Back when I was in grad school Andreas Ghez gave a colloquium talk about her work in my department. The basic idea behind the technique she used is that the resolution from images in modern telescopes is limited by atmospheric distortions. If you can get rid of the distortions, you can get much sharper images, and in her case, resolve the stars orbiting Sag A*.
Adaptive optics is one way to reduce distortions from the atmopshere. With adaptive optics, you observe some star (or more commonly a spot of light created by a laser) and rapidly distort one of the mirrors in the optical path to compensate for atmospheric fluctuations.
Andrea Ghez went further than this, though. Atmospheric fluctuations are fast, but they are not instantaneous. If you can take an observation faster than the atmosphere changes, you will get a much sharper image. You can sort of imagine that an any given instant, a star is a very sharp point of light (or actually several points of light due to refraction), but over time, these points of light move around randomly over an area to create a smeared out image.
Of course, there a number of huge technical difficulties to taking these very rapid images. But if you can manage it, you can resolve the orbits of stars around the supermassive black hole at the center of the Galaxy. (This technique is called speckle imaging.)
I also remember her joking that when most astronomers write observing proposals, they'll typically apply for a few nights of time on a telescope. But she was applying for a few seconds of observing time over the course of a decade.
Yes, I think this was more of an engineering problem than a physics problem. Building telescopes with enough sensitivity to be able to resolve a star ~100k light years away, and the necessary software to process their raw data, is no small feat.
> Andrea is fond of pointing out that one of the reasons for her success has been this tight and rapid loop between the needs of the astronomers and the engineers who respond to the challenge.
As an engineer I always had this fantasy that it would be super cool to do engineering at a place like this or (specific to my fantasy) CERN. As a product engineer I would like to think that making something one-of--engineering for engineers and scientists-- would allow me to focus on just getting the thing to work in one specific setting. No dealing with longevity or RMA or other similar constraints.
I guess it's a fantasy about practicing "pure" engineering and design.
I call it a fantasy because I'm not convinced that that is how it would play out in the real world. I would love to hear from anyone familiar with this field.
I work for Keck as a software engineer. What do you want to know? It's not all upside, working for pure science like this. We like to "fly the prototype", so to speak. Despite the challenges of it being a 25 year old pair of telescopes that are constantly climbing the far wall of the bathtub curve, it's a damn cool place to work at.
What do I way to know, eh? I guess the real thing I would like to know is how your job is different from product development for the public. I very much realize that it's an open ended question and is not answerable unless you've had experience doing what I do.
That said, here are some concrete questions. Do your projects get cancelled regularly for reasons other than engineering merrit? Do you get into a mode where you have to maintain projects that are "past their prime"? This one might be weird, but do you feel needed, as in do you feel like your contribution matters?
While I am asking these I realize that my fantasy bubble is busting and that maybe project requirements and management might not be that much different in our worlds, but I'd rather hear from someone in that world.
1) Is our job different from product development in the real world? Yes. I have worked for big business, I've done startups. It's categorically different here. We are responsible to maintain a set of telescopes and do science every night, that's our priority. Development of new capabilities is important but comes second after getting the "production" systems ready every day for observing.
2) Do projects get cancelled? Not from my perspective, things slow down sometimes and business priorities shift. We have new capabilities that come online over time and we spend development effort to get that integrated into the facility. It has a schedule and we spend our allocated time towards that.
3) Projects past their prime: yes, some things get really old and are built from unobtanium. Ever hear of a transputer? We make heroic efforts to keep that old stuff running and for the most part we get it done.
4) Does our contribution matter? Absolutely. Everyone is on task and really does a good job of keeping things running smoothly. I've been in industry for 25 years and I can see our team is on point and we are all doing a great job keeping it going. There is an enthusiasm I've not seen in other industries here. That's not to say it's all roses. We have shitty nights where everything seems to go wrong. That said we get it fixed the next day or so and keep on truckin'.
Don't take it for a fantasy, but it's a pretty neat situation we have here. If you are a software engineer we have a few openings. Working in Hawaii has it's peculiar challenges and it's not for everyone. Apply! You never know what could happen once you throw your hat in!
Replacing old stuff is a challenge. What is the effort to spin up a whole new widget that does the task, and how do we test it sufficiently to prove it has the same results as the old one? Do you need to run it on sky? If so, can you switch it on/off quickly so we can run the old stuff for science and run the new stuff on an engineering night to prove equivalency? It's difficult to replace but not impossible.
The best outcome you can hope for is some new instrument with better capabilities comes online before the old one dies horribly. For computing systems, we have projects to replace the old stuff with new machines and new OS's and take the time to work out the kinks and bits of tribal knowledge we realize all too late that we lost when so-and-so retired. This is not atypical in the computing industry.
In summary: yeah, things can be hard to replace. There's inertia associated with old systems you have to balance the thought to "string it along, ebay/graymarket a replacement" versus reinventing it.
I have done instrumentation work in US HEP. It is fun to do stuff nobody else (or very few) people are doing, in labs with very cool stuff, working with extremely smart people.
However there will be a few extremely smart jerks, and leadership seems to spend a lot of time applying/fighting for funding. I definitely got the sense that there was a pecking order. The PhDs are at the top, engineers and technicians don't get as much respect. Also in these days I hear that US national labs are moving to a more "flexible" model. In days past an engineer might be assigned to an experiment for the duration, but now leadership wants to be able to assign resources on a more fine-grained basis. (Though I think this is becoming more the case in industry too.)
All that said, if I was in a position to not mind a pay cut, I'd definitely consider going back to pure research.
I'm acquainted with someone that has a job that sounds like what you want. He's at a smaller university, runs a small machine shop, and mostly spends his time trying to creatively fab together whatever the science departments want constructed.
I’m a neuroscience PhD and this is the majority of what I end up working on. Large one-off systems engineering projects to enable new instrumentation for performing new types of experiments. It’s good fun, but there’s always a tension between doing something fast or right, and tape somehow always sneaks into the final product.
Interesting! I always thought, it would be just logical, that at the center of (at least) a spiral galaxy a black hole must reside, since it looks just like water draining down into the kitchen sink. But for science, obviously, obvious is not enough. It must be proven.
I do not like this clickbait title. Just tell us what the experiment was. This seems designed to build a narrative about something rather than telling people about an interesting experiment.
Barely. I would have liked to know much more about the details of the experiment, but also why people were initially so skeptical, and what happened to make these people come around.
Do you guys not read the article at all? How is it clickbait to describe what he did - the Nobel Prizes aren't that interesting as science goes (good science is good science from the day it was published until now), the human side is what draws many to an article like this.
There is the "Science is interesting, and if you disagree you can fuck off" approach, but why not give it some style.
Seconding this. There's a grand total of two sentences vaguely describing what she wanted the instrument to do but zero information whatsoever about how that differed from its normal operation, why nobody thought it would work, or why it posed a risk to the instrument. This article is just a bunch of compliments toward Dr. Ghez.
Well duh, it's an article nicely applauding the person who just won a significant prize/recognition, if you wanted to find out more about the research you're sitting in front of a machine which can help you find that.
Need more context why it considered it is a waste of time and how she overcame what would make a great story. Instead just apple shining. Sigh. Should be a great story like the one why he kill Pluto.
Something like the first comment here but more. Even the infrared one would be a good story.
The article is written like an 'in memoriam', except that Dr Ghez is actually alive.
Here's what the clickbait headline was referring to—
"NIRC [Near Infrared Camera at Keck Observatory] was never designed to do what Andrea needed—an ultrafast readout of images and then a restacking of the result to remove the effects of the atmosphere’s turbulence. But she was not to be denied—and we made the changes. And it worked!"
Tangential, but I was surprised that the article kept referring to her as “Andrea” instead of Dr. Ghez until the end where the author stated that they’re friends. From that perspective the author likely meant for it to have a warm, friendly tone. At the same time, given its publication for a wider audience and the occasion, it is perhaps more fitting to refer to Dr. Ghez more formally.
She might prefer being called by her first name and perhaps doesn't even like formal address. We don't know and if she isn't complaining about it then I don't know if we should.
I work at the Keck Observatory. The author of the article is the director. She's been working with us for 25 years and is as close to family as you can get. We're all on a first name basis here!
Good. You do you. My point is that you appear to be taking offence on her part when you don't know if any offence was actually caused. Or indeed whether, as her friend, the author knows that she prefers to be referred to as Andrea, even in the written form.
I'm not taking offence on her part, I'm saying that there's a degree of propriety for certain things and not acknowledging that for a wider audience is incorrect for the format. MIT editorial style guide for instance:
> The best way to deal with professors is to use the capitalized Professor Irwin (second reference) for formal writing. In more informal writing, depending on the audience, a professor may be referred to by his or her first or last name on second reference.
> Use Dr. in first reference as a formal title before the name of an individual who holds a doctor of dental surgery, doctor of medicine, doctor of optometry, doctor of osteopathic medicine, doctor of podiatric medicine, or doctor of veterinary medicine: Dr. Jonas Salk.
> If appropriate in the context, Dr. also may be used on first reference before the names of individuals who hold other types of doctoral degrees. However, because the public frequently identifies Dr. only with physicians, care should be taken to ensure that the individual’s specialty is stated in first or second reference."
Note that there is consideration for how the public reads these as well. If it's for more than just a personal note, then a more formal tone is likely more appropriate. Which is both obvious and all that I was saying.
i think goku99 was being sarcastic but thanks for clearing the downvoting up. obviously he's new around here... and my comment was a remote sensing joke.
clearly downvotes are used to suppress views that certain ycombinator members don't like. i would expect that behavior from china and russia, sad to see it here.
In the comment head, click on the time submitted (it's actually a link), just after the author. For example "9 hours ago". You will be taken to a readable version of any downvoted comment.
Adaptive optics is one way to reduce distortions from the atmopshere. With adaptive optics, you observe some star (or more commonly a spot of light created by a laser) and rapidly distort one of the mirrors in the optical path to compensate for atmospheric fluctuations.
Andrea Ghez went further than this, though. Atmospheric fluctuations are fast, but they are not instantaneous. If you can take an observation faster than the atmosphere changes, you will get a much sharper image. You can sort of imagine that an any given instant, a star is a very sharp point of light (or actually several points of light due to refraction), but over time, these points of light move around randomly over an area to create a smeared out image.
Of course, there a number of huge technical difficulties to taking these very rapid images. But if you can manage it, you can resolve the orbits of stars around the supermassive black hole at the center of the Galaxy. (This technique is called speckle imaging.)
I also remember her joking that when most astronomers write observing proposals, they'll typically apply for a few nights of time on a telescope. But she was applying for a few seconds of observing time over the course of a decade.