Why This Mars Rover Has Lasted 3,560 Days Longer Than Expected | Hacker News Reader

Why This Mars Rover Has Lasted 3,560 Days Longer Than Expected

sfgate.com

113 pointscryptoz57 comments

What an awful article. All factual information can be summarized in two sentences:

"The mars rover lasted longer than expected because we did not anticipate the strong winds blowing dust off the solar panels. Furthermore, the rocker-bogie suspension system kept the machine from toppling over."

Everything else was non-information towards the subject. Claims that their machines are American, exquisite, sturdy, creative, well designed, excellent, fantastic and well built explain nothing.

Tough criticism.

It was designed for a few months. On its 10-year anniversary, I think it's worthy of a news article.

Sure... not much to say, but then the article wasn't very long, either. A couple facts, and the rest was filled with pride and other light reporting, like the feature article that it is. I don't know much about journalism, but it wouldn't seem out of place next to a feature about a successful restauranteur or a dog that saved someone from drowning.

Is hard-hitting and just-the-facts the only acceptable form of journalism or something?

To me, this was at least worth the bits it was printed with. I'm glad to know that people still care enough to do something so well that they can launch it on a rocket, land it on another planet, and keep it running and sending data back to Earth for 10 years straight.

> It was designed for a few months.

It was designed to, at an absolute minimum, with everything they could reasonably expect, work perfectly for 90 days. This is a bit of a different statement.

A strange point: so you're saying that because they designed it to last longer, that lasting longer is not as much of an accomplishment? Like the optimism of the engineers detracts from the results when that optimism is realized?

Regardless, it doesn't take away from my point. Getting something to work in that environment for a decade (I suppose I should specify that it's an Earth decade) is quite an accomplishment.

Let a few people feel a little pride of accomplishment. And let the rest of us feel inspired that such a feat is possible.

Not too surprising. I was going to click it, then saw it was by sfgate. When was the last time they published scientific research in an informative way?

Always love this xkcd: http://xkcd.com/695/

Why does anthropomorphizing our machines always seem to lead to sadness?

There's an alternate take on the xkcd panel with a more upbeat viewpoint. I looked for it briefly but couldn't find it.

Found it http://twicsy.com/i/vTY5E

Bingo. Thanks.

Other alternative, oposite point of view: http://forums.xkcd.com/viewtopic.php?f=2&t=22741&start=4680#...

Oh man. If that isn't a timesink.

Because we treat them like slaves.

The word robot means slave.

Almost. It derives from robota serf or corvee labour. While either of those are certainly slave like, they are distinct from chattel slavery (which is what most people think of when you say slave). The key difference is that chattel slavery has the slaves as actual property. Serfdom implies a responsibility from the master to the serf, beyond that implied by chattel slavery. Corvee labour is essentially a tax in the form of labour.

I initially thought this was just going to be another pedantic argument but it turned out to be really interesting. So, thanks?

Actually it just means work.

Because we treat them like machines.

Reminds me of WALL-E

//slightly off topic

The rock, which has a white rim and center that is deep red, is unlike anything scientists have seen before on Mars.

Such statements insinuating that the scientists are somehow puzzled by that fact. Of course they haven't seen it before, because it never happened. How often do you get a rover on mars. Sooner or later a rock will get yanked from the wheels. Media is really scraping the bottom here.

edit: My guess, looks more like something metallic, like a part of the rower perhaps.

It sounds so much more dramatic if you say "Never in the history of Mars exploration has this ever happened!" welcome to breathless journalism. That said, if it is a jelly donut then someone is going to have some explaining to do :-)

You are being overly cynic. It is a new discovery.

Exactly.

In other news, scientists astounded the world with a new discovery on mars: a small, hexagonal rock with a threaded cylindrical hole.

Slam-dunk proof Mars had intelligent life at some point!

Even more amazing: it just happened to be sitting near the rover. And it wasn't there when the rover drove up to that point.

Proof the martians are just putting evidence right in front of us, while being sneaky at the same time! How clever of them.

I was always a little dubious of the 90 day mission length. It sounded like NASA were just giving themselves a "yeah we meant that to happen" get out if things went bad.

You need to keep in mind that the rovers also needed to survive the six months in transit in once piece, and the airbag-assisted landing in one piece, both of which add necessary reliability and durability factors to make sure that the thing lasts even the 90 days. Also, you need to keep in mind that those cleaning events came out of left field. Yeah, they were expecting to maybe get a few extra weeks out of the rovers, but because the wind behaved in a manner that was completely unexpected, the panels weren't as dust-caked as was expected, which had always been the real limiting factor.

