Hacker News new | past | comments | ask | show | jobs | submit login
NASA cancels $450M mission to drill for ice on the Moon (nature.com)
53 points by gnabgib 3 months ago | hide | past | favorite | 52 comments



I'm all for avoiding sunk-cost thinking, but considering that nearly half a billion was spent on this thing and its launch was barely a year away, it's hard to wrap my head around scrapping it entirely.

From the lost talent alone, this seems catastrophic - why would any of the country's best engineers want to work on ambitious NASA projects when they can be rug-pulled so close to completion?

Also, where's the accountability or at least lessons learned in this cancelation? Is there a single finding on how to reduce cost overruns for future missions?


The Superconducting Super Collider was hoping to get to that "too far along to cancel" point as well. They thought that reaching 80% tunnel completion would do it. However it didn't, and it was canceled. I'm sure there's plenty of other examples as well.


> avoiding sunk-cost thinking

The past is in the past. The half a billion matters insofar as it's given us our current playing hand. The right mentality is some combination of looking forward (we have many options, some of which were bought for half a billion, but even those might be worthless), backward looking (half a billion is probably worth having a retro over; was the outcome regrettable? If so, how can we do less of that?), and keeping non-obvious systemic effects in mind while doing both of those things (e.g., if you never finish a project, it doesn't matter if you were always working toward the "right" thing, and with that sort of thing in mind your bias toward cancelling projects should perhaps be weaker than the immediate facts might indicate).

Imagine you'd invested half a billion in Theranos. You really would have pennies on the dollar in return, and completely scrapping it (or otherwise selling it for parts) could be a logical move. The sunk cost fallacy comes in when you decide that because you've already invested you need to continue in that direction.

Your point about it being hard to wrap your head around scrapping it entirely is a little subtle. We very well might have bought something worthwhile for that half a billion. In that case though, you still want to look at the resources you _currently_ have available or expect to have in the future, not the fact that you lost a half billion (except insofar as you can use that as a learning experience to not fuck up so much in the future).


The original approval for VIPER was $433M. So at one time they thought that VIPER was worth $433M. Now they need $100M+ to finish it. If it was worth $433M at one point, surely it's worth at least $100M now?


That was $433M with less information than they have now. To be clear, my personal bias is that there's some sort of graft or other major problem, and it would be worth $100M, but we might have also spent that money as a high-risk, high-reward endeavor and simply found out later that the rewards aren't worth it. That isn't even necessarily a bad outcome, if true.


Not entirely on NASA, Congress slashed their budget causing a -$500M hole, with future years budget not looking to fare much better so they had to make some tough decisions. Sounds like other missions in the same program are going to pick up VIPER's objectives though.


Sure looks like they cut it 2% in 2024, but they've increased it every year over the last 10 years, even in real terms (maybe 2022/2023 was a bit under flat)

Year Budget Change

2000 13,428

2001 14,095 4.97%

2002 14,405 2.20%

2003 14,610 1.42%

2004 15,152 3.71%

2005 15,602 2.97%

2006 15,125 -3.06%

2007 15,861 4.87%

2008 17,833 12.43%

2009 17,782 -0.29%

2010 18,724 5.30%

2011 18,448 -1.47%

2012 17,770 -3.68%

2013 16,865 -5.09%

2014 17,647 4.64%

2015 18,010 2.06%

2016 19,300 7.16%

2017 19,508 1.08%

2018 20,736 6.29%

2019 21,500 3.68%

2020 22,629 5.25%

2021 23,271 2.84%

2022 24,041 3.31%

2023 25,384 5.59%

2024 24,875 -2.01%

https://en.wikipedia.org/wiki/Budget_of_NASA


The problem is that most of that budget is earmarked for huge money-wasters like SLS, and those earmarks are growing. So the non-earmarked budget shrinks.


This is a huge part of it.

Congress might increase the budget by 3% but when they force something like SLS on NASA that eats 6% of the budget in real terms it was a 3% cut.

And this assumes the employees don't deserve any raises or bonuses.


You’re looking at the Nominal Dollars (Millions) column which is meaningless. The 2023 Constant Dollars (Millions) has

  2010: 26,162 
  2015: 23,150
  2020: 26,559
Then the actual budget in 2023 is down to 25,384, and 2024 is 24,124 a 5% cut from the year before. https://www.usinflationcalculator.com/


In real terms that looks pretty flat for both the last 10 years and since the beginning of the list. Doesn't explain the project cut but also isn't really real budget growth either.


> From the lost talent alone, this seems catastrophic - why would any of the country's best engineers want to work on ambitious NASA projects when they can be rug-pulled so close to completion?

That's the point! By making government dysfunctional we can point out all the wasted funds and slash the budget /Reagan


With the recently discovered caves in the moon[0] I always thought this particular mission had unfortunate timing since the caves are pretty far from the south pole. Maybe its cancelation is for the best after all.

Huge sunk cost though, it might lower confidence in future budget talks.

0: https://www.space.com/moon-cave-lunar-exploration-radar-imag...


It could be a win-win. Just imagine if NASA canceled the Starliner before it launched. Boeing could have been saved the embarrassment they are currently enduring, but they weren't so lucky to have their project canceled.


While NASA helped fund development and lend expertise, Starliner is not a NASA project to cancel. It's a Boeing/ULA project, and it's likely in Boeing/ULA's interest to develop manned space flight capabilities regardless of what NASA's needs might be today or in the near future.


That's a very optimistic view on the history of the project. Of course Boeing wants to compete for these huge contracts and they'd love to have it be sustainable outside of singular NASA projects, but that still doesn't make it the reality of the situation.


