Hacker News new | past | comments | ask | show | jobs | submit login
Telescope That ‘Ate Astronomy’ Is on Track to Surpass Hubble (nytimes.com)
133 points by dnetesn on Nov 21, 2016 | hide | past | favorite | 80 comments



One of my friends works at the STSCI on various visualization and simulation tools (all web-based, written in a modern stack!) for planning missions for this thing. It's a really really neat project.

The telescope itself is "scriptable" using a (truly ancient) version of JS via a really old implementation that I've seen referenced but can't find right now. There's a lot of open information about the JWST, but it's not widely reported on. Definitely worth checking out if you're interested in space, technology, and the systems we actually deploy into the void!

Some papers: http://www.stsci.edu/~idash/pub/dashevsky0607rcsgso.pdf

http://arc-test.aiaa.org/doi/pdfplus/10.2514/6.2006-5747


Oh lord, JavaScript goes to space. You thought losing the MCO because of Imperial units was embarrassing? Just wait until this $9 billion telescope explodes because [] == ![] or something.

(This was supposed to be humorous, just making sure)


Seriously though, considering how deeply every other aspect of this project was deliberated (presumably, given the cost involved) - how they could possibly end up going with Javascript is bizarre. You'd think there would be a preference for arbitrary precision and against undefined behavior.


I can't find the evaluation paper they cited in the second linked paper, but it looks like it's using Javascript because they wanted a COTS script engine (as opposed to running compiled code) that was compatible with VxWorks, and the particular Javascript engine they selected (ScriptEase) was presumably the best option they found back in 2003. I'm not really sure what all options they would have had in 2003 though.


They should have hired Erann Gat to do it in Lisp.


Or just can't do a calculation right because single precision floats lack the granularity for it.


JavasScript uses double precision floats, unless you're working with float32 typed arrays.


Where does it say that the planning component is going to space? The web stuff seems to be for UI and is all ground based. i think that most Gui work these days is web based as we have forgotten how to do it in any other way.

Its like this: programmers who do ui move on to other things (mostly into stuff that gets you a higher sallary) and the new ones are left with the latest and greatest toolkits - all with their own set of weak points.


In both of the papers I reference, there is a javascript component running on the JWST.

Of course the mission planning stuff is all groundside.


Th spacesuit software runs on Node. The NASA guys are quite active on the conference circuit


What is the MCO?


This, I assume: https://en.wikipedia.org/wiki/Mars_Climate_Orbiter

Looks like it's worth the read.


Thank you.


In fact, one of the major reasons for the $4B+ budget overage was the rewrite from plain Javascript to Backbone in 2012, again to Angular in 2014, and to React last year.


Why Javascript?


Probably because Common Lisp has a bad rap at NASA.

http://www.flownet.com/ron/jpl-lisp.html


So why was lisp bad other than some bureaucrat saying it's bad.


Lisp isn't bad. Technically it was an unqualified success (and still is). It died at NASA simply because of politics.


Thanks for posting this interesting read.


Not that it answers your question, but JS dialects can be found in other strange embedded places such as Cognex's DataMan industrial barcode readers.


probably so astronomers can easily focus on their work - pointing the 'scope at stuff - and not worry about writing elegant/concise/reusable/succinct/$(feelgood adjective) code. I could understand it being used that way if it were for a limited scope (no pun intended) of operations, such as direction, focus, camera/sensor settings, etc. I'm sure they're not using JS to control thrusters or calculate trajectories.


Actually, us astronomers will never have to script anything with JS. They'll give us a GUI which will limit what we can do to strictly valid commands: e.g. point the telescope at this position in the sky and observe using this instrument for this much time.


There are plenty of languages that provide that same benefit of straightforwardness while also lacking warts and being even easier to embed; for example Lua, which was designed with express purpose of simplicity and ease of embedding.

The real answer is probably that in 2003 JS looked more interesting/good enough that it was chosen in this case.


So the Russians can hack it to self-destruct, obviously.


I can understand the sense of "we just spent 8.7 billion dollars on this thing; we'd better not screw it up", but I wonder what the replacement cost would actually be if they had to launch another? Assuming they spent most of the money on R&D, it might actually be relatively cheap. Maybe we could even launch a couple smaller versions just to have backups and do the science tasks that don't require a full 21' mirror.

There's an interesting contrast between NASA's and Elon Musk's idea of what space exploration should be like; the former spends most of its efforts on one-off projects, whereas the latter is focused on making things cheap and repeatable and achieving reliability by iterating on a design rather than getting it perfect the first go-around.

Both approaches are needed, and certainly NASA paves the way for others to come along and do the same thing cheaper once it's been proven.


"Why build one when you can have two at twice the price?"


great quote and i thought this too. there has to be another at least for spare parts or if the launch results in a rapid unplanned disassembly


Contact


We currently have the 80 cm Spitzer telescope in space that covers the same wavelength range, and the 3.5 m Herschel telescope also recently finished its mission at slightly longer wavelengths. So really to make breakthroughs in science to justify the cost of a space mission, we need the full 6.5 m telescope.


There is also SOFIA [1] (2.5m telescope on an airplane), which can do certain far-IR things (mid-IR as well iirc). For the VLT (8m ground-based telescopes) there is also VISIR [2] which is an mid-IR imager/spectrometer (and uses the same detector chip as one of the JWST instruments iirc).

[1] https://www.sofia.usra.edu/ [2] http://www.eso.org/sci/facilities/paranal/instruments/visir....


Oh yeah, those do cover the same wavelength range. The atmosphere of the Earth absorbs a lot of the light and has a strong thermal background though. The sensitivity of JWST is orders of magnitude better than ground/airborne observatories [1], so things that JWST want to do will be basically impossible from the ground.

[1] http://ecuip.lib.uchicago.edu/multiwavelength-astronomy/imag...


Oh I am perfectly aware. It's just not all projects require JWST-level sensitivity and they are somewhat less known.

Also I think SOFIA in visitor-mode is hard to beat when it comes to the coolness-factor.


While I will gladly rant about my issues with how NASA handles things: NASA and Musk are trying to solve different problems

Musk's goal is affordable space travel

NASA's goal was space travel (and exploration) period.

That means everything is a prototype. As for why we don't re-use them that much: By the time a project is over, so much has been learned that it and said knowledge needs to be applied to the next round.

In the case of space telescopes and the like: Repairs are made (if memory serves, Hubble was the one that famously launched with defects and needed to be fixed in orbit). As for launching more smaller ones: There just isn't a huge need for it. Getting time on a telescope is something even a grad student can do and the timescales are such that there is very rarely a "we need this NOW! Screw everyone else and block off a few weeks for us" situation. And if there is? Odds are a LOT of people want that data anyway.

One way to think of it is like a gaming PC. Some people buy mid-end and make incremental updates over the lifecycle of the machine. Others go all out and just build a monstrocity with an i7 every 4-6 years.


Why is NASA's approach a good idea?

If you have something cheap and repeatable, you can do a hell lot more science.


You may not be able to do the kinds of science that will test the hypotheses you want to evaluate. We could build a lot of Zero Gradient Synchrotrons for what the Large Hadron Collider cost, but the ZGS isn't anywhere close to being powerful enough to measure the properties of the Higgs boson. Another example: For less than USD 500 an individual may outfit themselves with an optical telescope more accurate, more precise, and more powerful than anything available to Galileo or Kepler or Newton, but such a "cheap and repeatable" device will never be capable of sensing the gravitational waves detected by the LIGO and Virgo laser interferometers. Also, that $500 Celestron is only possible because of all of the one-off telescopes that came before it. You have to start somewhere, and NASA is that somewhere for space science.


I was curious how the NASA budget compared worldwide. Summary: it is in fact a lot larger (2013[1] US $40b, China $11b, Russia $8.6b, India $4.3b).

[1] https://www.rt.com/usa/199480-space-budget-nasa-report/


Just because I can get 1,000 10 dollar optical microscopes for the cost of a single 10k electronic microscope does not mean that an equal about of science can be done.


This paper, from 2014, says that the Hubble space telescope is oversubscribed by 4:1 to 5:1. [1] That's a higher acceptance rate than most government funding grants these days (unfortunately). Perhaps it suggests that there is still a lot of unmet demand for good science to be done with it.

[1] https://arxiv.org/abs/1409.3528


Until someone figures out how to duct-tape 1,000 optical microscopes together in a giant interferometer, I guess.


> what the replacement cost would actually be if they had to launch another?

I wonder if they considered just launching a repair-buddy satellite ? So basically a robotic toolbox with a ton of spare parts that could cover 95% of the possible failures of the thing. Then if they get it up there and find out the panels wont open, or the mirror gets stuck, they could pop it open and get it ready before they send it to the lagrange point.