  > the panels weren't as dust-caked as was expected

Common, this is such a lousy excuse. If dust was a big problem I'm pretty sure they would have designed a dust-cleaning device.

Such as what, exactly? I'm sure it was considered, but it's not that simple. Would it involve air blasts? Liquid? A squeegee? The sand/dust content wasn't exactly known so a scraping motion would potentially scratch the glass surface and make things worse. Compressed air would take a tremendous amount of energy to compress and a liquid is heavy.

It was considered. Weight is a huge factor, and when the features were triaged it was dropped.

Or maybe they overspent and built something too good. They could have had a much smaller budget if they really wanted a 90 day mission.

Keep in mind that the lifetime of this mission is actually pushing other missions out of scope. The budget needed to continue support of Opportunity could be dollars spent elsewhere. So, overspending doesn't make sense. They're roughy, $128 million (combined for both rovers) over their operating budget [0].

[0] http://en.wikipedia.org/wiki/Mars_Exploration_Rover

It was $820M to send the rovers to Mars, and $128M to extend that by 10+ years; the main cost was in getting them there intact. Yes this money displaces spending on other missions, and there are diminishing returns for applying the same tools to the same small part of Mars, but mostly these years have just been bonus.

> Why This Mars Rover Has Lasted Longer Than Expected

It hasn't, it was designed to last this long.

But politically it's better to say it's designed for a short period in case things go wrong.

I'm sure the cost of getting the thing there overshadows the cost of construction, but it seems really fabulously overbuilt. I'm sure the martian environment is horrible for plastics, but heck, they only need to last 90 days. Let them get brittle and rot. (I'm assuming plastic is lighter and therefore cheaper to transport)

I guess it's the difference between doing everything as well as possible and then dealing with the limiting factor, vs doing just enough to meet the 90 day goal. A classic quality vs value trade off.

The hard part is defining what "just enough to meet the 90 day goal" is.

The rovers were things that had to be built to survive a flight to Mars, go though the atmosphere, land by bouncing on some airbags, and then go drive around in a very hostile environment - which needs a pretty sturdy robot.

Over-engineering the robot itself is also a form of insurance. Launching a rocket is hard. Landing a lander with airbags is tricky and very unpredictable. The time from finishing to build a robot and it landing on Mars is a really long time. Funding to pay for all this comes once or twice in a generation. Given all those possible points of failure in things that have nothing to do with the rover itself, why not engineer the rovers to be as resilient as possible to try and make up for some of the possible risks that could happen elsewhere in the project?

Put another way - if you're building a race car and the body costs $1000, why not spend $20 (as opposed to $5) on the engine to make sure that of all the things that could break, the engine won't be it.

If you absolutely want a 99+% chance it would last the first 90 days, then really the chances are that's it's going to last a lot longer. I actually don't think it's a typical quality tradeoff. Because you have to meet a high initial reliability, you end up with higher probability of hitting a endurance beyond your minimal requirement for the device.

I'm also guessing w.r.t the structures, you'll have to build to withstand a worst case environment during the launch and landing. Once you've survived those you're "overbuilt" with respect to rolling around the surface, but the strength was needed to survive getting there.

"fabulously overbuilt"... not quite, think about this: you have one chance to launch it into space, send it millions of miles away, reenter atmosphere safely and then survive a hostile environment for only (?!?!?) 90 days! There is no overbuilding, by the nature of the mission itself your Factors of Safety [1] are through the roof!!

[1] http://en.wikipedia.org/wiki/Factor_of_safety

Amusingly, in this case there were two chances.

I'm saying if Spirit and Opportunity were constructed less well, perhaps we could have afforded a third with the same budget.

I don't really have any honest way to value the difference between 10 years of extra data vs another location (i'd guess the more data is better, but that's hindsight bias)

Remember, every ounce of "factor of safety" is tens (or hundreds?) of thousands of dollars.

There's also the fact that the cost of design is quite likely a lot higher than the cost of construction. For one-off (or two-off) prototypes it's often worthwhile to overbuild so that the design's easier. Optimising for cost and exact lifetime is expensive.

Recognizing the premium for any mass on the mission, I've wondered why some sort of panel-duster (perhaps a spinning flail such as you'd find in a carwash) wasn't designed in.

Still, the longevity and data have been truly amazing.

You'd still have the same problem though: how do you keep a motor lubricated in the partial pressure of a pretty good vacuum? How do you stop the duster's brushes falling out or wearing out?

Not saying it's not possible, but if you consider things like reaction wheel failures, there's a lot of "might last a long time, might die in a year or two" problems in any motor in space.

