Penny for NASA (Centennial Challenges)

I remember the first time I was exposed to the fact that NASA gets less than 1% of the federal budget, and the suggestion that we could do so much more awesome stuff if NASA just got 1% of the budget like it used to. I was 16 at the time, and I was in my freshman year at BYU. I was so fired-up by this idea that I used the idea as the topic for a persuasive writing essay for my Creative Writing class I was taking that semester. The sad thing is that in the 17 years since then, the Penny for NASA arguments have gotten no more convincing than my paper (which probably only earned me a C+).

My biggest beef with the Penny for NASA concept is that I think NASA is doing a lackluster job of effectively spending its current ~$17B/yr. There are not so far-off technologies that could slice the cost of deep-space manned missions by 2-5x, but funding for those technologies are getting starved, while most of the money gets sucked-up by politically untouchable mega-projects. If NASA’s budget were doubled, Congress would probably run out of excuses, and send some money towards technology development and demonstration, but that funding would also likely be the first to get cut if NASA’s budget were ever reduced in the future. And seriously, who here thinks that if NASA’s budget were doubled it wouldn’t be come a lightning rod for cost cutting in the future? No, I’d rather see NASA more effectively spend the money it has, and is somewhat likely to keep, than hoping for solving all problems by shoveling more money at the agency.

What I think would be a more useful spending goal would be to advocate for spending at least 1% of NASA’s budget on prizes, such as the ones that have been run by the Centennial Challenges.

You’ve heard the benefits of prizes before, but to reiterate:

  1. Prizes only reward success, not effort. Far too much of NASA’s money is spent in a way that does not guarantee the taxpayer gets anything in exchange for huge investments.
  2. Far less of the money and effort associated with winning a prize is spent satisfying bureaucratic requirements/oversight than with even COTS-like contracts. There are some regulations such as FAA regs that have to be followed, but the overall percentage of prize efforts spent on hardware/operations versus paperwork is much more optimal.
  3. Well-structured prizes often help encourage multiple players in an industry even after the prize is completed.
  4. Prizes encourage creative, out-of-the-box solutions that might have been rejected by a selection committee for normal funding.
  5. Prizes can often encourage non-traditional players to compete.
  6. Prizes are often more exciting and engaging for the public than similar traditionally funded technical efforts.
  7. Prizes can often get ego-capital investments into risky new technologies that might not have happened based purely on financial merits.

That said, prizes don’t solve all problems, I’ll be the first to admit that. Prizes don’t guarantee that people will be able to raise the money to pursue them for instance. Especially when the prize value is too low compared to the likely cost to complete it, and when there’s no clear near-term commercial follow-on opportunity. Prizes tend to attract for more teams than it attracts credible teams. Of the something like 20 X-Prize teams, I think that Armadillo was the only other semi-credible team other than Scaled, and they were way behind Burt Rutan’s team. Of the something like 15 teams competing in the NGLLC, we did better, with about 5 pretty serious teams by the end (in addition to the two winners, Unreasonable Rocket, TrueZer0, and Paragon Labs all at least built and had stable tether tests with vehicles that could’ve won the prize with another year of practice and refinement). No comment on GLXP.

But while there are definite tradeoffs with prizes, it’s also pretty clear that they’re getting woefully underfunded compared to the rest of NASA’s spending priorities. At its best year, I think Centennial Challenges came close to getting 0.1% of NASA’s total funding, and most years it has received significantly less than 0.05%. While I agree there is room for debate on how effective prizes are compared to traditional contracts, I’m pretty sure nobody seriously thinks that they are 1000-2000x less effective.

So, I propose the idea of setting aside 1% of NASA’s budget for prizes.

Really, in the grand scheme of things, a 1% tax on other NASA programs wouldn’t be felt by most of them, only the programs that have been hit the most with spending cuts (like Planetary Science) would likely notice them at all. And having a pot of $170M per year for prizes would enable a wide range of prizes that NASA has so far been unable to even offer, including prizes related to technology development/demonstration capabilities they’d like to fund, including: reusable launch systems, nanosat launch, microreturn vehicles, interplanetary cubesat missions, Lunar/Asteroidal/Mars/Venus sample return and/or ISRU demonstration, cryogenic propellant storage, handling, and transfer technologies, deep-space human spaceflight issues (like radiation protection, artificial gravity, etc). When you’re only getting $4M/yr or less, you’re stuck with either funding prizes for really tiny efforts, or badly underfunding prizes for slightly bigger (but still quite small) efforts.

