This is not a NASA Website. You might learn something. It's YOUR space agency. Get involved. Take it back. Make it work - for YOU.
Exploration

NASA OIG Reports Mars 2020 Rover Sampling System as Largest Risk

By Marc Boucher
NASA Watch
January 30, 2017
Filed under , ,
NASA OIG Reports Mars 2020 Rover Sampling System as Largest Risk

NASA OIG: Audit of the Mars 2020 Rover Project
“The largest risk to the Mars 2020 schedule is the Project’s Sample and Caching Subsystem (Sampling System), which will collect core samples of Martian rocks and soil and place them on the planet’s surface for retrieval by a future robotic or human mission. At Preliminary Design Review (PDR), three of the Sampling System’s critical technologies were below technology readiness level (TRL) 6, meaning the prototype had not yet demonstrated the capability to perform all the functions required. Projects are evaluated during PDR to ensure they meet all system requirements with acceptable risk and within cost and schedule constraints. The immaturity of the critical technologies related to the Sampling System is concerning because, according to Mars 2020 Project managers, the Sampling System is the rover’s most complex new development component with delays likely to eat into the Project’s schedule reserve and, in the worst case scenario, could delay launch. As of December 2016, the Project was tracking the risk that the Sampling System may not be ready for integration and testing – the period when a spacecraft is built, undergoes final testing, and is prepared for launch – in May 2019, as planned.”

SpaceRef co-founder, entrepreneur, writer, podcaster, nature lover and deep thinker.

31 responses to “NASA OIG Reports Mars 2020 Rover Sampling System as Largest Risk”

  1. Bob Mahoney says:
    0
    0

    Not surprising. How does one design a ‘Tupperware’ system for another planet’s surface…without being sure when or even possibly how said containers will be retrieved? I hope they come with marker flags that while stick up high enough to clear the height of whatever years (excuse me, sols) of dust storms might deposit…

    • Jeff2Space says:
      0
      0

      It certainly does seem like they’re “putting the cart before the horse”.

      • Paul451 says:
        0
        0

        Particularly when the return vehicle hasn’t been funded or designed or even clearly conceived. Surely the two or three missions should be designed together?

        (Three missions if they stick with the sample-storer (2020), sample retriever-bot (2024?), ascent/return vehicle (2026?) architecture, two missions if they combine the retriever and ascent/return vehicle.)

        • fcrary says:
          0
          0

          According to the current Decadal Survey, it would be about one per decade. The scientific community wouldn’t have supported the idea of the next three planetary flagship missions all being dedicated to a Mars sample return. Spreading them out was generally seen as a way to get a foot in the door.

          • Paul451 says:
            0
            0

            So you send a rover to collect and store samples in 2020.

            Then in 2031, send a single-function ‘bot to collect the decade old sample containers and put them in a neat pile.

            Then in 2041 or 2043 send a MAV to pick up the pile of now two decade old sample containers for return to Earth.

            Wow that is bonkers. People thought that was a good idea?

          • TheBrett says:
            0
            0

            It’s ridiculous. Why not just send a stationary lander with a robot arm and small drill to land in an interesting area, grab some samples, and then launch back to Earth once that’s done? The biggest engineering problems with that would just be building the return vehicle – the rest of it is proven technology from prior landers.

          • fcrary says:
            0
            0

            The problem is that they don’t want to return just any samples, they need to return geologically interesting ones. The baseline is something like 30, 15-gram samples, and the goal is to revolutionize our knowledge of Mars. When you compare that to the quantity of samples terrestrial geologists collect, with far more modest goals, getting interesting rocks is a big deal.

            When you combine that with the criteria for a safe landing site and the uncertainty in the actual location of landing, a rover becomes critical. MSL/Curiosity had an estimated error ellipse of 20 by 7 km, and actually landed a bit over 2 km from the center. Even if you could identify a good site to sample, the odds of landing within arm’s reach of it are miniscule.

            Worse, safe landing sites tend to be geologically boring. Flat and as rock-free as possible is good for a landing site. But that isn’t a good place to find interesting samples. The base of a cliff, or a similar place, would be good for finding interesting samples, but that’s about as far from a safe landing site as you can get.

            So the idea is to land a rover at a safe location within a few to a few dozen kilometers from a geologically interesting sample collection site, drive out, collect samples, and return them to the safe landing site. This also lets you collect samples from multiple locations. Locations which could be a few kilometers apart. So you have the potential to collect several sorts of different, interesting samples.

            I will note that this logic assumes there will only be one sample return mission. It implicitly assumes you have to make the most of the one-and-only opportunity. Given the budget and flight rates, that may not be a bad assumption. But I do think the idea of making the most of the mission contributes to the cost and low flight rates. So there is some circular logic involved.

          • TheBrett says:
            0
            0

            I was figuring that it wouldn’t be the only such mission, and that if you kept the costs down you could do a series of them. There would still be the landing site issue you mentioned, but if you’re sending more missions you can risk some of them on trickier landing sites.

            If you’re only getting one crack at the apple, then you would need to pile on the capabilities and scale of the sample return.

          • fcrary says:
            0
            0

            Now you are getting into the issue of risk posture. Planning on multiple missions and making tricky landings means accepting the fact that some of them will crash. The traditional and accepted NASA risk posture is against that.

            I think of this as the “failure is not an option” mindset. It results in high costs. This makes the idea that there will be only one opportunity a self fulfilling prophecy. Personally, I would prefer ten missions, with a 50% chance of failure to one with a 1% chance of failure. My preference would result in five successful missions. The alternative would be only one successful mission.

            But the other side of this issue can not be ignored. My preference would mean five failed missions, and that’s five times NASA would have to explain to Congress why they failed. That could easily hurt future funding. I think that problem could be dealt with, but it is a tough sell. Someone would have to make a case for good, overall benefits, a good cost-to-benefit ratio and not fixating on bad press from the failures.

          • Michael Spencer says:
            0
            0

            That’s an awful lot of complication and opportunity for errors.

            Ive wondered about the rush- if the mission to actually pick up the samples and bring them home to daddy is many years away why not wait, pushing the collection mission out? Why not take advantage of more time to develop the collection mechanism?

            Moreover, as I read through your discussion I wondered about the collector mission having to land quite near the samples. Wouldn’t this imply the need for a fairly robust rover? And if so why not use that rover to find the samples in the first place?

          • TheBrett says:
            0
            0

            On top of that, do you really want to have samples that have been sitting in sealed boxes for two decades? I’d rather wait for a mission that can provide relatively fresh samples for return to Earth.

          • fcrary says:
            0
            0

            Mars is a fairly benign environment for storage. Take a look at the plans for storage and custody of any extraterrestrial material returned to Earth. The plan to keep the material around and usable for decades, so it can be reanalyzed as new laboratory techniques are developed (and they have done so with the Apollo samples.) Years of unspoiled storage in a cold, near-vacuum environment has got to easier than decades of unspoiled storage in Huston, Texas.

          • fcrary says:
            0
            0

            I think the logic is that, today, a reasonable level of funding will allow sample collection. Ten years from now, when the collection work is done, a reasonable level of funding will allow retrieving the samples. But it’s likely that a realistic level of funding will _never_ allow collecting and returning samples at the same time (or with overlapping schedules.)

            This isn’t too different from human spaceflight in the early 1970s. They wanted a space shuttle, a space station and a trip to Mars. The budget for that wasn’t there. So they decided to just develop a space shuttle in the 1970s and early 1980s. Once that was done, and development money freed up, they could worry about a space station. Once the space station was build, and development money freed up… That didn’t work out as planned, but you can see the similarity in the logic.

            As far as the sample recovery goes, I’m not sure what the current plan is, or how detailed it is. Precision landing is something they are working on for the 2020 mission. But even improving the current +-10 km (planned) or ~2 km (achieved by MSL) accuracy by an order of magnitude would do for sample recovery without a rover. +-10 _meter_ accuracy wouldn’t do that. Maybe you could use a very simple, low resource and short lifetime rover if you could land within a few hundred meters. I’m sure the 2020 rover wouldn’t last. Power from the MMRTGs decreases too quickly, and it won’t have enough power to operate after a decade on Mars.

          • fcrary says:
            0
            0

            Would you believe politics were involved? Not Republican versus Democrat politics, but Mars versus Titan and geology versus atmospheric science. The Decadal Surveys are supposed to be a consensus recommendation by the whole planetary science community, about what NASA should be doing in the next dacade. Since planetary scientists have very diverse opinions on the subject, making recommendations everyone can live with is a political as well as a scientific/technical matter.

            In this case, a Mars sample return within a decade would have a huge cost. So high that it could have ruled out anything more than a couple of Discovery missions to study the rest of the solar system. That would not have been something the whole planetary science community would have supported. At the same time, the Mars community would not have supported putting off a sample return (which is sort of a holy Grail for them) for at least another decade. The consensus was to do something, almost anything, to start work on a Mars sample return within the period covered by the current Decadal Survey (2013-2023) but to leave the actual sample return to sometime in the 2023-2033 or possibly 2033-2043 timeframe.

            I have real doubts about the current concept. I’d call it inefficient and possibly very optimistic (by requiring coordination of several missions spaced years or decades apart), but that was the best thing the authors of the current Survey could come up with.

          • Paul451 says:
            0
            0

            At the same time, the Mars community would not have supported putting off a sample return (which is sort of a holy Grail for them) for at least another decade.

            Which they did anyway. Are they morons? They gave up a useful mission window for the pretence of having a “sample return” mission this decade that won’t return samples this decade. Are they particularly dumb toddlers?

          • fcrary says:
            0
            0

            I don’t think there is any reason to expect the next decade would be different. This time, for the 2013-2023 Decadal Survey, they expected NASA could afford something like one flagship mission (~$2 billion), a couple of New Frontiers missions (~$1 billion each) and, with luck, five Discovery missions (~$500 million each.) That’s about $6.5 billion over a decade.

            The cost of one of the three elements of the sample return was expected to cost $2 billion. The original cost estimate for the 2020 rover was $1.3 to $1.7 billion (when announced in December 2012) and it’s now up to $2.4 billion. (All cost numbers are from this OIG report we started off discussing.) Arguably, that’s the easiest, and therefore cheapest, element of the sample return plan. Trying to do it all in one decade would be asking for ~$6 billion out of the ~$6.5 billion the authors of the Survey assumed would be available. There is no reason to expect this will be any different for the 2023-2033 or 2033-2043 period.

            As far as I can see, there are four possibilities:
            1) Give up on a Mars sample return for the indefinite future (which is unacceptable to part of the science community.)
            2) Give up on more-or-less everything else in the solar system for ten years (which is unacceptable to the rest of the science community)
            3) Dedicate a large (multi-billion dollar) fraction of the budget to technology development, to reduce the cost of a sample return in the next decade. I think this one makes the most sense, but it’s very hard to convince senior management or congressmen: This approach would mean a full decade of funding without any exciting “seven minutes of terror” events or impressive discoveries.
            4) Split the work up and fly it as multiple missions, with the time between each dependent on the available funding.

          • Daniel Woodard says:
            0
            0

            I agree that technology development is the best strategy in the long run. Indeed, that’s to some degree what the Obama Administration tried (unsuccessfully) to do with the space technology program. I particularly wonder about the JIMO mission. A flight ready nuclear reactor would provide a powerful source of energy for both instruments and propulsion that would be what we used to call an enabling technology. It would make a quantum difference in our ability to explore the outer solar system, and a reactor is much safer to launch than an RTG since uranium is, by comparison with plutonium, almost nontoxic.

          • fcrary says:
            0
            0

            Strictly speaking, a reactor is only safer than an RTG before you turn it on. Once you turn it on, decay products of uranium start building up and some of them are much worse than plutonium dioxide. JIMO would have been launched to Earth orbit and slowly spiraled out under electric propulsion. That means a fair amount of time on Earth orbit after turning the reactor on.

            The reactor JIMO was looking at was also couldn’t throttle down to low power levels, and shutting to off wasn’t really an option. There were some concerns about how to use the power when it wasn’t needed without causing thermal concerns. A good radiator design could probably have handled that, but they were seriously considering (or looking for) scientific instruments which used lots and lots of power. Just to have a way to load balance when the electric propulsion system was off.

            I’m not disagreeing about the value of a reactor for planetary science missions. But I think JIMO just wasn’t a good implementation of the idea. I feel the same way about carbon moderated reactors on the ground, and anything involving NaK as a coolant gives me bad feelings.

          • Michael Spencer says:
            0
            0

            5) Recognize that sometimes one’s eyes are bigger than one’s stomach. For reference, see Europa submarine.

            The Mars crowd has been spoiled. I’d much rather see a Uranus orbiter or Neptune orbiter- these are within closer reach, though distant.

          • fcrary says:
            0
            0

            Even the Europa orbiter (in almost all its incarnations) was an example eyes being bigger than someone’s stomach. This is a real problem with how NASA develops mission concepts. Typically, the scientists are asked what the big questions are, and what measurements are required to answer those questions. Then those requirements are handed over to engineers, who are supposed to develop a mission concept which does the job.

            It’s hit-or-miss how much feedback there is, or whether the scientists’ desired measurements are realistic. Some scientists don’t even think they should consider how practical or affordable their goals are: They aren’t experts, and don’t feel they should second-guess engineers who may find a clever solution. But this can let things get out of hand.

            The MSL/Curiosity mission is a good example. The 2003 Decadal survey listed it as the highest priority for Mars missions in their “medium cost” class. That meant they expected it would cost between $325 and $650 million. It turned out to cost closer to $2.5 billion. But there was no process to revisit that recommendation; no opportunity to ask the science community, “Now that it looks like it will be a couple billion, are you sure it’s still the highest priority?”

            I’m also afraid a Uranus or Neptune orbiter may not be in reach either. I’d love to see one, but there is one very tough problem. To get there in a reasonable time (say, less than twenty years) the spacecraft ends up approaching at far too high a speed to stop. The only practical solution I’ve seen is aerocapture. That certainly isn’t easy, and I can’t see NASA betting a flagship mission on aerocapture working the first time they try it.

          • Michael Spencer says:
            0
            0

            I’ve wondered about the sample return; yours is the first to assign some sort of motivation or explanation. But it seems off, at least to me; granted you’re part of the scientific community, so have a better reading.

            But surely there’s little support for a half-assed mission simply to get the ball rolling? Piling up some samples of dubious quality- to be returned decades later when collection tech would be far more capable, one presumes? Those scientists have to be a lot smarter than that.

            (And I would add parenthetically that the Mars boys have had an awful lot of love from NASA for many, many years; crying about a sample return is a little over the top).

          • fcrary says:
            0
            0

            I think part of the problem is the fact that it isn’t certain technology for sample collection and return will be better in ten years. (Since the Decadal Survey was written behind closed doors, this is just based on the public part of the process.)

            Technology development for unmanned missions tends to get a fairly small amount of funding. Historically, most of the work has been done when there was a specific mission, with specific requirements driving the effort. Without a mission with relevant requirements, the necessary technology might not be significantly better in ten years. These are some fairly specific applications, so you can’t count on improvements in general being applicable.

          • Daniel Woodard says:
            0
            0

            Sample return would drive technology development for the Mars Ascent Vehicle, but it would be rather specialized technology. Samples from smaller bodies like asteroids and comets would be highly informative without requiring a high-powered ascent.

  2. Tally-ho says:
    0
    0

    Does anyone think naming it Mars 2020 was a bad idea? How often does NASA nail project deadlines? The need to prepare some quick come-backs if they miss the 2020 launch date.

    • fcrary says:
      0
      0

      Except for Curiosity and InSIGHT, NASA’s track record for launching Mars missions has been pretty good in the 21st century. Viking did slip by at least one launch window, and I don’t remember what the schedule for MO was.

      On the other hand, I do remember the early 1990s joke that the next Russian Mars mission was called “Mars 1992+2n”

      • the guy with the cat says:
        0
        0

        MSL and Mars Observer both slipped (thanks for remembering MO!), so the track record is less than stellar.

        • fcrary says:
          0
          0

          MSL and Curiosity are the same mission (well, Curiosity is technically just the rover, while MSL includes the carrier vehicle and EDL systems.) So that’s three launch slips out of twelve mission, counting from 1990 and assuring I counted correctly.

  3. Daniel Woodard says:
    0
    0

    While I understand the rationale for sample return, the advances in situ analysis have been extraordinary, Ultimately the scientific productivity may be higher from pushing the number and capability of the rovers.

    • Michael Spencer says:
      0
      0

      Indeed. I always thought of sample return as kinda silly in favor of better remote sensing (well, in situ). On the other hand it’s a fine test for those Mars Mavens to test some hardware.

      • Daniel Woodard says:
        0
        0

        You’re right. It’s not always in situ per se since samples are sometimes dug up and put in analytical instruments in the rover.