I remember talking to a high up on the program at a FIRST robotics event who complained about them running so long. They weren't making any major new discoveries and sucking up resources and mindshare stopping him for kicking off and focusing on new projects. He said nobody would take the PR hit to decide to just end a mission like this so he was stuck.

Sounds like that person doesn't really understand that NASA is a political entity and the Public Relations of an ongoing successful mission is more valuable long term than more scientifically pragmatic experiments in the immediate.

Sounds like s/he perfectly understood that, but still didn't like it. Often (always?) politics get in the way of actually getting things done in a smart fast efficient manner.

The thing is, you can substitute "politics" for "people" and get the same statement, which is why NASA's political work is at least as important to human space exploration (and, I would argue, long-term species survival) as the raw scientific work.

People have to be continually convinced that exploration of the solar system is a worthwhile expenditure of a lot of money that could be spent on other things. Seems obvious to those who are already thinking in the direction of where humanity is going, but most people do not.

It's still possible to simultaneously acknowledge the fact that we need to repeatedly convince people of space exploration being important while also disliking that this is necessary.

Politics also affords the collective resources to engage on such projects. It's more chicken-egg than most think.

Not to mention, science without proper PR has a disastrous level of near-term utility. See: climate science

What's the limiting factor on just running with reduced staff support? (Honest question.)

A small handful of people can still easily cost millions of dollars per year. And DSN time to talk to the rovers is expensive, probably in the neighborhood of $2,000 per hour.

How much DSN time do you need. Given the deley, I assume the only use is batch transfers, so cost per bit seems like the more useful metric (along with how much data is involved)

$5M a year seems like it would be very reasonable. The original 90-day price tag was $820M, and they've spent an additional ~$200M over the past decade and a half. Maybe it's really not worth it to spend $5M/year to extend the life further, but the scientific returns would have to be quite severely diminishing.

I had the same question, I don't imagine it takes too many people to decide what it should do and give the robot commands while monitoring it.

Looks like according to an old article linked on wikipedia they spent another $100+ million over 5-6 years to keep it going. So not wildly expensive, but not cheap either.

https://web.archive.org/web/20120504171453/http://www.msnbc....

> I remember talking to a high up on the program at a FIRST robotics event who complained about them running so long.

Not surprising at all. Coached a FIRST team for several years. At this point they sell licensed robotics kits with canned instructions for kids to put together and those kids win because of time limitations. The teams that build from scratch and actually contribute discover or learn anything at all are intentionally blocked from winning. The entire program is total crap for retarded children. Meanwhile smart kids are off doing their own projects that are 1000x as complicated as the most challenging FIRST challenges. It's just turned into a money burn.

Retarded is a bit much, don't you think? We are talking about children here after all.

...what? They've always given out the Kit Of Parts and I don't recall those teams that relied on them ever winning anything substantial.

> The teams that build from scratch and actually contribute discover or learn anything at all are intentionally blocked from winning

This is ridiculous, and I'm saying it is from the perspective of having been a student whose team finished fourth in FRC in a recent year (and won multiple events per year), and having volunteered and assisted multiple teams.

While it is true that a lot of lower resources teams may be better off going with commerical off the shelf (COTS) components, such as the Greyt Elevator and Greyt Intake, teams that perform at a higher level will build everything from scratch. Whether this means utilizing gussets or welding, building your own drive train or using the AndyMark prebuilt chassis.

While it is true that some kids may be better off doing research projects, there is nothing wrong with participating in a team. From my three years as a student, I was exposed to developing solutions to split messages up and verify data integrity when communicating over serial with external microprocessors, utilizing version control, setting up Continuous Integration on our repositories, first deploying apps from Cordova to developing Progressive Web Apps, which forced me to learn to use a VPS on Digital Ocean, running a docker swam on hyper.sh, and storing files on S3 and using RDS.

FIRST is what you make of it, and I'm personally glad that someone holding your opinions towards the program is no longer actively interacting with kids in the program.

It's well known that the best teams are the ones that eschew using the kit of parts and have their NASA/SpaceX/Autocorp sponsors laser cut their robot chassis for them.

That story is very much over sold. They defiantly designed the things with a longer lifespan in mind. NASA simply had more reasonable minimum specks so if they died on day 220 nobody would call the program a failure.

Oversold?