Here are a few suggestions I’d have for making this better funded Centennial Challenges program more successful:

  1. Place a cap on the fraction of the budget that can be spent on administrative personnel. I’d suggest say 5%. 5% of $170M still works out to over 40 full-time support staff.
  2. Setup Prize Manager positions in the Centennial Challenges program in a method similar to how DARPA handles Program Managers, where they are only brought on for a set period of time (typically about 3 years). Encourage bringing people on from academia, industry, and military space, not just NASA employees. This encourages more diversity of thought and cross-pollination, while also decreasing the ability for people to empire-build.
  3. Make prize winnings tax-exempt (I’m not sure if they already are, but they ought to be–that’s one way of making them more lucrative that doesn’t cost much extra money).
  4. Set aside some money for paying the partnering groups that run the prizes. I’d still limit this to no more than say 5-10% of the total budget. But providing some financing to the prize groups would mean that the prize groups wouldn’t be spending so much of their time trying to raise money for themselves instead of running the prizes effectively.
  5. It might also be worth setting aside one last 5% of the budget for media/promotion of the prizes and teams. If the prize organizations aren’t having to raise money to cover their own cost, and media groups aren’t having to pitch Hollywood on funding media efforts, there’s a lot higher probability of being able to gain visibility for both the prizes and the teams involved.
  6. Require that the other 75% of the budget can only be spent on prize payouts, and keep the current Centennial Challenges ability to retain money unclaimed from year-to-year.

While there are probably tweaks or adjustments to this plan that could make it work better, and while 1% is a bit ambitious, I still think this is much better than just trying to up NASA’s topline budget by a factor of 2x, and a lot more likely to result in lasting benefits than doubling NASA’s budget.

 

This entry was posted in Commercial Space, NASA, Prizes, Space Policy. Bookmark the permalink.

16 Responses to Penny for NASA (Centennial Challenges)

  1. kert says:

    I think its useful to note that neither of the previous space X-prizes have lead to successful companies yet – most of the participants have dissolved their teams and moved on to something else.
    I think if you want to set up a successful prize program, you have to plan ahead on how to sustain the interest in the area and lead to further development.
    One of the very good examples of a successful prize program is RoboCup – the amount of research , engineering and teams built around it is nothing short of impressive, and technology is continuously improving. The key to its success however is that it keeps the cores teams together, year after year, even with people joining and leaving as it always happens. IMO, successful teams and organizations actually matter more in the big picture than individual technical achievements.

    I think doing something with Cubesats would be a very useful frontier for space prizes – there is already a large innovative community around the world building them, mostly in universities.

  2. Jonathan Goff says:

    Kert,
    One thing to keep in mind regarding the X-Prize, NGLLC, and GLXP is that most of the teams just flat-out aren’t credible. This isn’t like a high-school or even college robotics competition. Building safe, repeatable rocket-powered vehicles costs a lot more money, and most of the teams were kidding themselves. I know that sounds harsh, but less than 25% of the X-Prize or NGLLC teams ever successfully flew anything, even once.
    If you only look at the teams I would call credible, the numbers are a bit better. Of the teams that won, 2/3 of them are still around, and all three of them benefited significantly from having competed in the prize. Armadillo is the only one of the three companies that won one of the prizes that went out of businesses. And if you look at their history, it’s pretty clear that the NGLLC opened up a bunch of opportunities for them, but they got unlucky with their decision to drop all the contracting work and shoot for the 100km vehicle directly. Had they kept with all of the projects they had brought on due to their NGLLC fame, they’d probably be a successful company to this day, probably cashflow positive, and making gradual progress towards their end goals. Of the non-winning teams that were credible, the record isn’t as good. Most of them did end up folding.

    But is the goal of prizes really to keep teams together? In some cases, sure! And I’d love to see some fraction of this proposed 1% for Centennial Challenges budget spent on prizes of that nature. But there are other prizes where the more important concern is proving out a capability in a way that leads to a follow-on commercial capability.

    Is the record mixed? Sure. Scaled and Masten’s wins have not yet translated into operational capabilities. But really, even in the simplest of industries, failure rates for startups are really high. Just cause we’ve tried something inexpensive twice, and gotten inconclusive (so far) results, doesn’t mean it’s a bad idea. For comparison, is CC’s record as mixed as say MSFC’s record with multi-billion dollar space launch projects over my lifetime?

    Not saying your suggestion is a bad one–just saying that not all prizes ought to fit that mold. They have their place, and I’d love to see say 15-20% of CC’s budget go into ongoing prize competitions like that in the cubesat, UAV, robotics, or nanosat launch ares (for instance), but there are other prizes that I think should still be done based on 1st and 2nd across the finish line approach.