There are plenty of other motors on the landers, presumably the rovers haven't been telekinesing merrily over the marscape. The brushes would be a far lower duty-cycle item.

Depending on the brush design: something sufficient to clear the panels periodically (say, based on noted solar degradation and/or self-inspection photography) might work. Clearly, incidental wind has proven sufficient.

And there's a lot of engineering work on Earth which could go into testing out the design. As I said: it's just curious that this seems to have been left to chance. Though nobody really expected the rovers to run as long as they have.

> There are plenty of other motors on the landers, presumably the rovers haven't been telekinesing merrily over the marscape. The brushes would be a far lower duty-cycle item.

But far more exposed, though. If not the motors themselves, then the mechanics of the wiper arm.

More so than camera servos or other arms on the probes?

Again: while high-reliability, low-pressure lubrication is a challenge, it's clearly not an insurmountable one. Providing a means to ensure your only source of energy remains viable would be useful. Failure of the dusting system of itself wouldn't result in a mission failure. Success of it could prolong it, and/or allow for smaller collectors (as I recall the solar panels were oversized to allow for possible dust accumulation).

I'm thinking of some sort of rotary duster which might pass horizontally above the panels. Less a windshield wiper design than a dual-arm rocker.

I've read the sand is very abrasive, and any kind of "wiping" would have scratched the panels.

should have gone all DSLR self cleaning sensor and vibrated and tilted

That was another thought I had.

As for abrasive sand, if you're whisking it off lightly enough, you should avoid most problems. Coatings might address that. And you've got the choice (addressing grandparent): unscratched PV under mountains of sand, or lightly scratched PV exposed to sunlight?

Google is your friend --

http://www.newscientist.com/blogs/shortsharpscience/2008/11/...

Thanks. And in a Connections moment, that's Henry Spencer of "Those who fail to understand UNIX" fame.

I didn't realize the Rover program was so rushed.

Yes, you know, I had forgotten about the rush, and I worked at JPL during that time. Obviously not on the Mars rovers, or I would have remembered ;-). Curiousity, the current rover, missed its original launch window and had to be delayed -- but by then, our confidence in ability to land spacecraft on Mars had been restored, so there was patience.

The really funny thing is, after an article patiently explaining the constraints (need to manage risk after losing two Mars-bound spacecraft, time line for Earth-Mars launch window, weight on the airbag lander) there are still people (see the comments in the article) who want to solve the problem with the methods dismissed in the article above their comment.

E.g., use "a can of compressed air" ("put it in a hollow leg, it would add no more than a few grams"), or a "beater bar" or "squirrel fan". It's like bikeshedding, but by people who don't get that it is a spacecraft and not a bike shed.

It's one thing to ask a question ("Why didn't they...") and another thing to assume the engineers missed something obvious ("Fools! They should have...").

When the U.S. puts as much money into something as the mars rover (that isn't for killing or spying), red-blooded 'mericans want to know they're getting their money's worth. If you design a rover that's supposed to work for 10 years and it dies in 10 months, you F'd up. However, if you publish a Scotty estimate of just 6 months and it lasts 10 years...

I don't buy, even for a moment, that the mars rover wasn't designed to last. It was indeed well designed and built. The estimate for it's expected lifetime, however, was more a function of politics than engineering.

Is it too much to ask that a journalist be able to convert 10 years to days correctly? I mean, it's true that it lasted 3560 days more, but I'm guessing they were shooting for 3,650 days. (or 3653 days)

it's 10 years minus 90 days (the original anticipated mission length). 3650 - 90 = 3560 days.

Thanks; my own reading comprehension failed me. I hate when I fall victim to Muphry's Law. [1]

1. http://en.wikipedia.org/wiki/Muphry's_law

I find it more likely that the year value was rounded and the day value, more precise.

This is why is seems so terribly short-sighted to me that our most recent rover is nuclear-powered.

Using RTGs for power gives the rover a ton more flexibility though, and helps make the most use out of the time that it has. I remember from reading the Mars and Me blog (written by one of the rover drivers) that the cyclical nature of solar power put some pretty big constraints on the rovers. It wasn't just a day/night thing, they had to be really careful about the orientation of the rovers' solar panels during the martian winters. A lot of the time they'd just park the rovers for the winter. Plus, the waste heat from the RTGs can be used instead of electric heaters.

I suspect there is an upper limit on the amount of solar energy that can used to power a rover at any given point in time. I'm assuming the nuclear option provides a much stronger power source.

Reply on news.ycombinator.com