Got doubts about Curiosity surviving Mars landing

[jackmell]

That landing manouever sounds too complicated, too susciptable to errors. But surely they know more than me and feel it's the best way. Still, I think it will fail. That's my prediction. Wasn't there a less error-prone way of setting down on the surface? Maybe though it was a budget issue.

 

[marcus]

Did you watch the YouTube "Seven minutes of terror" ? It's great.
Of course there is a substantial chance of failure.

I could be wrong but personally I think you are mistaken about the landing method being chosen for budget reasons when there was some obviously better way.

The way I see it there is no obviously better way. They wanted to land a 1 ton vehicle on a layer of dusty soil in rather strong gravity (considerably stronger than Moon surface gravity).
Landing with rockets all the way to ground would have raised a lot of dust and made craters in the surface you are trying to land on.

Maybe they could have encased the vehicle in a shell to protect against dust and flying debris, and landed it in the shell, sort of like a big egg with rockets and legs. And then once the egg had landed open the belly bay door and lower the vehicle to ground.

But that, to me, seems inelegant compared with what they are doing. It is innovative but UNTRIED. If it succeeds it may turn out to be the method of choice in future. If it fails, well it fails and we've learned something. New technologies often look risky and impractical. You have to try them and allow for some chance of failure.

That's just how I see it and I may be wrong, of course.

 

[jackmell]

I've seen it. I believe the cabling suspending the rover right before landing will get caught up in the rocket exhaust causing the rover to get entangled in the cabling and causing a bad landing.

 

[marcus]

So you've seen the YouTube. Great! Let's share the URL:


I'd like to hear what other people estimate the odds of success are.

I think the probability of successful landing is better than 60%.
It's risky and without being at all expert or knowledgeable I'd put it at 70%.

Anybody else have a guess?

 

[jackmell]

So you've seen the YouTube. Great! Let's share the URL:


I'd like to hear what other people estimate the odds of success are.

I think the probability of successful landing is better than 60%.
It's risky and without being at all expert or knowledgeable I'd put it at 70%.

Anybody else have a guess?

70% ! Maybe then I don't understand. Personally I'd give it 1 out of 10 and that's only because I have great respect for Carl Sagan and would not have wanted him thinking poorly of me. Surely though the mission crew have worked out the success probability very precisely. Does anyone know what their figures are? They may not want to admit it but I suspect they are privately thinking one out of three.

Oh yeah, I predict it will land but the first picture will be mountians in the background and in the foreground, a broken wheel entangled with a bunch of cabling.

 

[Chronos]

I'm unreasonably confident the landing will be successful. NASA has a great track record and I doubt they allowed for a 30% chance of landing failure in mission planning.

 

[marcus]

I'm unreasonably confident the landing will be successful. NASA has a great track record and I doubt they allowed for a 30% chance of landing failure in mission planning.

So you estimate the odds as BETTER than 70% chance of success. Great! Glad to hear it!
I'm really curious to know what probabilities of success other people reckon for the mission, at this point.

 

[marcus]

Chronos, I guess that means if we were betting you would put up $7 against Jack's $3.
It would be a fair bet because you could expect to gain $2.1 (.7 times 3) at a risk of $2.1 (.3 times 7).
Actually that would seem an advantageous bet because you judge the probability of success to be more than .7, so your expected gain is more than $2.1 and your risk is less than that.

I've been listening to Jim Hartle's concept of generalized quantum mechanics in which he reformulates QM in terms of betting odds based on "consistent histories" partitions of the different paths the world could take. He does away with the classical outside observer and boils it down to dividing the possible histories up according to what we care about and feel the need to bet on.

 

[RichZitola]

I don't have a good guess on the probability of success, but the biggest risk in my mind is the dynamics of the lander/rover two-body system. After watching the full scale drop test, you can see there is a bounce and oscillation:

http://www.youtube.com/watch?&v=YasCQRAWRwU

How will the lander handle that?!? Without feedback from the rover, the lander might end up amplifying that oscillation. I think the mass of the rover is comparable to that of the lander, so the possibility of uncontrolled oscillations seems obvious. Although cable entanglement seems very unlikely, I could envision the rover going through wild swings in orientation that might be enough to capsize it on touchdown.

Part of me says, well, of *course* they must have thought of that, and programmed the lander to dampen those oscillations, but that's got to be a really tough problem to solve.

If the landing goes badly, my nickel is on this scenario... just imagine the rover on the surface, intact, but inverted... wheels to the sky! oops!

Fingers crossed... :)

 

[Drakkith]

My confidence level is over 9000!

 

[CosmicEye]

It does look complicated enough with the influence of how dramatic they were saying it.

 

[Waterfox]

Nearly every online article and video I see talking about EDL have comments from people who are uncertain about the complexity of the landing. You are definitely not alone.
Check out the previous "Six Minutes of Terror" of video
Back then many people were unsure it would work, but it work flawlessly twice!

I feel confident in the MSL landing.

 

[drrail]

How can we give odds on an untested system? We can only speculate on certain aspects of the plan as parts of that plan. Break it down into the most probable failure points, and weigh those points against known facts.

The weight of the rover should aid in the success of the crane system, however depending on how the cables are 'rolled up' in their casing, that will determine the likely hood of the cables getting caught up or entangled on themselves inside the casing. One cable getting ‘caught up’ would more than likely cause complete failure. I've seen the test of the crane lowering the rover, but I'm not sure of the internal engineering with regards to the specifics of the crane cable housing mechanism.

http://www.quora.com/Did-NASA-do-a-dry-run-of-the-Curiosity-rover-landing-sequence-here-on-earth

Another failure point of importance to mention would be the parachute deployment. If the rover is unlucky enough to get into a spin or roll on entry, the parachute would have the chance of entangling itself around the rover/craft and dropping like a lead balloon. Then the Crane really doesn’t matter.

 

[jackmell]

I'm unreasonably confident the landing will be successful. NASA has a great track record and I doubt they allowed for a 30% chance of landing failure in mission planning.

Unless their budget was cut which is sometimes the case. I suspect they would have preferred an apollo-type landing but because of budget constraints, elected to dangle precariously the rover from the business end of powerful rocket engines. Not unreasonable to suspect it was "either we do it this dangerous way within budget or we don't have a mission," and then just accept the high probability of failure and then don't really admit it to the public.

 

[Drakkith]

Unless their budget was cut which is sometimes the case. I suspect they would have preferred an apollo-type landing but because of budget constraints, elected to dangle precariously the rover from the business end of powerful rocket engines.

An apollo type landing is not a good option.

Not unreasonable to suspect it was "either we do it this dangerous way within budget or we don't have a mission," and then just accept the high probability of failure and then don't really admit it to the public.

Oh please, go spread this BS somewhere else. It is 100% unreasonable to suspect something like this. NASA isn't going to waste millions of dollars on a mission with a high risk of failure just because they couldn't afford the "good" way of doing it. They either find a way of doing it within budget and with a reasonable chance of success or they don't do it at all. (Or they end up overrunning their budget and getting the money anyways by cutting other programs)

 

[jackmell]

NASA isn't going to waste millions of dollars on a mission with a high risk of failure just because they couldn't afford the "good" way of doing it.

I disagree with that. It's always a matter of money and practicality. And please don't provoke the mentors into locking my thread until at least I've been given the opportunity, in the unlikely event this succeeds, of admitting I was wrong.

Can you explain why the apollo-type landing is unreasonable?

 

[256bits]

I disagree with that. It's always a matter of money and practicality. And please don't provoke the mentors into locking my thread until at least I've been given the opportunity, in the unlikely event this succeeds, of admitting I was wrong.

Can you explain why the apollo-type landing is unreasonable?

Well, if you accept the fact that they did not want rocket engines to kick up dust from the surface, then the tether system was judged by them to be the way to do it.

For the Apollo landings on the moon, the command module was set into orbit around the moon. To slow the curiosity mission to an orbit around Mars would require more fuel to be brought along. Sending down Curiosity to the surface from Martian orbit with rockets would also require more fuel. To do it the apollo way would then entail a larger initial spacecraft right from get go at the launch pad on earth.

Then the question is whether or not that type of landing ( Apollo style ) has actually a greater chance of success than the tether system. Since Martian landing missions have a success rate of around 30-40% , would have substituting one complex expensive system for another complex sytem, have inreased the odds to 100% suuceesful mission. No.

I think the heat shield, parachute, tether system looks quite promising will be successful.

 

[joema]

...Can you explain why the apollo-type landing is unreasonable?

If by "Apollo-style landing", you mean a single-body, legged lander using powered descent to the surface, it's not unreasonable. In fact Viking landers 1 and 2 did this in 1976, and they weighed about 600 kg, vs Curiosity's 900 kg. So they were in the same approximate mass range. It was technically feasible to do this for Curiosity.

However, Viking landers were fixed and non-mobile. This facilitated a combined lander and descent stage. When landing a wheeled rover/lander, it's different. Regardless of whether you use a two-body "Skycrane" method, or a single-body Viking method, the wheeled lander and descent stage must eventually separate.

The options are:

(1) Use airbags for final descent. This works for small landers, but for heavy landers like Curiosity the airbag system is just too heavy.

(2) Put the a legged descent stage under the wheeled lander, land as a single body, deploy ramps and let it drive off. This necessitates tricky engine cutoff procedures, redundant landing gear (legs plus rover wheels) and entails risk of blocked ramps due to terrain. A blocked ramp nearly prevented deploying the rover on Mars Pathfinder.

(3) Suspend it from cables (Skycrane method). This avoids redundant landing gear, avoids covering the lander with dust, avoids tricky engine cutoff and touchdown sensing, and allows very gentle powered landing. The tradeoff is added complexity of the Skycrane system, but no matter how landing is done it entails a lot of pyrotechnics and sequencers working perfectly. The Skycrane system incrementally adds complexity to gain certain benefits, but an "Apollo/Viking"-style landing is not free of complexity or perils.

For a detailed historical review of landing options, see "The Challenges of Landing on Mars": http://www.engineeringchallenges.org/cms/7126/7622.aspx

 

[singapura]

Well it worked. Case closed.

 

[Chronos]

Score one for the optimistis - http://gizmodo.com/5932039/live-cov...landing-from-nasas-jet-propulsion-laboratory/

 

[jackmell]

Ok. I was wrong. Nice going guys. Score another for Carl Sagan (I mean I think lots of the guys working on the program grew up admiring him).

 

[litup]

Ok. I was wrong. Nice going guys. Score another for Carl Sagan (I mean I think lots of the guys working on the program grew up admiring him).

As an ex Apollo tech (Apollo tracking and timing), I am QUITE pleased you are wrong! The reason they had to do that multi-stage technique was due to the fact the Martian atmosphere is 100 times less dense than Earth's, so if you had something like the Space Shuttle try to land on Mars the way it lands on Earth it would just crash into the surface doing a couple thousand miles an hour, not a great way to land:) So they had to use the heat shield which took off about 12,000 mph, still leaving 1000 mph left (It would have hit the ground at about 1000 mph if left to just the heat shield) so a supersonic parachute was next, followed by jettisoning the heat shield, then at about the same time, starting the retro rockets and releasing the parachute, then as the probe got lower, in order not to cover the solar panels with dust and thus lower the amount of power generated by the sun, the cables lowered the probe to the ground, the cable disconnects, and the rocket exits stage left.

All done under local computer control because the time delay was 14 minutes for the radio signals, 28 minutes round trip for those signals to get a handshake so controlling the descent from Earth was impossible. The increase of computer power and software over the last 50 years has been incredible, computers on board tens of thousands of times faster and stronger than the old Apollo days. That is what made the difference, what allowed all those separate steps to happen under local control. Without it we would be stuck with probes weighing only a couple hundred pounds max.

Also, this technique can be upgraded to something more like 10 or 20 tons with larger parachutes and higher thrust rockets and so forth for human landings. They can be actual pilots if the computers die on the way down.

 

[sophiecentaur]

I reckon that the level of computer control, these days, must be orders of magnitude better than it ever was in the past. Rugged / reliable systems, suitable for that sort of application will be quite a few years behind their games-console counterpart, simply because they will have been chosen and developed over a really long timescale. This is the same problem that weapons engineers have - most of the operational gear is ancient by home computer standards.
But, even if you go back ten years, the available computing power was still pretty nifty. I would be prepared to put my money (if not my life) in the charge of the sort of computers that they are likely to have on board the curiosity vehicle, whereas I definitely wouldn't trust an autonomous system to have landed the space shuttle reliably.
That tether arrangement may not strike people, intuitively, as a good solution but how many of us have done any of the sums?

 

[jackmell]

Ok, that all makes sense now to me guys: no apollo landing. I'm still amazed that it worked. The mission members are tops in my opinion now for succeeding and I admire their competence.

 

[sophiecentaur]

All this makes it painfully obvious just how much more of a big deal it would be to do a manned mission. Just how many tonnes of payload would you need to get onto the surface so that your crew could get back up again?
It would require yet another landing solution, for a start.

 

[marcus]

All this makes it painfully obvious just how much more of a big deal it would be to do a manned mission. Just how many tonnes of payload would you need to get onto the surface so that your crew could get back up again?
It would require yet another landing solution, for a start.

I'm not sure why one would consider sending meat to Mars. Robots seem to be doing a pretty good job so far.

My candidate for a manned mission (if there is to be one) would be Ceres. Outer layers appear to be mostly water ice. Chemical sampling at various depths in the ice should be fascinating. One could hollow out places to stay that could hold breathable air at the right pressure... One would not have to bring all one's housing. And relatively easy to land and take off in Ceres gravity. A 1000 km diameter dwarf planet in the asteroid belt...

If we have to send humans somewhere in the solar system, I'd vote for it being Ceres at least at first.

 

[drrail]

Wouldn't be a bad place to start. We'll get more info in February 2015 Dawn ( spacecraft )

http://www.nasa.gov/mission_pages/dawn/main/index.html

 

[Waterfox]

I would be prepared to put my money (if not my life) in the charge of the sort of computers that they are likely to have on board the curiosity vehicle, whereas I definitely wouldn't trust an autonomous system to have landed the space shuttle reliably.

Didn't the Soviet Buran land autonomously?

 

[sophiecentaur]

I'm not sure why one would consider sending meat to Mars. Robots seem to be doing a pretty good job so far.

I agree. Desire for manned space missions just strikes me too much like shootem-up wild west expeditions. They're just such bad value in terms of information per buck.

Didn't the Soviet Buran land autonomously?

One swallow doesn't make a summer, though. It would only be after a lot of successes that a system could be said to have proved itself. As intelligence (machine) improves, the success rate should improve in leaps and bounds.

 

[twofish-quant]

I agree. Desire for manned space missions just strikes me too much like shootem-up wild west expeditions. They're just such bad value in terms of information per buck.

But done right, they are good value in getting more bucks. I'm a very strong proponent of manned space flight. The question is goals and strategy. I think the big priorities would be to get together a LEO infrastructure, then get to the moon, then get to Mars. One reason LEO is important is that it's a laboratory for how the human body can respond in space.

The problem with manned space flight is that if we mess up with unmanned probes, we don't have to notify the next of kin, so it's possible send up a robot, watch it crash, figure out what went wrong, repeat, in a way that would be totally unacceptable with people.

One swallow doesn't make a summer, though. It would only be after a lot of successes that a system could be said to have proved itself. As intelligence (machine) improves, the success rate should improve in leaps and bounds.

The other thing is that thick atmospheres make it easier for autonomous systems. Thick atmospheres create a negative feedback which means that your computer can be slightly off, and not cause a crash. Autopiloting/autolanding aircraft is pretty mature technology.