    ~Jon

  3. Lars says:

    A good plan, but I think it would be much more sustainable if it was scaled up over a period of years.
    A thing about prizes I have noticed, is that credible teams don’t put a lot of effort in them without a really good shot at winning. It is a lot of risk to take on if there is no second place. I feel that there is a limit on the amount of prizes an industry can take in a short period of time without becoming collectively ‘burned-out’ and becoming more risk averse.

    I 100% agree that a NASA that was spending 17 billion properly would not have problems with funding.

  4. kert says:

    >>It is a lot of risk to take on if there is no second place.

    Yes, exactly – but a very workable “second place” is the opportunity to come back next year, to slightly more difficult challenge goals, and take another swing at the front runner – learning from and absolutely eliminating the one (or five ) small mistake(s) you made.
    I have been in that slot myself, more than a decade ago. In a robotics compo we didnt pay much attention to our cable connectors, and on the critical run we had one come loose. Regardless of the superiority of our overall solution which we demoed ten times over after the actual compo, we lost – and not just the competition.

    I still think if you want to make progress in a technology area, its much more important to set the incentives and goals that promote formation of good, healthy and sustainable teams, rather than set particular technical goals.
    IMHO even Darpa Grand challenges, both urban and desert one, were much better set up than NGLLC for instance. Both had multiple teams cross their finish lines with very workable tech, opening the doors for sponsorships and further tech incubation even if the team didnt “win” on the competition metrics.

  5. Nathan says:

    What surprises me a little is that no private individual has set up a prize giving space foundation. Given the wealth of Bezos ($25 billion) as well as the other space billionaires: Branson, Allen, Musk, they could easily afford it.
    A billion dollars fund could give out ~ $40 million per year in funding, which isn’t as much as you are talking about but would be still be considerable amount.
    They could also devote say half of that amount to bigger longer term prizes, for more difficult challenges. A challenge meant to take 5 years to accomplish would therefore have a prize fund of $200 million behind it.

  6. DougSpace says:

    Prizes tend to be one-time payments at the end. There are no payments for milestones achieved. I think that this is one of the biggest problems with prizes. There is a limit to how far companies will go out on a limb before the competed chance of winning. Payments for milestones gives a partial or even full return on their partial investment so that participating companies can afford to continue developing large, complex, and risky programs.

    LunarCOTS.com

  7. Karl Hallowell says:

    What surprises me a little is that no private individual has set up a prize giving space foundation.

    There’s the X Prize foundation.

  8. Dennis Tito has all but given up raising the funding for Insirpation Mars because everyone he spoke to wanted to know why NASA wan’t behind him. This led to his diasterous testimony before Congress and the whole “we consider it a NASA project now” madness.

    So, does the private space industry need more prizes and “support” from NASA? Sure, if you’re one of these people who think the solution to too much government is more government.

    What the private space industry needs is a clear statement of disinterest from NASA and a clear business case. That answers the key question: why should I give you money?

    What’s great is that we have actual examples of this today.

  9. Paul451 says:

    I’ll third Lars/Kert’s points. Prizes need to be regular, scaling, ongoing. Single prize/single goals don’t really work, IMO. (You even see that with the big international goals, like Apollo. Once the “space race” was won, everyone lost interest.)

    DARPA’s Grand (later Urban) Challenge illustrates the advantage; in the first year, not a single competitor succeeded in the simplified challenge. In the second year, most teams finished the more complex challenge. And now they’ve added traffic and an vastly more difficult urban environment.

    The original Ansari X-Prize was a single prize. And no one even made an attempt for 8 years. Then it was won, and no one has bothered to repeat the effort for the last 9 years. Not even the winner. So I wonder if it would have been better to instead have a $1m prize every year for a decade for merely the highest manned/reusable flight each year. (Not necessarily over 100km.) Would that have stimulated more competition, and more development? By the tenth year, would you have a bunch of teams flying well beyond 100km? And running tourist flights in the “off-season”?

  10. Karl Hallowell says:

    Then it was won, and no one has bothered to repeat the effort for the last 9 years. Not even the winner.

    Paul, Virgin Galactic has been steadily working towards constructing (via contract with Scaled Composites) and testing a number of SpaceShipTwo vehicles. Glancing at Wikipedia, it is stated that the SpaceShipTwo prototype has flown 26 test flights by now, including two powered flights. In addition, there have been, again according to Wikipedia, 141 test flights through December 11 of this year by the carrier plane, VMS Eve. So I’d say that someone is indeed following up on the effort of SpaceShipOne, the winner to the original X Prize.

  11. Paul451 says:

    it is stated that the SpaceShipTwo prototype has flown 26 test flights by now

    Mostly carry and drop tests. Two short powered flights. Not a single flight over the 100km, the original prize. That was my point. They may be intending to do so, but they’ve been “intending to do so” for nearly a decade since the first two successful prize flights. The Ansari X-Prize did not stimulate the sort of competition and iterative development that, say, the DARPA Grand Challenge did.

  12. kert says:

    DARPA DRC just showed again IMHO that they have a pretty good model for really fostering innovation. They are not picking winners and not spending a ton of up front cash, but they really do catalyze things. From what i understand they will now drop about $8M across 8 winning teams and significantly up the ante for the “finals” – which will be very representative of their end goals. At the end, they will have enough design DNA and talent pool around to start Clone Wars if they like.

  13. Karl Hallowell says:

    But here’s the thing, Paul. SpaceShipTwo is a commercial vehicle intended for safe transport of humans who aren’t test pilots in a vehicle that has a lot of things to test. What should we expect at this stage aside from a lot of testing?

    My view is that jumping straight to 100 km flights just wouldn’t be serious. There’s too much that could go wrong and which can be tested and fixed without the full mission profile. It’s a huge, costly gamble.

    For example, starting to work the bugs out of the White Knight 2 carrier plane doesn’t require a lot of 100 km flights with SpaceShipTwo. Even with the full mission profile, White Knight 2 has to land and do other significant activities without a working SpaceShipTwo attached.

    Similarly, the ground crew needs practice prepping and running flights and maintaining the vehicles between flights. Pilots need to build up flight time. So these configurations, activities, and crew training and experience can be done and tested without requiring the full blown setup for a fraction of the cost and risk.

    For example, Project HARP and the X-15 are examples of two government-run aerospace programs that took this healthier approach to developing a novel technology and working the bugs out of it.

    My view on this is that it isn’t enough to “bend metal”, but one also needs to do extensive, incremental testing of the resulting product. Short cuts result in stuff that doesn’t work and causes spectacular accidents.

  14. Karl, Paul,
    I think that VG took the wrong approach with SS2. They would’ve been better off doing an SS1.1 that upgraded SS1 to something that was safe enough for operations, and then started early operations while developing SS2 in parallel. Trying to simultaneously triple the size of the spacecraft while increasing its quality for spaceflight participants looks like it was a bridge too far.

    But the fact that a market with this few players is having delays shouldn’t be a massive surprise to anyone. Even of low-risk startups, the mortality rate over 10 years is >80%. I don’t think that’s the prize’s fault at all.

    ~Jon

  15. Paul451 says:

    Karl,
    “My view is that jumping straight to 100 km flights just wouldn’t be serious.”

    And you make my point. The $10m Ansari X-Prize did exactly that. And thus it took eight years for someone to claim it, and not a single group has repeated the feat in the nine years since.

    Whereas, IMO, a $1m annual or $2m biennial challenge for the highest repeated flight in a reusable vehicle would have delivered much more before the money ran out; even if not a single competitor had reached 100km by the end of the challenge series, I really think the field (the industry) would have been healthier.

    Jon,
    I don’t “blame” the X-prize for the lack of competitors, I just don’t think it was an efficient way of stimulating development. I feel the same about Google’s lunar prize, and many singular NASA challenges (and even more NASA programs.)

    [And if I had $10m to spare, I'd certainly put my money where my mouth is by funding an annual highest-flight challenge. I even tried to interest NASA in it, when they were running one of their public-idea contests some years back. But I think I was listed down with the "bUiLd aNtI-gRaViTy!!1!" people.]

  16. Jonathan Goff says:

    Paul,
    While I can see some advantages to doing a series of yearly “best results” prizes, I can also see some challenges.

    For example, with the X-Prize, they only ever raised something like $4.5M or so in prize money (they got to the full $10M via an insurance bet). If they had done $1M prizes for the highest flyer of a given year or two, they could easily have spent all of their money on dead-ends without ever achieving their overall goal.

    Had the NGLLC had say two smaller prizes for the longest duration hovers by any team, I still don’t think you would’ve seen faster progress. In both cases there was some minimal level of capability you had to get to before anything happened, and after that the performance of L1 vs L2 wasn’t that big of a deal compared to just getting to a “reliably hovering rocket”. Sure, if there had been further follow-on prizes, that would’ve likely helped Masten, but AA already decided that prizes and contracts were a waste and they wanted to push straight for a commercially viable product.

    Basically, I agree that there really are cases where “continual yearly competition” prizes make sense, but that there are other cases where a “1st (and maybe 2nd) past the finish-line” approach are a better fit.

    ~Jon

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>