove people can fly the missions they are flying-its a non-sequtor, but so what? Science had been part of the ISS budget until about 2001-just as it had been a part of the unmanned budget-that was the reason for being- but in ISS that money was taken long ago and continuously ever since-and no one outside of NASA stepped forward to pay for science which was NASA human spaceflight’s fervent hope-and besides NASA made it so difficult to fly science that they ran off any prospective experimenters. Orion everyone wonders what its for. It is not for any science that hasn’t been done 50 years ago. And the rate at which Orion has come on line, and its cost, as compared with competitors like Space-X, Boeing, Dream Chaser, make everyone wonder why NASA gets to spend so exorbitantly with so little to show for it. But Congressmen still like the funding in their districts, so it continues. Space aficianados looking for great space breakthroughs will simply have to wait for the commercial suppliers to do more spectacular missions-which they are working towards at little real taxpayer expense.
Somebody is funding science on the ISS. Here is the list of experiments going on aboard the ISS:
https://www.nasa.gov/missio…
A lot of these are not science. Mr. Bigelow’s BEAM for instance is an attempt to sell goods to the government or to others. Its the latest inflatable design in a series of iknflatables projects that go back more than 50 years. ISS has been ongoing in space for nearly 20 years. Which of these are current and which are from long ago?
Overall I do agree with you, but the part where you talk about how human space flight under performs after being given less than they asked for reminds me of a Dilbert Cartoon: http://dilbert.com/strip/19…
A snarky headline about the costs of flagship unmanned missions is the height of irony, given the huge costs of NASA sending people to endlessly orbit the Earth with little to no scientific value.
The primary failing of human spaceflight has been failure to reduce cost to a level consistent with what customers are willing to pay. Hopefully with Commercial Crew that is beginning to change. Obviously the situation is different in The Program That Cannot Be Named.