A group of engineers put a working robot on a moving object 140 million miles away, and that robot trucked along for fourteen years.

I honestly do not see any aspect of this which is not totally and completely astonishing, marvelous, and awe-inspiring.

They designed it for a winter which means at least 687+ earth days days was considered achievable. Dust was expected to be a larger issue and in many ways they lucked out, but more so as a calculated risk than blind luck.

Distance makes little difference past the first day. Doing the same thing on the moon would not be easier becase it’s closer. It’s local conditions that make this hard more so than simply getting to that point.

I can assure you that the distance took its toll on the rovers, in my understanding. Their wheel surfaces eroded, oil pumps failed, bearings degraded and they ended up dragging dead wheels in the end.

Are you talking about the distance the rovers traveled while on mars? Because that's not what Retric is talking about. Retric is talking about the distance from the earth to mars (hence why they compared it to the moon).

Ah, my mistake; I misunderstood their comment.

To me, that they designed it to last that long, and it actually did, just makes it more impressive. Otherwise it would just be blind luck.

Sort of.

I think the book "Roving Mars" covers this the best. They were worried about a dust storm like the one that killed Opportunity, so they designed the rover to last 90 days even in the pitch darkness of a dust storm. To make sure they'd succeed even in this worst case, they added two additional petals worth of solar panels (IIRC, about a 50% increase in area) to the MERs.

Reading a contemporary description, they seemed to be very scared they wouldn't make the 90 days - hence this major and risky redesign. (The new panels had to open, and if one failed it would block off both the new solar panel, and one of the original ones.)

It worked, and so they had enough extra power to last through that first Martian year, where the winds blew the panels clean. And the rest is history.

I think they solve the power problem by bringing a nuclear reactor as the main power source.

Different rover. Spirit and opportunity aren’t RTG powered; curiosity is.

Yes, they probably designed it to last for a year. No one builds bridges for the exact expected load either.

There is no chance anyone involved in that program expected the rover to last for 5 years, let alone 15.

I disagree. Even without accounting for expectations, making a machine that works for years without regular scheduled maintenance in extreme conditions is a monumental feat. Ask people who work in the oilfield, extreme heats, extreme humidity, etc.

The discussion is about 90 days, not years. Years is awesome. 90 days was a planning goal used to set engineering priorities.

Kinda/sorta.

90 days was because they didn't think a rover would last over 90 days due to dust storms. They figured that after 90 days they'd encounter a bad dust storm that would cover the solar panels and make them unable to collect power, and thus the battery would be drained and unable to charge after 90 days.

They didn't account that the winds on mars would be able to clean the solar dust off the solar cells so well that it turned out dust storms weren't a huge issue until the latest one.

And from my understanding it wasn't the intensity of the latest storm that killed it but the part of the rover's computer that kept track of time was shot, so it wasn't able to optimize when to go to sleep and when to wake up properly, causing it to run out of charge while it was dark.

Expectation management is sometimes as important as the actual delivery.

It's pretty much what you do in any successful enterprise. Under promise, over deliver.

That's just startup and sales speak. I've found that delivering exactly as promised is more than enough to keep customers happy.

> That's just startup and sales speak.

Nah, it's a method for giving you a lot of wiggle room if you (or your business) fucks up in something.

I wasn't just commenting on the physical durability of the rovers themselves although that is a huge part of it. It is also impressive from both a political and an organizational standpoint to keep a program like this running successfully for so much longer than initial plans dictated.

NASA has numerous programs running for this length of time. One of the people working on Voyager showed up at my grad school to get a degree between Uranus and Neptune.

NSF, too. The VLA opened in 1978 and the first data recorded by it can be reduced by the current radio astronomy analysis software. In fact the software data format developed for the VLA (FITS) is now popular with high-end camera people... 41 years later.

> One of the people working on Voyager showed up at my grad school to get a degree between Uranus and Neptune.

So, did he make it beyond Pluto or not?

She.

Underpromise and over deliver :)

underpromise and way over deliver. This is unheard of.

JPL wrote a fascinating reflection on what engineering practices allowed this to happen. For any of you who have not experienced the NASA Knowledge Gateway, I would recommend. https://llis.nasa.gov/lesson/1743

Nice concise read. I think many lessons are well known but not acknowledged. For a high quality product...Test a lot, experienced people, do things in house.