You might as well just send up a second one, rather than send something that is 95% of the first one plus a very complicated system of robot arms to do arbitrary repairs and replacements.


I believe NASA is considering this, but the technology won't be possible in the next few years.


It is interesting, but also perfectly appropriate within mainstream economic theory that a private corporation be responsible for the latter goals, and the former are with appropriate government subsidies.


Hubble design was designed to be compatible to some extent with military spy satellite. Perhaps here it is again the case ?


I can assure you that there is very little design commonality between JWST and surveillance satellites. JWST is a very unique mission that is completely different from that of spysats.


This is probably unlikely as JWST is designed to fly in L2 and not in a geosynchronous orbit. Additionally, at the wavelengths it is operating at (mid-infrared), if it pointed towards Earth, it'd just see the heat of the Earth's atmosphere. Probably not very great for spying.


Hubble design is that of a spy satellite.


Last week Dr. Michael Ressler at JPL gave a talk about the Webb telescope, mostly focused on the mid-range infrared imaging sensor:

http://www.jpl.nasa.gov/events/lectures_archive.php?year=201...

If you happen to be in the LA area, JPL hosts these talks once a month at their facility in Pasedena. They're free and open to the public, and they're always interesting. And if you can't attend they're also posted to the JPL website.


Is there a way to get email notifications when they're holding an event?



> A House subcommittee once voted to cancel it. Instead, the program was rebooted with a strict spending cap. ... The major change, said Jonathan P. Gardner, the deputy senior project scientist, was to simplify the testing of the telescope.

Ugh, that does not sound encouraging. Organizations love to cut testing and QA. After all, all they do is cost money, cause delays, and 'create' problems. Cutting them is an obvious way to bring a project under budget and on schedule.

Does anyone know what changes were made to testing?

We're going to be unhappy, and the advancement of astronomy delayed a long time, if something goes wrong. I don't see the next President and Congress wanting to raise revenue and spend more on science.


I believe the initial estimates for the budget were extremely optimistic. I believe after Congress almost cut funding on it, they came up with a very realistic budget and schedule and have been following it very closely since. I know people who are working on the testing, and it seems like they aren't cutting corners. So I wouldn't worry more than normal for deploying a 6.5 m telescope in space.


Thanks; that informs the discussion.


Is the top comment in this thread seriously someone claiming that "the telescope is "scriptable" using a (truly ancient) version of JavaScript via a really old implementation"? It's like the author of The Martian united with some oldschool C programmer to create a horror scenario.


"Ate astronomy" means it costs more than all the proposed big ground-based telescopes combined. The Overwhelming Large Telescope was considered too expensive at $1.5bn.

It's the F-35 of astronomy.


I've never been able to wrap my head around the scientific premise of hoping to observe radiation from the early universe. I studied physics in university but I was very crap at it and never graduated, so that might explain it.

The reason I can't is because I assume that the EM radiation that resulted from the Big Bang happened before we did (we as in our neck of the cosmos), and that radiation left the epicenter of the Big Bang before we did. Assuming that radiation has been moving radially outward since then, how can we hope to observe it yet it has a head start on us?

The Big Bang is not an event that happened in the universe. It's the event that brought forth the universe.

It's possible for us to observe, say, a supernova that exploded even before our solar system, for example, existed. But that supernova was an event in the universe, and it's light may take billions of years to reach another part of the universe and for entities in that part of the universe to now observe it, even if those entities didn't exist at the time the supernova exploded.

That I understand. But the Big Bang? Isn't that different? Isn't trying to observe the Big Bang a bit like trying to observe your own birth?


Disclaimer: I Am Not A Physicist.

The Big Bang didn't happen in a specific location some distance from us - it created all of spacetime; the epicentre was at every point[1] in space.

[1]: for want of better phrasing, given that we don't know if spacetime is discreet or continuous.


> I've never been able to wrap my head around the scientific premise of hoping to observe radiation from the early universe.

It's not "hoping." The humanity actually observes the radiation since 1964:

https://en.wikipedia.org/wiki/Discovery_of_cosmic_microwave_...

The guys got Nobel in 1978 for that, almost 40 years ago.

What you have to understand is that looking in the distance means looking at the older events. Eventually, we "see" the "glow" of the Big Bang (actually receive it with the radio telescopes). It's a specific glow of one specific phase of the development of the Universe, not of the "zero point" but the exact moment when the "baby-universe" was at the state to emit these photons. That happened some 380,000 years after the imagined "zero point of time."

The confusion that you have is that a lot of those who know that the Universe already existed before that glow that we observe are the researchers who when they speak about the glow aren't interested in the phase that will never be visible (and don't worry, there are also other researchers who are). So the popular accounts shorten it to "we see the Big Bang." Which is also true, as nothing else but the Big Bang can explain the glow. It's not "I said he said" it's an immense amount of very precise calculations of different very precise measurements that all have to fit, and the alternative "explanations" simply don't give the results. So the Big Bang it is. It's comparable to the fact that the humanity started to produce the globes ("the 3D models of how the Earth looks like from the distance") almost 500 years before it was really possible to see the Earth from the distance. It was all a lot of the measurements and a lot of the calculations.

https://en.wikipedia.org/wiki/Globe#History

And it matched (the better the measurements and the calculations were, the more it matched). If you want to learn more about the Big Bang radiation, it's easy to find, it's called CMB:

https://en.wikipedia.org/wiki/Cosmic_microwave_background

"The CMB is a snapshot of the oldest light in our Universe, imprinted on the sky when the Universe was just 380,000 years old. It shows tiny temperature fluctuations that correspond to regions of slightly different densities, representing the seeds of all future structure: the stars and galaxies of today."

It's fascinating.


http://spacenews.com/op-ed-the-approaching-battle-over-the-j...

This article contains a graph of the JWST programme's appetite.

While I'm looking forward (eagerly!) to that telescope's imaging, I am also really looking forward to what other projects become possible once it is no longer on the teat. (And the sooner the ISS burns up in the upper atmosphere the better for the exact same reason).


The deployment of the telescope looks quite a complicated process.


Yeah, honestly, I think I got anxious for the engineers just watching it. If any of that unfurling just hits one minor snag, it's 8.7 billion up in smoke.

Does anyone with more information about the telescope know how they are able to keep such tight tolerances given how the telescope deploys? For example, I remember the original Hubble telescope had a spherical aberration that amounted to less than the width of a human hair. Surely the way the side mirrors deploy, and the way the front mirror unfolds into place, can't have tolerances that tight. How does the telescope account for that potential variance?


The Hubble's problem wasn't tolerances but the mirror got produced with the wrong numbers. For the wrong numbers it was within tolerance :-). But yes, I am not sure if I could handle the stress on working on something like this for years and then watch it deploy without any possibility of fixing problems. This must be incredibly stressful.


Bringing up Webb is a six-month process after it gets on station. According to [1], in months 2 and 3 the mirror positions are adjusted to bring them into optimum optical alignment. I haven't found anything on the type of actuators used for this.

[1] http://jwst.nasa.gov/faq.html (in section 16)


https://m.youtube.com/watch?v=bTxLAGchWnA

Launch, transfer to the L2 orbit and deployment will take ~30 days.


Please use a safe rocket to get into space. That would be one stressful rocket ride for the project scientists on earth.


I was thinking the same thing - JWST is slated to ride an Ariane 5, which has had something like 75 consecutive successful launches.


As much as I admire SpaceX...not the job for them yet.


Not this satellite. Not yet.


well, you can't put a price tag on knowledge and discovery :)


8.7 billion USD.


and beyond!


So, 8 nuclear submarines, give or take? Or a couple dozen F-35s?

I'd say it's a better investment.


I know you're being facetious but the program to create the f35 was $55bn and that doesn't even get you a jet off the production line. And we've build jets for decades.

NASA must be doing outstanding work for what they get out of the money.


Virginia class! you could get 3 nuclear submarines.


This may be obvious to everyone else, but today I learned that https://en.wikipedia.org/wiki/James_E._Webb is not the same person as 2016 Democratic primary candidate https://en.wikipedia.org/wiki/Jim_Webb

I feel a little better know. I was wondering why we would name a space telescope after that guy.


TLDR: The James Webb that the scope is named after was the NASA administrator during the time of Kennedy and the heroic days of the US space program.


Perhaps we should start naming space telescopes after politicians, so that NASA's budget has an easier time in congress.


This is precisely what happened. Edwin Hubble was an astronomer, James Webb a NASA administrator.


That made me wonder where the Jodrell Bank Observatory got its name from, turns out it was an archer: https://en.wikipedia.org/wiki/William_Jauderell


The Robert Byrd Green Bank Telescope

https://en.wikipedia.org/wiki/Green_Bank_Telescope




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: