Why the sky crane isn’t the future for Mars landings

Sky crane delivers Curiosity to Martian surface in artist’s conception. (Credit: NASA/JPL-Caltech)

Given all the attention that it is receiving, the innovative technology that will place the Curiosity rover on Mars — the sky crane — may seem like something that we’ll be seeing much more of during future space missions. Yet it’s not. In fact, there’s good reason to suspect that it will be a long time before the sky crane is used again on Mars, if ever.

Of course, its prospects do depend on the success or failure of the Curiosity landing, but let’s hopefully assume the best. [Update, 1:36 a.m. EDT, Mon.: The best occurs! Success!] Instinctive skepticism has always greeted the plan: it is complicated and unorthodox, and a mishap anywhere along the chain of feats in involves leads to disaster. Even those of us enthusiastic about the sky crane have often conceded that it sounds crazy but might just be crazy enough to work. Even that skepticism, though, isn’t exactly why the sky crane won’t be selected for many other missions.

The absence of the sky crane might seem all the more surprising given that NASA’s rationale for using it with Curiosity has always been that it had no good alternatives. As I explained in my SmartPlanet column about it, technologies used to land other probes on Mars hit their limits with something the size and weight of the Curiosity rover. Parachutes can’t slow the craft enough in the thin Martian atmosphere for a soft landing. Airbags can absorb the force of a landing impact for the 400-lb. Spirit and Opportunity rovers but not something as big and sensitive as one-ton Curiosity. Rocket thrusters, the old reliable standby, can do it but at the cost of disrupting and polluting the landing site.

So then why wouldn’t the sky crane be the method of choice for upcoming probes? Because the sky crane is an expensive technology developed by NASA, and NASA is temporarily getting out of the Mars lander business.

Artist’s conception of the MAVEN orbiter. (Credit: NASA)

NASA expects to get years of good results out of Curiosity, so it won’t be idle on the Mars front. Nevertheless, the only mission concretely on its schedule now is the Mars Atmosphere and Volatile Evolution (MAVEN) orbiter, set to launch in 2013, which will study the planet’s upper atmosphere. It won’t be sending anything to the surface at all. NASA has long-term Mars exploration plans that would repeatedly take it to the surface — with balloons, aircraft, deep-drilling probes, more rovers, and even rockets capable of returning samples to Earth — but none of those has been scheduled or funded yet, and the cloudy condition of the economy makes it unclear when they will be.

Originally, NASA was planning to participate in the ExoMars program, a pair of joint lander missions with the European Space Agency and Russia (which was initially a minor partner) scheduled to launch in 2016 and 2018. ExoMars has been battered with financial problems and shifting plans throughout its history, and those only got worse early this past February when NASA confirmed that it was dropping out of the project because of budget constraints.

The two ExoMars missions never needed a sky crane. The one launching in 2016 was always planned to deliver a relatively small stationary instrument package. The 2018 launch will be sending a rover, but one that’s only about 200 lbs. heavier than the Spirit and Opportunity rovers. That much extra mass doesn’t justify all the complex contrivances of the sky crane, though it probably does exceed the tolerances of what airbags alone could handle.

Scale model of the proposed ExoMars landing-stage vehicle, showing the three sets of thrusters that will slow and guide its descent. (Credit: Anatoly Zak/RussianSpaceWeb.com)

The ExoMars partners have therefore developed an entry, descent, and landing system of their own. Like the Mars Science Laboratory and Curiosity, after using its heat shield to slow atmospheric entry, the ExoMars Entry, Descent and Landing Demonstrator Module will release a parachute for further deceleration. After cutting loose from the chute, the ExoMars vehicle will use small liquid thrusters to guide and slow its descent — much like the sky crane descent vehicle except that it goes all the way to the ground. That landing only qualifies as “semi-soft”: whereas Curiosity arrives on Mars with a kind of suave grace, like a tuxedoed James Bond arriving by jetpack, the ExoMars landers will hit the surface with a bit more of a thump and count on airbags to absorb the worst of it. Nevertheless, it’s a neat, elegant system that bears more than a little similarity to the arrival of a flying saucer.

The ExoMars vehicle will enter the atmosphere of Mars at a speed of about 21,000 kilometers per hour. (Credit: ESA)

The craft will deploy a parachute when it has slowed to a speed of Mach 2, and will drag it down to subsonic descent speeds. (Credit: ESA_

Three sets of hydrazine thrusters will allow the ExoMars vehicle to make a semi-soft landing on the Martian surface. (Credit: ESA)

(Sadly, Europe’s financial problems call into doubt whether ESA will be able to mount these missions at all. The ExoMars partners will need to collect commitments on all the needed money by November.)

So no other mission will use a sky crane through the end of this decade at the least. In theory, sky cranes could reappear after 2020 if Mars missions would involve payloads massing on the same order as Curiosity or larger. Any human missions would surely meet that weight requirement. Yet even for some heavy missions, a stepped-up version of the ExoMars system might turn out to be a preferable compromise.

Any decision to go with a sky crane rather than a thruster-assisted landing has to be made by assessing the risks of the more complicated procedure against the priority of not disrupting the landing site with rocket exhaust. For example, imagine that you are planning the early human missions to the Red Planet. If the lander will include rocket systems for returning the astronauts to orbit, will you care whether the sit is disturbed? Will the sky crane solution be able to scale up to handle something the size of a return vehicle? Would it be easier and safer to let the vehicle’s own rockets handle the landing?

Sky cranes may not disappear altogether from NASA’s plans, but they will always represent just one solution among a mix of several from which mission planners will choose. They may not be the future, but they will probably stay as just a part of it. And they surely are all-important to the stage of Mars exploration immediately ahead of us: much of what we hope to learn about Mars over this next decade will all depend on how well the sky crane delivering Curiosity a few hours from now performs.

Related Posts Plugin for WordPress, Blogger...
This entry was posted in Space, Technology. Bookmark the permalink.

33 Responses to Why the sky crane isn’t the future for Mars landings

  1. Pingback: Live-Blog zur Ankunft von Curiosity auf dem Mars « Skyweek Zwei Punkt Null

  2. Everyone’s talking about the crazy sky-crane like it’s an engineering mess but I don’t know if it’s that dangerous. Miguel San Martín, one of the engineers involved in the entry, descent and landing process gave a talk at our university. He was confident that this system was going to work. According to him, it was simpler than the airbag system used by Spirit and Opy and much safer. The physics of a pendulum is perfectly understood and it has a lot of negative feedbacks that work to make the whole thing more stable.
    Overall, the simulations they did told them that this mechanism is about 99,9% safe. Similar simulations done for Spirit and Opy gave “only” a 99,6% chance of success (or similar numbers, I don’t remember exactly).

    • John Rennie says:

      Good points, thank you. I don’t doubt that if the NASA/JPL engineers didn’t have extremely high confidence that the sky crane would work, they wouldn’t have relied on it it. Nevertheless, missions to Mars have an historically high rate of failures in practice. The physics and engineering of the sky crane may be straightforward, but the simple fact that the MSL’s plan for entry, descent, and landing has so many phases, each of which is completely dependent on its predecessor, means that it gets riskier than a method with fewer steps would be.

      • Mike says:

        The physics and engineering of the sky crane may be straightforward, but the simple fact that the MSL’s plan for entry, descent, and landing has so many phases, each of which is completely dependent on its predecessor, means that it gets riskier than a method with fewer steps would be.

        No. A method with “fewer steps” may easily still have a higher aggregate rate of failure than a method with more steps. Also, the failure rate of each method in a step doesn’t go up just because methods are chained together. We don’t really care about the number of steps, we care about the aggregate rate of failure. An example: If you had a choice to spend one minute in a cage with monkeys armed with deadly lasers, would you rather go into a cage with a single monkey that hit her target 50% of the time, or a cage with three monkeys that each hit their target 1% of the time?

        Also, you make it sound like there are many, many differences between the two methods, when in fact the two methods are nearly identical until you get to the end of the parachute phase.

        • John Rennie says:

          The methods are similar in their early phases, yes — which is why adding additional steps becomes riskier. Additional risks associated with the sky crane include: possibility that the thrusters of the descent platform will fail; possibility that the the tethers will fail to lower; possibility that the system miscalculate the descent speed and drops the rover onto the surface too hard; possibility that the explosive charges will fail to detach the tethers properly; possibility that a malfunctioning descent platform will fall onto the rover or damage it during its departure. Even when the odds of any of those problems occurring is extremely low, that’s a fair amount of extra risk to aggregate.

          All we can really say is that the total estimated risk for the sky crane seemed to be lower than for any alternative systems NASA considered. (The airbag systems wouldn’t have worked, so they weren’t contenders.)

          Now, for any mission payload that can’t be delivered in any less risky way than using a sky crane, sky cranes will by definition be what the engineers will favor. But theoretically, a future mission to deliver a smaller payload, like another Opportunity-size rover, could be delivered by sky crane rather than by airbags. I’ve never heard anything to suggest that would happen, though, because the aggregate risks for doing it by sky crane would be higher.

          But I’m not unyielding on this. If the great experience with the MSL’s landing gives NASA reasons to think it would want to use sky cranes as something other than a method of last resort, great. Then we could certainly expect to see it turn up again. For now, though, I think that’s a questionable call.

          • Mike says:

            I can only reiterate what Miguel San Martín and other members of the EML team have said.

            1) As pointed out earlier in the thread, Miguel San Martín, one of the EML engineers said that the sky crane was simpler than the airbag system, and cited the simulation numbers. Yet you keep saying that the sky crane was riskier, even though he says it was not.

            2) I think you’re forgetting that the “airbag system” relied upon retrorockets and altimiters to being the vehicle to a mid-air stop. It also had four airbags with six lobes each that had to be inflated with gas generators, deflated, held together with ropes, had small explosives to release the lander’s special nuts and bolts, accelerometers, motors, hinges, petals, and then the rover had to be driven off the lander. No wonder Miguel San Martín said the sky crane was simpler.


            3) At a news conf. Monday the EML team revealed that the sky crane had multiple backup systems, including for flight.

            4) If you toss a coin a hundred times and it comes up heads each time, there is a psychological tendency to believe that the odds of it coming up tails on the next toss are higher. But the odds are still 50/50, because the odds are independent of what came before. The same is true here. The odds of each step failing are independent of the odds of the previous step failing, and if the odds of each step failing are small enough then it could be safer in the aggregate, just as Miguel San Martín said it was according to the numbers that came out of the simulations.

          • John Rennie says:

            Fair enough, all good points.

            I will be curious, however, whether NASA will elect to use the sky crane again when it has an actual choice. As I said earlier, suppose it’s 2025 and we have reason to land another Mars rover no bigger than Opportunity. Will NASA use airbags again or will it rig a sky crane for a smaller payload? Would they make that choice on the basis of aggregate safety or cost? Or might the experience by then with something like the ExoMars EDL system argue for a completely different approach?

            My guess is that landing systems will continue to be a varied lot, and I won’t be surprised if sky cranes turn out to be disfavored more often than not. But we’ll see.

    • pgc says:

      So far as I know those percentages refer to a fully working system. They take in account only unexpected variations from nominal conditions (a gust of 3-sigma wind, landing on a large stone, etc.), not malfunctions of the system.

      The problem discussed here is different: how RELIABLE is the sky crane system respect to the airbag one? Curiosity demonstrated it was but this is not statistics but a single sample.

      I have no answer for that question – neither wish to suggest the sky-crane is unreliable – but just to make it clear.

  3. Sammy says:

    I was really interested in learning about the Sky Crane. Extremely fascinating science… congrats to humanity! :-)!

  4. Success! Very exciting, but will it translate through the media? There is a great deal of patriotic fervor, it will undoubtedly bolster funding efforts to follow through on a manned mission in 20 years. I do agree that the sky crane is risky, there must be a way to shorten those “7minutes of terror” but it is an important step, an opportunity to learn and improve. Terraforming, here we come!

  5. Stuart says:

    what happens to the sky crane and all the other junk used for the descent. Are we already poluting Mars?

  6. Andrew says:

    Why are scientists worried about getting too close to the ground with a propulsive descent? This IS A ROVER, which can move off to examine untouched terrain. Isn’t landing on Mars difficult enough without creating ‘risks’ that are actually pretty minor? Propulsive descent to the ground didn’t seem to phase the Viking landers or Phoenix. Actually in the case of Phoenix, it helped by exposing the ice directly beneath the lander. :)

    • John Rennie says:

      There are also issues of not wanting to risk having the swirling dust damage the exposed rover, as I understand it.

    • amphiox says:

      Also, there are still limits to how far and fast a rover can go, as they don’t travel very fast. You get more science time for your buck if you land on an undisturbed landing site, assuming that one of your missions is to examine an undisturbed site, as is the case here.

  7. Jeff T says:

    The Skycrane worked brilliantly. I have a few questions I wonder if anyone knows the answers to.

    Why was it decided that the skycrane woukld be f no further use when it had done its job? Why was it flown off to crash 200m away? Could itr not have been programmed to fly away and land itself? How much fuel was left in the Skycrane after it did its job

    ? Seeing that the cost to get each kilogram of this expedition to Mars maybe well over 1 million dollars – if there were several kilograms of fuel left which could be used in later missions to Mars – why was it sent away to crash?

    It might have been useful for another purpose in the eventual manned landing – or even for spare parts.

    • John Rennie says:

      Good questions, and I can only guess at the answers. It seems likely that there was very little fuel left in the platform’s thrusters after it delivered Curiosity to the ground. As you said, having extra fuel would have been expensive, and engineers could probably anticipate fairly exactly how much fuel they’d need for the job (plus just a little as a reserve). The platform wasn’t built to do anything else, so it’s job was over. They could in theory have built it to do more, but then they’d be adding more mass to the equation. Remember, the platform isn’t built for a soft landing: if it could do one, NASA might not have needed the sky crane maneuver in the first place. I don’t know how fast it was going when it hit the ground, so it’s possible that the platform’s value as a future source of spare parts (which is a somewhat fanciful justification for watching out for it) has not really been compromised by much.

      • Mike says:

        Some of this was covered in during the news conference on Monday. As for the remaining fuel, they reported that they had 140 kilograms of fuel left, out of 400 kilograms. The primary objective is to get the sky crane away from the rover, so as soon as it disconnected it flew away to the north. They actually ended up with much more fuel reserves than they expected, partially because they were over cautious, and partially because the vehicle didn’t have trouble finding a good landing spot.

        Packing extra fuel for a safe landing for the sky crane, and planning for a safe landing of the sky crane, would have added extra cost, and served no real purpose. The US isn’t planning a manned mission until sometime in the 2030s. Unlike the moon, Mars has extremely unforgiving weather that make any equipment left there over 20 years unusable.

        Also, a reasonable estimate to get one kilogram of fuel into space is ~$5000/kilogram, not one million dollars.

        • Jeff T says:

          Ok – so there is (was?) possibly 250kg of fuel sitting in a possibly unbroken canister close by? That’s a lot of fuel that could be put to a useful purpose at some stage I’m sure.

          If the skycrane is made from titanium – 20-30 years of even Martian weather would not make much difference to it.

          In regard to my $1M/kg figure – I calculated that roughly on total mission cost/ kgs landed on Mars ($2.5B / 2000) – and it seems 2000kg is being generous .

          I think this could have been better thought out .

          • You want to use the Skycrane? Ok, then you not only have to put more than just 200kg of fuel (which wouldn’t last long) but you also need to pack in a computer, transmitter, receiver, cameras and, if there’s still space left, scientific instruments.

            The skycrane was not the only piece of machinery that was sent crashing into the surface. Parachute, heatshield, backshield, course stage (and many other stages)… all were lost in the voyage. But that’s the point, they were tools for landing this 900kg robot into the surface of Mars. THAT’S where the science happens.

  8. Graham says:

    Can anyone tell me what happens the Sky Crane onec the rover has been safely landed?????

  9. Janis says:

    One minor nit:

    “So no other mission will use a sky crane through the end of this decade at the least.”

    That’s only 8 years, which is really not that much time. It’s like saying that the skycrane isn’t the future because we won’t be using one again in the next 15 minutes — okay okay, an ever-so-teensy exaggeration. :-)

    8 years isn’t that long at all, and this thing, insane and overengineered as it still seems to me, worked like a charm even without having been field-tested. (At least I’m assuming it wasn’t field-tested. It’s not like they have a spare Mars stashed in a closet someplace at JPL where they could test it.) They’ll find other uses for it in the future. It’s just that the future is longer than 8 years.

  10. Allen says:

    Wish I could have been there to see the sky crane crash onto Mars surface.

  11. Daneel says:

    I think we can only wait. As I interpreted it, Miguel was very enthusuastic about this mechanism and I woldn’t be surprised if they used it (or some variant) in future missions.

    But in the end we can only speculate.

  12. Kaputnik says:

    John, I’m afraid there are quite a few misunderstanding in your article, interesting though it is.
    The main omission is the reasoning behind choosing the Sky Crane system.
    This was driven principally by the need to have the rover able to reach the surface easily. A rover mounted atop a platform would need ramps; this works well enough at the smaller sizes of Sojourner or MER but it does not scale well and would have added complexity and mass at the scale of MSL.
    Secondly, landing the rover on its own ‘undercarriage’ is an elegant way to save a lot of mass (and complexity). The MSL descent stage was just a chassis housing engines and tanks. A complete landing platform would have been much heavier, and in some respects more complex, because it would need to support and protect the payload, have its own undercarriage, and provide egress equipment.
    Thirdly, a conventional platform style lander cannot deliver the payload quite as smoothly to the surface: because of plume effects as its nears the ground, plus the instability caused when one leg first touches the surface, such a design has to cut off its engines just above the surface and drop the remaining distance. This has of course worked in the past (even with the LEM) but it leaves you with the need for a very robust and stable landing platform- which is always going to be heavy.

    • John Rennie says:

      Thanks, you’ve made a number of excellent points that further illuminate NASA’s choice of the sky crane for delivering Curiosity.

      Yet I don’t really see anything in what you’re saying that fundamentally alters my main point, which was that the sky crane doesn’t represent the primary way that we’ll be landing future probes on Mars. The sky crane made sense for the Curiosity mission because NASA needed to deliver a big rover and didn’t want to be saddled with the weight, expense, and other headaches associated with a sturdy lander platform. Those same priorities won’t necessarily apply equally well to future missions, however.

      For example, even when NASA was involved with the ExoMars program, was there ever a plan to use a sky crane to deliver its rover (which will be bigger than Opportunity but much smaller than Curiosity? I don’t know the answer but I haven’t seen anything to suggest that’s true. Instead, ExoMars’s managers have seemingly always favored the semi-soft landing solution, in which thrusters and airbags will handle it. If someone has evidence to the contrary, I’d be glad to know.

  13. Kaputnik says:

    Yes I do agree broadly with your main point.
    Sky Crane is quite a specific method that suits large rovers. A static lander can possibly be integrated into its descent stage at lower overall mass/complexity.

    The really interesting developments in future Mars EDLS will come earlier in the entry phase- lifting bodies, deployable decelerators, hypersonic drogues or ballutes, or supersonic retropropulsion.

  14. Pingback: » NASA’s Curiosity Rover Landed On Mars, What Next? - Their Planet

  15. Larry Yoder says:

    You are so wrong,it worked and you can handle the fact that it landed so perfect ,1.7 mph tap down and it can handle even heavier payloads I know because I used the same principal the the army that is where this idea is from. Like the airbag landings ,accurate is not a NASA idea contrary to popular belief

    • John Rennie says:

      That’s fine, Larry. I didn’t argue that the sky crane wouldn’t work and I’ve never discounted the considerable technological genius involved in building and executing this kind of landing. What I said was that, no matter whether it worked with Curiosity or not, I don’t expect this kind of landing to be used very frequently by future landers on Mars or other planets and moons for the reasons outlined. We’ll see what happens.

  16. Larry Yoder says:

    I can tell you a story of a chance meeting in south Florida with a NASA engineer in the 1990 or 91 this chance meeting included an other man that was explaining how the air bag landing would work

  17. Larry Yoder says:

    I was introduced and was asked a question.. ifi I was going to land something on another planet how would I do it and I told them I would use sky crane looked at me quizzativly and I told them of a technical we used in the army called slingloading