Meanwhile software industry is... Ship to prod ASAP, hire junior devs to reach head count, outsource to overseas and contractors.

I realize the risks are different... But management gets pissed when software causes financial lost and wonder why.

One systemic problem is that the ownership and management of most companies shifts significantly over a 15 year time span, whereas the people designing a new NASA program know that they (or people that they care about) would be picking up the pieces if they are sloppy.

I wonder whether family owned companies are closer to NASA in their testing/hiring practices?

It's a testament to the engineering excellence happening at NASA.

Also a testament to the value of under-promising. :)

If I remember correctly the reason they did not expect the mission to last longer than that was because they expected the solar panels to get covered with dust and thus render them useless. I don't think they had banked on martial winds to regularly clean them.

Your recollection is correct. Consider too the dust accumulation predictions for the moon landing. It would appear that we (as humans) are not very good at dust estimates (dustimates?). This is reinforced by the continuing struggle/inability to manufacture dirt. As an avid composter with a ton (lol sorry) of experience, I can make you hummus for days. I can turn 90% of your organic household waste into growth medium, bokashi tea, or pre-dirt in anywhere from 6 months to 2 years. But if you want me to make dirt? I'll need a decade to get you a production level process that requires raw stocks, pre-processing, vermiculture, aquaculture, and many rounds of late stage funding. Dirt production stymies every modern industrial process because as it stands, you cannot reduce the time necessary for conversion from hummus to dirt. I find it beautifully tragic that terraforming Mars will require mass dirt transfer from earth (consider this tonnage conundrum and revel in the irony).

Don't take my word for it. Go out and try to make earth yourself. If you can reinvent this fundamental, elementary process, you will join Jonas Salk and Fritz Haber, and very few others, in changing The Game in a fundamental way.

Kim Stanley Robinson's Mars trilogy talks a lot about the "can't make dirt fast" problem, and I was surprised and impressed at how fascinating such a mundane-appearing problem could be. I had never thought about it before.

Note: I think you mean humus. Hummus is "a Levantine dip or spread made from cooked, mashed chickpeas... blended with tahini, olive oil, lemon juice, salt and garlic."

Autocorrect has outed me as a desirer of dips and spreads more so than an urban farmer. The shame and ignominy is overwhelming.

Hummus? Or humus? Because one I'd like to eat with pita bread and a nice shawarma.The other...not so much :)

You are correct. Incidentally, humus is a delightful compliment to a sound horticultural spread, not unlike it's etymological cousin. It can make things better, but it cannot be the main course.

Jumping on the opportunity to have somebody who knows about those things.

I used to have a huge garden, and made a compost by barely dumping anything organic on a big pile. It worksed great: low maintenance, never one problem, plenty of compost produced.

I now live in a flat, and have composter using worms.

I can't make it practical for the love of me.

If I put it inside, I get flies, no matter how much carbonated material I put in it.

So I put it outside, but it dies from either heat, or cold.

Even when I manage to avoid the worm genocide, the process is so slow. It does produce a fantastic liquid fertilizer in mass, but the volum of organic matter consummed is nowhere close to what I need. I eat a lot of vegies and fruits, and in 2 weeks, I have to put most of it in the trash can, waiting for the worms to work on the legacy pile of trash.

Any advice ?

Definitely! First off, I am biased in my compost preferences and it is always good to hear different opinions. Second, what works for one fails for another, so no sense in throwing more effort into a process that isn't working.

On to your situation. In home composting is always tricky. My wife and I do indoor/outdoor because we have a yard. In your situation, I'd advise against vermicomposting for the same reasons you listed. I was the technical manager for an enterprise vermicompost startup. Our facility was a rowhouse basement and the owner of the property was a founder. Flies are unavoidable, as is the occasional vermischwitz. They are fragile guys and minor mistakes have major consequences. The exact ratios escape me, but 1lb of healthy, mature red wigglers can consume half a pound of Green (fresh) cellulose in a day or two. We were engineering soil additive, so we supplemented heavily with Gray (dead) cellulose (cardboard, paper, etc.) In an attempt to hit profitable output. I will spare you the further details of our failure. Suffice to say that vermiculture is tricky in the best circumstances.

My suggestion to you is twofold: use the Bokashi Method (https://www.planetnatural.com/composting-101/indoor-composti...) and find a friend with a garden to offload your Bokashi Tea. That link is one of the first hits on Google (read: not vetted) because if you go this route you are going to be reading a lot and there are many roads to Rome.

This is very short due to medium, but if you'd like to discuss in depth and at length I'd be more than happy to. Both composting and waste neutralization are passions of mine. Just let me know how to contact you should you desire.

I would love to be able to connect re composting. My email is my user name at gmail. Thanks.

Great, can you email me at lesametlemax on gmail ?

Your main point stands, but, for terraforming purposes, a decade is a very short time period.

In terms of exoplanetary interchange, time recedes and tonnage expands as the concern. Without a space elevator or MAC transfer solution, you'd be effectively FedExing teacups of dirt to a green house in a Sahara contained on Antarctica... on Mars.

It's also a matter of federal budgeting; when the rovers landed, NASA definitely did not have enough funds budgeted to keep tasking the rovers for 8,000 days. But as the rovers continued to operate, NASA requested more funds to continue, and Congress provided them.

Once a robotic mission accomplishes its primary objective, NASA has a great story to tell about marginal cost vs return of additional funds. It's not that uncommon; heck they still get funding to keep the Voyager program operating.

Yes. For more:

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

and:

https://www.jpl.nasa.gov/missions/mer/images.cfm?id=1823

The second link shows about 18% power drop after 90 sols ~= 90 days.

Then that was the mother of all under-promises!

If they make it over 9000 days, they can state the just did a error on the exponent

Then the units would be time squared. That doesn’t make any sense!

Nah, just use scientific notation: 9e1 vs 9e3 days.

Oh, OK

Just a humorous take... This is SOP for government projects. "No worries, it'll be complete in 90 days." 8,000 days later....

The 90 days was, as mentioned, a lower bound, not an upper bound. They were planned to last at least 90 days.

Framed this way it's not entirely surprising they lasted longer, especially given the fortunate turn of events with the martian wind cleaning the solar panels of dust.

This of course does not in any way diminish the engineering and ingenuity involved with designing and operating the rovers.

^ this guy drives an accord

I’ve never found it impressive, it’s a horrible padded estimate meant to cover NASA and make them look good. A real estimate needs to be useful and somewhat close to the actual outcome. There’s no way they were this off without it being on purpose.

Imagine if your retirement planner or accountant was off by 44X in their number crunches for you. Would you say they were good at estimating?

Jesus dude, can’t you enjoy a crowning achievement of human and American excellence?

You think NASA is overly concerned about public perception to the tune 44x an original estimate?

Crazy sometimes human could be. But let us celebrate instead.

I celebrate the achievement just fine, it’s the horrible estimate that is shoved in my face constantly I have issues with. It represents a dumbing down of science for the masses that is just getting worse year after year that I resent. The rover isn’t good enough as is, we are told to be impressed by a nearly 50-fold wrong estimate that is all anyone knows about the rover. I knew it would be one of the top comments on this post. As a scientist I appreciate precision and good science, not hero worship and bureaucratic PR estimates.

Is it not acceptable that they were conservative in their predictions about the effects of Martian dust? I imagine they didn't have a whole lot of data points to inform their estimate.

I don't disagree with your overall point, but if you care about accuracy, you shouldn't also conflate an engineering goal with an estimation. I couldn't find the original estimation, but considering the unknowns we're talking about and the sunk costs, surely a 4-8x safety margin (1-2 years estimation) is not unreasonable. So we're at most talking about a 15-fold wrong estimation, not 50.

Another comment explained this. They set a goal of 90 days to get the data they needed. When they presented the proposal for budgetary approval, 90 days of data collection was considered worthwhile enough to greenlight the project. So they accomplished it. However once you already have a working rover on the surface you might as well operate it as long as it lasts because the marginal costs are minimal compared to the initial costs. So they kept requesting additional funding to keep the program running.

I see nothing wrong with this.

No one predicted that the rover would break on day 91. They designed it to last at least 90 days. Between standard rocket-science safty factors and preparing for the pessimistic end of all the uncertain risks, there is much room to go beyond your original mission.

Imagine you design a bridge to support 100 tons. No one would accuse you of mispredicting if it holds up at 200 tons. They will accuse you of negligence if it collapses at 105 tons.

If my retirement planner predicted that I had 90 days to live and I ended up living for another 15 years I imagine I'd be quite pleased.

It was never an estimate. It was the minimum plan.