Former NASA astronaut and Shuttle boss weigh in on fixing Hubble Space Telescope

The outlook continues to look a little bleak for NASA's veteran Hubble telescope as a former astronaut and a Space Shuttle manager weighed in on repair options and the possibility of a fix. NASA has remained silent on the fate of the Hubble Space Telescope (HST) since an admission last week that back-up computer hardware …

[Anonymous Coward]

So you're telling me there's a chance

As an eternally optimistic space geek, I choose to believe the NASA update

Let the boffins take all the time they need. No beers yet, just coffee.

(and how can an IT site not have a coffee icon)

[Mark 85]

Re: So you're telling me there's a chance

I suggest beers first, coffee later. Seems with many tech types they're more creative after a couple of beers. But win or lose, those engineers deserve a beer. They have to be racking their brains and putting in long hours.

[Anonymous Coward]

Re: So you're telling me there's a chance

It's called the Ballmer Peak.

It's a real thing, but also obviously parodied on xkcd: https://xkcd.com/323/

[Tom 7]

Re: So you're telling me there's a chance

Any idea what those units are in Uk ish? I used to be able to play pool like a god between 8 and 10 pints of ordinary bitter. I once beat the uni darts champion when someone moved me from the 'oky to the wall and back for my go after more beers. A friend plays some of the best guitar you've heard in the two pints before falling asleep standing up with his head on the PA. Never really tried it at work but we did used to have a liquid lunch on a Friday - some of the best Adnams ever - at BTRL and then have a sort of seminar while we sobered up to go home and came up with some stonking ideas, many still stonking on the following Monday.

[Dismas]

Re: So you're telling me there's a chance

> Any idea what those units are in Uk ish?

I reckon percentage blood alcohol concentration is a pretty international unit, but I think I know what you mean. I was curious what alcohol intake for peak programming skill that graph actually implies. It turns out, for an 80kg male, to be 4 and a quarter pints! My working follows:

The blood alcohol concentration (BAC) produced by consuming x amount of alcohol is commonly estimated via the Widmark formula*:

BAC = [ Grams of alcohol consumed / (body weight in grams x r) ] * 100

Where r is a constant depending on sex: female r = 0.55, male r = 0.68

The XKCD graph appears to put the programming skill peak around 0.14% BAC.

Therefore, for an 80kg male:

Let x = grams of alcohol consumed

Apply Widmark formula:

0.14 = [ x / (80000 * 0.68) ] * 100

x = ( 0.14 * 54400 ) / 100

x = 76.16g of alcohol to produce BAC of 0.14%

Now to put in terms of pints of bitter, assuming 4% alcohol by volume:

Density of alcohol (ethanol) = 0.789g/ml

76.16g of alcohol = 76.16 / 0.789 = 96.5ml of alcohol

(96.5/4) * 100 = 2412.5ml of bitter

= 4.25 pints of bitter

Well I’m no coder, to the point it might not make much difference, but I think I would not want to be doing skilled work after 4 and a quarter pints. Getting ‘big picture’ ideas in good company, sure, detailed work not so much.

For reference, the UK legal driving limit is 0.08% BAC, which gives an estimate of 2.4 pints for an 80kg male (n.b. this is an *estimate*, the only completely safe driving limit is zero pints).

*Widmark EMP. Principles and applications of medicolegal alcohol determination. English translation of 1932 German edition, Davis Biomedical Publications 1981.

[NoneSuch]

Re: So you're telling me there's a chance

The US has the secret X37-B that can be used. It is crewless and would need a cabin fitted in the 2.1 x 1.2m cargo bay. Big enough for one person and spares.

[john.jones.name]

the USA have other scopes with same sort of design pointing at earth

I suspect there is a lot of shared systems.

maybe get them on the phone and point out the same things could happen to their equipment...

[MyffyW]

Re: So you're telling me there's a chance

In fairness el reg do have a "coffee spilt on keyboard" icon, which is somehow more appropriate

As an eternally optimistic flapper, I hear "Robot Arm, Airlock, 6 people" as a challenge I'm sure certain private enterprises are equal to.

[John Robson]

Re: So you're telling me there's a chance

I question the need for six people - Gemini managed spacewalks with just two.

I still think the least un-likely* approach is a modified dragon, carry a robot arm, and maybe an airlock, in the trunk (37 cubic metres to fit something in - maybe a little bigelow module?)

* i.e. it's still not likely at all, but I think it's at least technically feasible.

[imanidiot]

Re: So you're telling me there's a chance

All previous Hubble Service Missions were executed with the crew of 6. 2 on EVA to perform the work, a spotter inside the shuttle to assist with their respective work, keep track of checklist items, what work was performed, etc, one person operating the Canadarm to maneuver one of the astronauts or equipment around as they worked and a person to monitor and operate the orbiter. IE, during the actual servicing all 6 crew members were busy on tasks. While it might be possible to reduce the crew count somewhat, you're not going to do a servicing mission on something as complex as Hubble with a crew less than a minimum of 4 imho. A 6 member crew really hits the sweetspot of "everyone has work to do" without overloading any individual.

And even IF the shuttle had been still flying, preparing a service mission after a fault like this would probably still take another year or 2 to train and prepare tools and hardware.

[John Robson]

Re: So you're telling me there's a chance

Thanks - I hadn't actually looked much at the schedule of previous missions.

You do only list 5 jobs, but I can forgive that ;)

Dragon does have seats for seven (NASA only use four at the moment), although I was hoping to use a couple of those to store EVA suits during take off/landing.

"And even IF the shuttle had been still flying, preparing a service mission after a fault like this would probably still take another year or 2 to train and prepare tools and hardware."

Which is still possible - it's "only" the payload computer that has failed, so hubble can hang out for a while waiting.

I'm not suggesting that dragon could go up tomorrow, there is a huge amount of work that's needed (the designs for the hardware of the arm and an airlock simply don't exist yet for one thing), I still think it's the least unlikely servicing possibility.

The next one down is probably starship getting to the point where it can yoink stuff from orbit - though I assert that at that point simply putting a better telescope up is probably worth the marginal cost.

[imanidiot]

Re: So you're telling me there's a chance

I think I listed all 6 ;)

2 on EVA

2 spotters

1 canadarm operator

1 orbiter operations/pilot/engineer

The problem with doing spacewalks by venting the entire dragon is that everyone has to be in EVA suits, and there's no-one left in shirtsleeves environment to do the checklists and spotting and such. This seriously hampers operations. On top of that you lose a layer of safety, because what if you can't get the spacecraft re-pressurized? The pressure suits the crews wear during launch and docking operations are not suitable for long term wear (I think they're about 8 hour max) and depend on the capsule providing air and cooling. On shuttle they had a contingency plan of riding down in the cargo bay if the airlock failed, but that's not an option on dragon. AFAIK dragon also doesn't have the provisions needed to purposefully vent and re-pressurize on orbit, nor do I think the forward hatch is really suited for going on an EVA in terms of maneuvering space inside the craft (EVA suits are big, bulky and not a lot of flex). Gemini was a lot simpler in that regard, just open the hatch and stand up. Then sit back down and close the hatch (Which almost went wrong on Gemini 4)

The only thing I could imagine is either SpaceX designing a hab module that can be launched on an F9 containing an airlock that they can then rendezvous with in orbit, but while the design and development speed of SpaceX has been is impressive I doubt they'd have that ready before HST re-enters in 2028 (barring a robotic rendezvous and orbit raising mission). The cost of such an undertaking would probably approaching building a new Hubble class space scope too, so not really worth the effort. Alternatively IF starship development takes a leap and they have an orbital vehicle, I can imagine them outfitting an early model of this as such a hab module and use this or using an early cargo version to capture and raise the HST orbit to create some extra time. However, right now I don't see Starship getting developed that fast (even with the Human Lunar Lander system development)

[John Robson]

Re: So you're telling me there's a chance

2 spotters - I read " a spotter inside the shuttle to assist with their respective work" as one.

The airlock would need to be a serious bit of kit - I'm thinking something bigelow like (inflatable hab module) that can be used to move into, change into EVA suits, then exit from.

And yes - there were serious issues with several of the Gemini spacewalks... it was never really an option given the size of hatch they'd need.

The fact that HST has hard points for docking would make a robotic extension mission pretty easy.

I'm optimistic about starship development. Not quite Elon time optimistic, but optimistic nonetheless.

[imanidiot]

Re: So you're telling me there's a chance

Ahh, that's an easy misunderstanding and I see where you're coming from.

AFAIK it's usual for each EVA astronaut to have their own spotter working with them through their respective tasks and checklists, keeping track of parts and tools, mission time, etc, such that the EVA astronaut can focus on actually doing the work with as few distractions as possible. Especially on HST SMs where each astronaut had a long list of tasks to perform. In some missions the second EVA astronaut is more in an assistant/backup role and there might be a single person serving as spotter inside the shuttle or ISS, or one of the spotters might also be the Canadarm operator but HST service missions were so complex with so many tasks that I think it would be one of those missions where these are separated tasks and there really were 4 people needed inside the spacecraft.

[slimshady76]

Re: So you're telling me there's a chance

There's also the issue of maintainig the relative positions of HST and the capsule/launch vehicle. I guess a cable tether could work in a pich, but moving between them would still be at least risky... The Canadarm provided both a stationary framework and a mean to "crawl" between the shuttle and the HST.

[Flocke Kroes]

Dragon limitations

Crew dragons fly with the trunk empty so the trunk can passively stabilise the capsule during an abort. Without an airlock, if anyone needs to enter or leave then everyone else must be in flight suits while the capsule is depressurised. Depressurisation is intended to be survivable but I have never heard any mention of it being safe enough to include in a mission plan.

You could try a mission plan with a crew dragon and a cargo dragon. The full docking adaptor spec is androgynous but implementations often are not. Either of the dragons could be upgraded to a full spec docking adaptor so they could dock with each other, transfer some crew to cargo dragon, undock, then depressurise the cargo dragon so people can get out. Barely started and already the plan requires fearless astronauts and has no redundancy.

Try again with a cargo and two crew dragons. Lets call them Fear, Surprise and Ruthless Efficiency...

[John Robson]

Re: Dragon limitations

"Crew dragons fly with the trunk empty so the trunk can passively stabilise the capsule during an abort."

I didn't think that was the case - I know it's stretched for stability.

But SpaceX's website says:

"Dragon’s trunk not only carries unpressurized cargo but also supports the spacecraft during ascent. One half of the trunk is covered in solar panels that provide power to Dragon during flight and while on-station. The trunk remains attached to Dragon until shortly before reentry into Earth’s atmosphere."

And the cargo version has deployable panels, not panels on one half of the trunk, so they must be referring to the crew version - or have I got something really wrong?

[Pascal Monett]

Re: So you're telling me there's a chance

And I choose to hang on to that chance with all my might.

Gotta learn how to cross my toes . . .

[JDPower666]

So a bloke who has, by his own admission, zero inside knowledge of what's going on has declared it unrepairable? And his words are supposed to carry some weight just cos he's been to space?

Do you wanna do an article on how repairable I think the 737 max is? I've no inside information but I've been on a plane.

[Dave 126]

Poor analogy - the 737 engineers conduct their work on the ground. To do any physical work on Hubble, it is an astronaut's EVA skills that are required, plus the vehicle to get him to the work site. Mr Anderson has 38 hours EVA experience.

Now, fantastic things have been done remotely to space probes' computer systems, remotely updating software, routing around malfunctioning modules etc in the past. Let's hope engineers succeed on this occasion. Even if they do though, eventually Hubble may need some new components fitted.

[JDPower666]

Not quite, he has said its not repairable by any method, not just physical access. He has no knowledge of the system software or hardware involved, or of exactly what the fault is and having EVA skills does not make you a software engineer. If he'd said parts can't be replaced cos there's no way to do it, there'd be no arguments. And no article cos that's not exactly a revelation.

[Anonymous Coward]

38 hours - about one working week - doing anything other than spacewalking is a laughable amount of experience. Can you think of any other profession where a week's experience is even worth mentioning, let alone in hushed tones?

Fortunately, like aircraft engineers, astronauts gain most of their experience on the ground - or in their case, swimming pools.

[Anonymous Coward]

"Can you think of any other profession where a week's experience is even worth mentioning..."

Presidency... they walk in with 0 hours of experience :-/

Has anyone asked why is it worth repairing instead of cannibalizing or replacing? However, I do believe that it's probably a great experience trying to fix it just for the analytical thinking, similar to how I've spent countless hours trying to figure out how to create Smell'A'Vision (I'm CLOSE!!).

[Anonymous Coward]

If you can cannibalise it you can repair it, so there's likely no chance to do that. Building a replacement (the two NROs are both emoty of electronics and not really even potential replacements as their focal lengths are shorter) would take years and cost billions. Presumably, if Hubble is not fixable, that is what they'll start on.

[Annihilator]

So... while he has zero knowledge of what's going on, he'll understand the process and politics that govern Nasa, so I'd give him a bit of a break. What he's probably got insight into is the governance around the repair work (hence his "DOLLARS!" comment). They've probably got very finite resources (read 'people') allocated to the problem and probably consider that Hubble is now close to a zero-rated asset (ie it's more than paid for itself already) so unlikely to throw lots at it.

The updated statement shows the efforts required for the diagnostics and fixes - they've basically got to do dry-run tests of the failover procedures for each diagnosis they're making. All of that requires man-hours, access to labs, possibly simulators and that's just the stuff I can think of - all of that is resource taken away from other programmes, including programmes mentioned elsewhere in the comments of telescopes that will broadly replace Hubble's abilities (not James Watt).

I do hope that it can be repaired though, or at least repairable within whatever budget restraints they have.

[SealTeam6]

Faulty backup computer

As both the main and backup computers appear to be faulty I doubt that the problem is fixable without physically visiting the Hubble and fixing it onsite. We no longer have a spacecraft with a robot arm and an airlock that we could use for the job.

A real pity because Hubble is an excellent telescope.

[imanidiot]

Re: Faulty backup computer

Since both the main and backup show the exact same problem the thinking is now the problem lies outside the payload computer itself. As indicated in the article the main suspects now are either what amounts to basically a co-processor/IO board or a power supply issue keeping the memory and compute modules from working. There's spares for both of those aboard too, but "cutting over" to them is a very complex procedure nobody has ever done before (If I understand correctly they've never even tried this on the hardware analogs here on earth and getting it wrong will likely render the whole payload computer definitively out of action) so understandably they're taking their time to make sure they get everything right and try it on the hardware here on earth first before sending the commands to HST in orbit

[sebacoustic]

Re: Faulty backup computer

methinks the other possibility is that both computers went kaputt with the same failure mode: cosmic radiation could have killed the spare while it was turned off.

[imanidiot]

Re: Faulty backup computer

I'd find it strange radiation damage (which, by it's very nature is random) would damage both computers in exactly the same way.

[sebacoustic]

Re: Faulty backup computer

no problem, who does on premises computing anyway? They should just do a cloud migration.

[keith_w]

Re: Faulty backup computer

In addition to the up and down vote buttons the Reg should have an LOL button, which could work in conjunction with the other two buttons.

[Anonymous Coward]

Re: Faulty backup computer

A cloud migration is exactly what you don't want to happen if you're in space...

[Chris 239]

Re: Faulty backup computer

I want to upvote (because funny) and downvote too (because I hate cloud)!

[Swiss Anton]

Re: Faulty backup computer

I think the fact that we once had the ability to fix Hubble was a bonus. There will be no such options when it comes to the James Webb space telescope.

[Potemkine!]

He's dead Jim

Nothing is eternal, especially in space. 31 years of good service is a fantastic achievement.

[MyffyW]

Re: He's dead Jim

Damn it Jim, I'm an engineer not a doctor!

[bazza]

Build a New One

It's the quickest and cheapest way, if Hubble is broken and there's science demand that the JW telescope won't satisfy.

[imanidiot]

Re: Build a New One

WFIRST/Nancy Grace Roman Space Telescope is in development. It's a much wider field of view, but much higher resolution instrument and iirc covers similar wavelengths of photons.

[JK63]

Re: Build a New One

If I am remembering the details correctly... The Roman Space Telescope is going to function as a survey telescope and was originally named WFIRST. While the frequency range is comparable to Hubble, survey telescopes serve a different function.

The iconic images from Hubble are from pointing at a single spot in the sky for extended periods of time.

The NRO donated 2 units with 2.4m mirrors, one is being used as WFIRST. I have no idea what the plans are for the second. A direct Hubble replacement implemented with 21st century technology would be awesome! I don't know if I'll be around to see it.

[Anonymous Coward]

Re: Build a New One

The NRO telescopes have much shorter focal lengths and hence wider fields of view. That's why they're suitable for NGRST/WFIRST and not direct HST replacements. To change that you need a new mirror, and at that point you might as well just build a new telescope.

[Greybearded old scrote]

Re: Build a New One

They have a couple.

I never heard of anything being done with them, but it will most likely take a few years to fit them out for science instead of spying.

[imanidiot]

Re: Build a New One

See my reply above. It's on the books and in development.

The Roman Space Telescope will be based on the donated NRO spy sats

[Anonymous Coward]

Re: Build a New One

See several replies above. Shorter focal lengths -> not direct HST replacements, useful for survey telescopes.

[Brewster's Angle Grinder]

Who hasn't been there trying to fix a system while the peanut gallery make fatal pronouncements? "Your data is not gone until I say we're out of options. Now let me work."

[Binraider]

Hubble has a load of computers on board; not all of them for control (which are the ones that are failing), so there are options yet worth exploring to re-route workloads, where possible.

What's the line? "I don't care what a system was designed to do, I care about what it can do!"

A manned repair option is still a possibility, even in the absence of Shuttle. HST and ISS are on fairly different orbits, but it would not be implausible to send a large MEV and tow it to the ISS. (Yes, yes, I know, relatively large delta-V needed to do inclination changes... Not impossible though. And mass to orbit is comparatively cheap right now).

Much cheaper than building a new optical telescope... Unless theres a later-gen keyhole design ready to roll that can be repurposed just as HST was derived from the spy satellite design.