There's a lot more parties interested in putting things and people into space than just NASA...


Previous discussion: https://news.ycombinator.com/item?id=40991699

I'll note that it sounds like they were relying on a 3rd party to make their descent module, but it won't be ready in time. So if they were to go ahead with the launch they wouldn't have a way to land the thing.


450M just for building the robot seems a bit steep; its not like they're actually getting it to the moon for that price. How much cheaper is it than just sending 3 dudes in a tin can to just dig the holes themselves?

Edit: I guess NASA says manned missions will cost billions. Still seems like half a billion is a lot for a glorified DJI drone thats only gotta make it to the moon lol. Regardless, money spent towards any space project tends to be money well spent, so compared to all the other spending I'd like to complain about, this isn't even on the radar.


If the documentary, "Armageddon" taught me anything, they'll do it if you just agree that they never have to pay taxes again, they can sleep in the Lincoln bedroom, and that you will cancel out a few parking tickets.


You're paying for provably guaranteed reliability.

When you're paying billions per rocket launch, you simply can't afford to have your robot break down on day 1 because of a software bug or because a component was accidently installed upside down.

(Also, "only gotta make it to the moon" makes it sound like that's a trivial task. There's a lot to unpack in that small sentence.)


You can if you send 10 cheap robots instead of one half-billion dollar one.


weight


I think you underestimate the engineering complexity a bit of these rovers :)


The very first Apple Watch cost a billion dollars to make. Costs got quickly amortized after that first unit.

If NASA was in the habit of building hundreds of robots at a time, the cost per unit would drop dramatically.


When you put it in that context, the rovers look like an absolute bargain!


> a glorified DJI drone

Where "glorified" means "can withstand temperature, radiation, and dust conditions that would brick a DJI drone in short order.


The risk of losing 3 dudes is greater than $450m.


People put valuations on lives all the time in risk analysis and I've never seen a cost even close to $100m.

DOT puts it at 13.2m: https://www.transportation.gov/office-policy/transportation-...


Factor in the PR impact on the entire space program given the public visibility.

We've never had an astronaut crew get stranded on the moon. though we got close with Apollo 13. If/when that happens for the first time, you'd better believe the entire planet will be paying attention.

Just the congressional inquiries alone will set the space program back by decades.


This. It seems average Americans, including average hn commenters, have their future outlook confined to the next fiscal quarter.


Astronauts are very visible deaths for politicians whose currency is points in the polls. Having your photo not show up on the news next to a photo of a dead square-jawed captain america astronaut is probably worth 100m of other peoples money.


They wouldn't be "astronauts". They'd be cargo.


Finding someone who agrees to go to the moon with a risk of 1 out 2 of dying, and 1 of 2 of becoming a historical hero is really doable.

Plenty of volunteers, and no need for 100M USD.

People go to war for less than 50K USD.


The city I'm from, there are people who'll kill for USD equivalent of $100. That does not mean we should encourage it.


The odds of dying in a war are less than 1 in 2, and for joining the military in general it is far far less.


?! I was like, "oh really?" and I checked... you're right, "a German soldier had approximately a 1 in 3 chance of dying during the WW2 conflict"

Dying to protect a political system vs dying to advance science, better pick science.


And that is an exceptional example. For most modern wars or major conflicts it's nowhere near as bad as 1 in 3.


In WWII young men volunteered to join RAF bomber crews and faced an almost 1-in-2 chance of dying, with a far higher chance of getting wounded or captured.


Thats for the average person - whats the cost to replace the average astronaut? I've seen estimates that it costs $15M just to train 1 astronaut, and the pool of qualified candidates is likely extremely small. I would figure a guess of $100M per astronaut is not unreasonable.


Risk avoidance has its place, but fear didn't get us to the moon the first time and I'm confident it wont work this time either.


You're probably right in this specific example, or at least you're probably in-line with NASA's leadership's thinking.

I am willing to bet that other parts of society (e.g. traffic planners and automobile manufacturers) have a much lower cost of human life.


For NASA the cost of dead people includes bad PR


Not really, SpaceX would just sell that as an unscheduled disassembly + “we collected data”.

It’s widely accepted that you may die going to space in experimental vehicles, nobody ever said it would be safe, and nobody can reasonably think it is risk-free.

Apollo 1 folks died, not a problem for reputation of NASA.


I’m sure there’s more to the story, but half a billion dollars seems like a staggeringly expensive price tag for one piece of a moon mission. What does all that money go to? What’s the ROI?

We’ve been sending rovers to other planets for decades, so I assume there’s a good base of knowledge to build on. This isn’t basic research level, it seems like it should be a lot closer to production work. Satellites go up nearly every day, so I’d guess the space-spec supply chain is healthy.

Do post-mortem documents get published on programs like this? I really can’t understand it.


We can spend $300M dollars to grade-separate one intersection between a road and a train.



I think with SpaceX we have the situation similar to computing - ie. whether to start a 20 year computation today on a $1B hardware or wait 5 years and complete it in 5 years then on a $100K hardware - to proceed to fund the $450M+ for the many years ahead or just wait a couple years until SpaceX gets to it and has it done for $10M in half a year.


When a project like this is cancelled, where does the hardware go? Not just flight hardware which I’m sure is ITAR EAR restricted but also the plethora of test and supporting hardware to go along with it?


seems like we don't have any active or planned missions to land on the moon anymore. bummer. (happy to be proven wrong) https://www.nasa.gov/missions/?terms=10828%2C10873%2C10900%2...


Artemis III will land on the moon (not before 2027 though).

Maybe faster if someone has an old spare Gameboy to lend to them.


Must be the alien base on the moon




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: