I’m continually impressed by space missions that last several times their original life expectancy - the Hubble was expected to last 10–15 years and yet here we are 30+ years later expecting at least that yet to come.
There is a good reason for that. In order to have an amalgamation of systems with design life of a given duration, every component must have a design life of much, much longer. When you get lucky, a system can go for a really long time relative to its design life.
Great examples of this are Spirit and Opportunity. The latter failed at ~61 times its specified design lifetime.
Having been a small part of some space-mission design and larger science collaborations, I can state with confidence that you really don't want your subcomponent to be the one that causes a failure.
And Voyager 1 and 2 are still zooming as far away from our solar system as they can get and happily sending back teeny bits of data every now and then. Pretty wild to think their voyages are now over 40 years old.
For a lot of stuff there's redundancy too— like rovers with six wheels but that can drive with any four, and the Curiosity and Perseverance missions having dual computers and dozens of cameras. I'm sure they run scenarios where significant chunks of the rover's systems fail and they figure out how to carry on anyway— thinking in part here of the famous HGA issues on Galileo:
People are scared and want to save their asses so they make super low estimates. That way it sounds really good in the media when things keep working.
I used to do the same when I did software development “I will be done with that in 2 days” then proceeded to build the thing in 2 hours chill 3-4 hours and hand it in earlier than expected.
Everyone was happy.
There is truth to this in aerospace: components will be qualified to the design life of the vehicle, but the qualification campaign is intentionally conservative to envelope all possibilities and uncertainty. When the actual conditions are more benign, more life can generally be expected.
Additionally as another case, Hubble has been continuously losing gyroscopes since it was first launched. Some or all gyros were replaced during most shuttle servicing missions. Seeing the writing on the wall for servicing missions post-Columbia, NASA developed software to operate with fewer gyros at a time, allowing for fewer to be spun up at any time, but importantly, allowing for the telescope to continue to operate after more than 3 of 6 gyros had failed. The key here is that engineers can often coax more performance out of a damaged (or otherwise limited) subsystem given the incentive- this has been the case in my aerospace experience, and seemed mirrored with Hubble too.
> People are scared and want to save their asses so they make super low estimates. That way it sounds really good in the media when things keep working.
I think that's an unnecessarily pessimistic take. Any system can have its life expectancy modeled over a range, e.g. a 99% chance it will survive 1 month but only a 50% chance it will survive to 1 year. When you're just giving one number to the media or other stakeholders, e.g. "expected life span", (or, for that matter, a single estimate to your boss), to be able to give a value with high confidence you will have to pick a number that is in the 99%-ish range, which means there is a good chance your system will survive much longer.
Underpromise. Overdeliver. You didn't have to do a crunch and the people waiting got something ahead of the deadline. That's my practice whenever possible as well :-)
As to your related point, I probably wouldn't use the term "scared." But, yes, there are a lot of incentives to meet/exceed goals and strong disincentives to fail by not meeting a somewhat arbitrary lifetime of a probe. Of course, some probes have clear primary objectives and you want to hit those but you may not want to "promise" you can hit a bunch of less important secondary objectives as well even if you think you probably can.
Well, that is one way to do it. I think it was necessary for HST because it was such a complex instrument that had to be capable of doing many things, and reliably -- because specifically you have to have a mirror of a certain size to make it worth it, and so it calls for a certain level of engineering reliability (and cost).
On the other end of the spectrum, the Mars programs (lately) were incentivized to do the opposite -- cheap, fast, and good, even if some of them fail.
Different circumstances may call for different incentives.
It proved Hubble right about the expansion of the universe, so all in all I think we can say with absolute certainty it was a smashing success.
Apparently the total cost for Hubble is on the order of 10 billion. That's just a fraction of what a tech company like Apple makes every quarter, so I think it was a totally reasonable expense relative to its discoveries.
And luckily we learned building giant, precise mirrors with 1980s technology maybe wasn't the best decision. The James Webb design is totally different and hopefully less prone to such problems.
A different mission, but Jake from We Martians did a really great twitter thread[0] about what goes in to the mission length. I'll summarise the thread here (mostly just quoting verbatim):
Firstly, this is about why solar powered Mars missions (specifically InSight) do not have solar panel cleaning devices, but it touches on how missions can be extended and what thinking goes into that.
Robotic missions to Mars are competed like any other; they have to "earn" their chunk of the budget by being effective science missions at effective prices. This causes a very normal tension between "do a lot" and "don't cost a lot".
InSight in particular is part of the Discovery program at NASA, which technically has a cost cap of around $500M. Program managers thus do everything they can to reduce the cost without sacrificing science. You want to avoid taking instruments off, for example.
One really effective way to reduce scope is to shorten missions. This has two big benefits:
1) Every year a mission doesn't operate is a year you don't have to pay salaries for those who operate it
2) Hardware on the spacecraft can be made cheaper if it doesn't have to last
That second point is really critical.
The testing program you have to put stuff through doesn't scale linearly. A mission that lasts twice as long costs more than twice as much to test. It spirals fast.
In InSight's case, they also saved money by reusing a spacecraft bus design from the Phoenix mission, which was also designed for a short mission - just 90 sols in fact.
See the similarity?
So all of these reasons (staying under cost cap, reducing operation cost, reducing testing costs and saving money by reusing spacecraft busses) combined to drive a decision that InSight's prime mission is just 1 Martian Year (about 2 Earth Years), which it completed in November.
So the short answer to "why not add solar panel cleaning devices" is that they don't need them!
By the time the dust takes down the spacecraft, it will have completed its mission. It's an element of disposability, though NASA probably wouldn't characterize it that way.
Now, often the vehicles last longer than their prime mission. Spirit and Opportunity, for example, blew past their prime mission of 90 sols. All of NASA orbiters (ODY, MRO, MAVEN) have gone past their prime mission, too. Last November, InSight also surpassed its prime mission.
When this happens, NASA reevaluates the vehicle's health & cost against further science it can do, and decides to either re-fund or end a mission.
InSight got the funding, and so here we are making the best of that!
NASA isn't scrambling to save a mission hampered by dust they forgot existed. They're doing their best to squeeze out more return on investment for the taxpayer on a spacecraft that has already completed its mission!
> So the short answer to "why not add solar panel cleaning devices" is that they don't need them!
> By the time the dust takes down the spacecraft, it will have completed its mission. It's an element of disposability, though NASA probably wouldn't characterize it that way.
I don't think that's a very compelling answer. The thing to ask, I guess, is whether they would have added them anyway if they were very cheap and lightweight.
If yes then "it was a waste of resources for the designed mission" hits much closer to the real reason. But if no, it means the real reason is "bureaucratic maneuvering caused a longevity problem on purpose" which is really bad.
Sounds like they didn’t do a good estimate of the lifespan. Grossly underestimating it to impress the general public is quite lame if you think about it. If you were 60 times off on your estimate that was a horrible estimate.
I think I'm going to genuinely cry the day they deorbit Hubble. Maybe it doesn't make financial sense but if there was ever an instrument that deserved to be retrieved and given a hallowed place in a museum, Hubble is it.
The original plan was that it would be de-orbited aboard the Shuttle and put in a musuem.
I'm not sure what the current plans are, but the last servicing mission installed the Soft Capture Mechanism, and the current understanding is that an uncontrolled re-entry would pose an unacceptable risk to human life. So, it seems like there is still a chance to put it in a musuem.
I wonder how expensive it would be to design an "affordable" space telescope which could be produced in a small series. The Falcon 9 should be able to launch a roughly hubble-sized telescope. So the launch costs would be a fraction of the shuttle costs back then. This means, more than one telescope could be launched and the telescope to be more designed to be "expendable". Which once again reduces costs a lot.
I think one driver for the enormeous cost of a lot of space missions is the amount of overdesign happening as due to the high costs, failure is not an option. Just designing for 5 years and allowing for the early failure, costs should be so low that a failure is no longer a big issue.
Also, if there is one basic platform, sending different telescopes with quite different sensor packages would be an option.
A lot of the cost for space based missions is also in instrument design. When you make a telescope, you don’t need to just design the mirror and the stuff that holds it up, and a camera with a series of filters. You need to design spectrographs and other complex systems that need to be “space-hardened” and extremely well tested. A “sensor package” doesn’t really cut it. See for example this spectrograph:
Huge instruments like these that actually power today’s science is another reason why ground based astronomy is required, and why we can’t just move all observation to space if satellite clusters ruin the night sky.
Hubble is mainly a set of cameras attached to the telescope, it doesn't have anything comparable to what you linked. Yes, I am aware of all the advantages of ground based astronomy, which just got a huge boom by segmented mirrors basically removing the mirror size limit. The flexibility and costs of ground-based astronomy are superior for anything which can be done well from the ground. I have nowhere suggested to stop or defund ground based astronomy. We need ground based astronomy and space based one. So we need to come up with a replacement of Hubble, thats where my suggestions come from.
And yes, I am aware of the challenges of space-going equipment. But if a system is designed for a life time of rather 5 than 30 years, things become a bit easier. Also, if you look at the camera equipment of Perseverance, a lot of it is off-shelve, only slightly enhanced for its intended purpose.
As another commenter said, they are designed to work for 5ish years. That we have had Hubble so long is essentially a miracle. Same thing for the Mars rovers that limped on for years despite only being designed to last months.
Also your claim about the instruments on Hubble is incorrect.
Uh, missions are designed to last max five years, so you're not gonna find any savings there. Maybe if you cut the high reliability requirements to five months you can skimp on some stuff, but I'm not sure that actually relaxes demands so much as that's still a lot of thermal cycles to maintain instrument alignment over.
A giant mirror is not expendable, and takes more than 5 years just to make. So until the astronomers tell the engineers, "hey no problem, we can do good science with little cheap mirrors", engineers are going to be stuck engineering the rest of the telescope around the economics of the mirror.
At about 2 meters of aperture, you can quite well still do a single mirror at reasonable costs. If the launch costs like 60 million, the telescope would realistically have a budget of 60-120 million for the platform. On that scale, a series of 2m mirrors should be affordable - you would of course order more than one mirror, so you have a reasonable supply.
Considering that a single shuttle launch costet close to 1 billion and the Webb telescope costs many billions, spending 1 billion on a set of 5 telescopes including launch costs sounds like a worthwhile program. Which could be scaled to more telescopes as long there is budget.
2m is pittance for professional astronomy. Hubble may have a 2.4m mirror, but remember that is a 1980s era telescope.
Current and future telescopes target 10-30m mirrors, purely for light capture reasons. We’d need to put something that big in space for it to be worthwhile.
No, even Hubble still delivers great contributions to science which are not quite matched by earthbound telescopes. Those win on aperture for sure, but Hubble is the only large telescope outside of the atmosphere. This gives access to wider wavelengths and removes all other atmospheric influences. That is the reason, the final passing of Hubble would be a huge loss, if not replaced.
There are several aspects we could improve on the current situation: make "better" telescopes or make more telescope. Having only 1, soon 0 space-based telescopes is cutting it very thin. Having 5-10 would increase our research a lot as more things could be observed.
Of course, eventually there will be a need for large space telescopes too, probably those will be enabled by Starship. But first we should use the currently available technology to make sure we don't fall back into the level of the 80ies when Hubble dies. And use modern technology to make this a "cheap" effort vs. the extremely expensive Hubble.
AO improvements over the past decade have made ground-based astronomy provide better images, with large enough apertures, than Hubble.
You are correct - the NIR is blocked by the atmosphere. That is why we are launching Webb. Optical, however, is very much now the domain of ground-based astronomy.
> Having only 1, soon 0 space-based telescopes is cutting it very thin. Having 5-10 would increase our research a lot as more things could be observed.
Apparently they went the other way and instead of multiple Hubble-equivalent telescopes they build one which has equivalent sharpness of images but a much larger field of view, so can take images 100 times as large as what Hubble can.
James Webb space telescope is scheduled in 6 months and it has a 6.5m segmented mirror.
A 2m telescope is pretty much useless, it doesn’t have enough angular resolution and it’s smaller than HST current mirror.
I was more thinking of roughly hubble-sized telescopes, but as a program optimized for having several telescopes. Like spending 1 billion in total for a larger amount of telescopes, which can be easily replaced should one fail or new instruments are required.
Also - we are currently launching another Hubble, called WFIRST. This was provided by a 3 letter US agency and required a mirror redesign. Launch costs are not a factor now, and weren’t then either.
"Currently"? It is worked on and might launch after 2025. Also, it is different from Hubble, so not a direct replacmenet. But indeed, with a total budget of over 3 billion dollars, launch costs are no longer the problem. Remains the question: why is it so expensive? Because it will have been worked on for more than 10 years and probably is over-engineered like hell, as you must not fail with a 3 billion dollar telescope.
You are saying the magic word: prototypes. That explains the huge budgets, time and cost overruns. They certainly have an important place in science, because sciences often is about doing things the first time. But there is also a place for production runs. Because you are no longer just making prototypes but start to reuse the experiences and expenses.
Hubble is 30 years old, and there isn't any direct replacement in sight at all. At latest now, with cheap launches and much improved technology, it is more than time to replace it and as it is more or less a direct replacement, use the scale of production. Don't build one telescope for 3 billion, build 10 at like 100-200 million each. If one ore more of them fail, so what?
Well the reason that we would build the one 3 bn telescope, rather than 10 100-200 m telescopes, is because they can tell you fundamentally different things about the Universe.
In many ways, to move forward, we must go bigger with the tech, not wider. What you are essentially suggesting is a survey - we already have instruments and telescopes to do exactly that, with new ones coming online soon (see the LSST). Deep observations like Hubble was meant to do can actually be done better from the ground now (e.g. the VLT, ELT, etc.). As such, focus has shifted to those for optical observations.
The community at large, however, has decided where it would like to spend its budget. That was on Webb for the IR wavebands, to study so-called 'high redshift' objects that Hubble struggles to see, and leave the more detailed imaging to those ground based observatories.
Look, I would love another Hubble. That would be really useful! And that's why WFIRST exists. But an in-place replacement doesn't make sense. The tech on Hubble (not necessarily the mirror, but the _instruments_, the spectrographs, the interferometers) is really old. You can't take an off-the-shelf component, because it needs to be built to withstand space. So you need to create a one-off, 'prototype', space-hardened instrument. It is also not a case of worrying about them 'failing', but you also must worry about them slowly drifting out of a calibrated range. On the ground, you could fix the instrument, or re-calibrate it. That's not possible in space. That's not a 'random' style error, where we can get out of it by upping the number of them - they are systematic, so all 10 of your proposed telescopes would be ruined.
NASA's OSAM-1 (formerly known as Restore-L) mission is planned as a technology demonstrator of robotic servicing [0]. The target is the Landsat 7 satellite; but if the technology is successfully demonstrated by OSAM-1, it might be used for robotic servicing of space telescopes in the future.
Originally targeted to launch in 2020, the project has encountered delays, and NASA is currently aiming for January 2025 as the launch date [1]
Ha. The James Webb telescope is implementing a ton of new technologies and manufacturing techniques. Building a second one would presumably cost significantly less than the first, although I have no real basis for making that claim.
The Hubble Space Telescope (HST) is basically a civilian version of the KH. And the National Reconnaissance Office (NRO) is launching one of those KH satellites every couple years. I bet the latest one (USA-290) has much better tech than Hubble and would make a good replacement for Hubble, with few modifications required.
JWST is fundementally a different type of space telescope from Hubble, all of which contribute to its costing:
* It's being deployed at L2 Lagrange point instead of LEO - so the launch costs are even higher.
* HST had the insurance of knowing that repair was possible - JWST does not (both due to it's location and lack of capability) and so everything drags on even more to try to reduce risk
* JWST is technically more complicated. The thermal management system needed, and entire origami unfolding mechanism alone would make it more complicated and expensive. This doesn't even start looking at the optics.
I don't know much about how the STS or ISS EVA airlocks are configured, but I wonder if you could do it with a slightly modified Crew Dragon docking nose-to-nose with another one. One acts as a fallback and home for the non-EVAing crew, and stays pressurized, while the other is used to stage the EVA and ultimately evacuated so that the astronaut(s) can leave via the hatch. You could launch one unmanned and have it wait in orbit for the other one, with crew.
There's a fair amount of room in those trunk sections for any gear or parts they need to bring up, too. I guess my biggest concern would be whether the hatches are big enough for real EVA suits to get in and out easily, and whether the life support system can support multiple repressurizations without significant modification. They might also need to modify the docking adapters, which are "supposed" to be androgynous but IIRC actually aren't.
