Campus research collaboration and NSF grant opportunities (John Hoinowski & Brian Murphy)

>> Our first presentation is about campus research and NSF opportunities related to that We have John Hoinowski and Brian Murphy from Butler John is a senior network and security engineer His IT career began over 16 years ago in the U.S. Army In addition to the military, he has served as an IT resource in mortgage insurance, healthcare, cloud service provider, and higher education industries John is currently working toward his MBA at Butler and strives to be a bridge between business and IT He’s working with Brian Murphy, who is a professor of physics and astronomy And I hope I have this correct — six years ago Brian secured funding allowing for Butler’s membership in the Southeastern Association for Research and Astronomy It’s the telescope consortium and this particular endeavor is one of the science drivers in Butler’s interest in applying for NSF funding this year So join me in welcoming John and Brian to the stage [ Applause ] >> Can you guys hear me OK? All right So, I’m a network guy, and they usually stick us in caves, so excuse my lack of presenting ability Yeah, so for the next hour we’re going to talk about campus research, collaboration and NSF opportunities A little over a year ago I-Light and Internet2 approached us and started talking about opportunities that NSF provides, and some of the success stories from Wabash [assumed spelling] and Earlham [assumed spelling] They were very inspiring, so we decided to look at the NSF, and we’ve been working on it for about a year, waiting for the 2016 information to come out And it came out two weeks ago, so good timing for us Presenters today will be myself, senior network and security engineer I started at Butler about a year-and-a-half ago — very good place to work I’ve had the opportunity to meet the next presenter who’s Brian Murphy, professor of physics and astronomy He’s been there for about 23 years I mean this guy’s part of the groups that talk about whether or not Pluto is a planet — very exciting I’m an astronomy nerd myself, you know, I watch Ancient Aliens and How the Universe Works You name it So the agenda for today is, you know, I’m just going to give a quick intro here At some point we’re going to Brian up in a few minutes He’s got some really good slides that talk about in-depth about what his research is And I know a lot of you who may be perusing the NSF grant solicitation You might be wondering what you have that would meet the requirements of the National Science Foundation for these grants So hopefully this will be a good example for you guys to think about how, you know, what research you have that meets the requirements After that, Brian’s going to hand it back to me Well he’s going to end at first talking about some of the pain points from his perspective Then he’ll hand it back to me and we’ll talk about pain points from an infrastructure perspective — storage, network After that we’ll talk a little bit about the process we used to make the rubber meet the road You know, it’s one thing to have these — have your research that you’ve identified, which is pain point one — how do you extract that information? How do you reach out to your faculty and staff to get all that And then, part two is how do you bridge that between some infrastructure requirement and actually make it a reality so that you can submit a grant request And then we’ll have a little bit of time for questions based on our time availability So a little bit about Brian Murphy He’s an extremely busy guy for one He’s a member of the Southeastern Association for Research and Astronomy, and these are universities across the country I know they have some in Chile I checked the website I can’t remember exactly where more of them are, but there’s about 13 of them, I believe And Brian’s going to come up and talk a little more in-depth about himself and some of the collaboration he’s doing with astronomy,

which is very important to emphasize in this grant So without further ado, Brian Murphy [ Applause ] >> OK. There we are Thank you, sir There we are OK. First off, I think he told you I’ve been here at Butler at least for 23 years Before that, actually, I was a graduate at Indiana University, so those of you from Indiana University you might work with their astronomy department And in particular I worked with Richard Durisen — Haldan Cohn, if the IU group knows any of them But the — they got me into theoretical [inaudible] I’d thought I’d start out as an observer Went to theory — and theory is all about obviously computational physics, so it involves a lot of super computing and things like that And obviously todays — what I used back then is equivalent to almost an iPhone as far as power And the reason I did that, stayed in theory, is because back in the 1980s we were still making the transition from photographic plates to digital cameras Astronomers were among the first — astronomers and NASA were the first to use digital cameras We’ve been calling them CCD’s forever, OK? And they started to come into play in the mid-1980s By that time I’d made my switch to theory But my first postdoc was in Utrecht in the Netherlands My colleagues there said “You know you can get some time on these big brand new telescopes in the Canary Islands — you and your colleagues from the states, since you are now in Europe.” I said “OK I’ll apply.” I had no idea what I was doing I was a theorist OK? I knew how to use a computer, but didn’t know how to use a big telescope I didn’t even know how to do digital imaging Luckily I had some very good people surrounding me from all across the world when we were doing these observations And that was my first foray into real observing with these — the largest telescopes at the time in the world Well, after that I came back to the States — had a postdoc at Cornell University, then came back actually, by chance, to Indiana, here at Butler University And the interesting thing was at the time I’d read the ad for Butler University, and it said OK, we’re looking for someone who’s a theoretical astrophysicist who knows how to operate a telescope and be an observatory director And I said “Well, I do know a lot about telescopes and I know how to run smaller telescopes, and I am a good theoretical astrophysicist, at least I can, you know, program my way out of a bag, or code my way out of a paper bag,” all right? So OK, it looks like they’re looking for me Evidently they were I was the ideal fit because they wanted someone who knew a bit about observations, who could get their telescope, which is — at the time was the largest telescope in the state IU temporarily had the advantage of us, and now I think they sold off the telescope they had in Morgan-Monroe State Forest, where back [phonetic] being the largest telescope in the state It is a one meter mirror And since the time I came here we’ve secured several grants And what happened in 2002 — I was reading the Financial Times, actually a colleague of mine was, and he happened to post something on his bulletin board outside his office that was listing the alumni in our department, who have gone on to get PhD’s And one of the alumni said that he had stock options worth a billion dollars Then, of course, that catches your eye at a university because most of us that go into academics, advancement departments don’t go after us, OK — because they know we’re in academics Don’t make a lot of money But those who don’t go into academics and go into industry usually make a heck of a lot more This guy did very well In fact, his company is called Finisar I don’t know if anyone’s ever heard of it It makes most of the fiber optic switches in the world today And he graduated from Butler in 1974, and now he’s a philanthropist, and so we contacted him, and after much going back-and-forth, he’s also an astrophysicist, PhD And so, he talks like an astrophysicist The only idea [inaudible] and I had to sort of steer him back to what I wanted Eventually we sort of came to a conclusion and said well, why don’t we try to get into a telescope consortium That would be great for a school like Butler University, where we don’t have a full-fledged giant astronomy department, say like IU But we can put ourselves in the consortium that makes us a very large astronomy department, with access to telescopes And so, that lead to him giving us a donation and creating an endowment to pay our dues to join the Southeastern Association for Research and Astronomy And so, after we joined this — and it was actually 2008 I believed we joined,

the first telescope they had at the time was Kitt Peak This was funded in 1994 It was actually those of you at IU — if you’re familiar with the telescope they have at Kitt Peak Arizona, it is a four meter telescope This telescope used to be on the side of that, so to make way for it they had to move it And the National Astronomy Organization told us if anyone wants this scope it’s yours, but you have to move it and rebuild yourself a new observatory Well, four universities in the American Southeast here took that over, got a National Science Foundation grant, and built this observatory here with this first telescope that was on Kitt Peak Arizona This is the first national telescope ever to exist in this country built in 1959, I believe So, by national I mean that anyone could use it if they applied for time Prior to that, universities had their own telescopes, and we still have this program today where you can apply for time on national telescopes It’s starting to shift back more to private universities again, and that’s why this was an important step for us So, at the time, there was only one telescope out at Kitt Peak Arizona, a one meter telescope similar to the size we have at Butler University But one of the things that enticed us to join was they said well also with your — let’s call it admission fee, we can get you — we’re going to get another telescope refurbished that belongs to La [assumed spelling] Observatory in Cerro Tololo, Chile That would give us both sky coverage of the Northern and Southern Hemisphere So, that led us to getting another observatory in this consortium That led to 60 nights a year on telescopes Now you might say well that’s a lot of travel time Well, we don’t travel to these observatories We use them fully remote I can use them from my laptop, watching TV, and, in fact, I have — sometimes with a margarita, sometimes on my deck And but — anyway, now we are up to actually three telescopes We just commissioned the Yacob [assumed spelling] [inaudible] telescope We have become one of the premier consortiums of using small telescopes, taking old telescopes, refurbishing them, putting great cameras on them, and you can see now that we have — we’ve got a pointer here, the universities Is anyone from Ball State here? OK. How about Valpo? OK. All right, he’s got the whole crew right there You guys are in our consortium, too, and I consider them part of our big astronomy department OK? I know almost every astronomer in this consortium, which is about 50 of us, and we do act like a large astronomy department, exchanging data between each other And you can see the list is getting longer and longer We’re probably adding two more people, or two more universities to this in the near future And now there’s competition to get in it And just a quick thing about this consortium, my colleague and I, [inaudible] Hahn at Butler University He’s another astronomer there We were doing a calculation because we were looking at the WIYN consortium, which is what IU belongs to, and that’s nothing against IU, but they have a very large, very nice telescope, which I’ve used out at Kitt Peak We were calculating the amount, the cost per photon to observe That is they have a larger aperture It collects more than photons for their telescope We have three telescopes of smaller aperture But if you add the cost per year, ours was one-quarter of the cost OK? Per photon coming into the telescope So in that way, we are one of the most cost effective consortiums in the word, if not the most cost effective consortium in the world in collecting light from space So anyway, the telescopes we have — first of all, the one Kitt Peak out in Arizona, and we actually send students We usually got out there once every year or so with students, just so they can get the feel for what it is observing on a mountain top, which they suddenly realize is roughing it Because at Kitt Peak Arizona they have radio telescopes also, which means you cannot have any Wi-Fi, which means they cannot use their phones, which means they’re separated from the outside world and it’s fun to see that OK? To see college age students — my phone, what? I can’t — I said no, you actually have to take this cable and plug the computer into the wall It’s called an Ethernet cable And they’re totally amazed that you actually have to do that But the Cerro Tololo telescope is sitting down here, and this is also the site — near the site where they’re going to build what’s called the large survey synaptic telescope, which is going to create — I think its 10 terabytes of data a night, all right? And that’s going to have to somehow get from here up to North America The same thing happened with our telescopes We have to get our data, and we get 80 nights a year on these three telescopes, but we have to get out data from these three telescopes to Butler University The bottleneck is usually not necessarily Butler University It’s usually getting out of Cerro Tololo here for instance The reason is is because everyone is putting telescopes all over the world, and remote observing is now the hot topic to do They even have robotic telescopes that do your observations for you And everyone tends to push their data through the pipe, off the mountain top,

at the same time a day, right after sunrise And there is — I think South Korea put a telescope on there that would pushing through several terabytes all at once, which was slowing everything down And there was one point — I think I was getting something in the neighborhood of 50 bytes per second download to Butler And I sent an email to the director, who is an IU grad also, and he said — and I told him I said “You know, I could actually fly down there with my flash drive, get the data and come back to Butler from South America, and it’s still a faster rate of getting data across.” OK? Nothing like an SUV full of flash drives going down the highway, all right? So anyway, so one of the issues we deal with — and, by the way, remote observing is wonderful I mean it is absolutely wonderful My students realize that when they actually have to go to the mountain top — it is sort of literally roughing it They have to sleep in something that looks like a dormitory, not even as good as — nowadays they live in Hiltons practically, the students do But they get out there to these 1950s dormitories and realize oh, OK, — watch out for scorpions and occasional mountain lions, things like that But this telescope here, just to tell you one of the nice things about astronomy — they tend to be in wonderful locations — our observatories So it is nice to travel to these And I was actually at this one here last fall We traveled out there This it the board of directors — some of them Not everyone could make it to the dedication of this telescope We are now hooked in with the Spanish They are hooked in with our organization We send students there now And so, this telescope here — it looks a lot smaller, but it’s about a 50 foot high building, has several levels in it It’s basically an academic building in its own right And this sits 8,000 feet This is the Atlantic Ocean down here, 8,000 feet above the Atlantic Ocean One of the premier sites in the world, and the fact that a small school like Butler, say Valpo, Ball State, we can all get access to these facilities is really remarkable Now, one of the things that we deal with in choosing our research program, you know, remember, I was a theorist And we were looking for something that was good for Butler Well, you know, I had — by the way, our donor, Frank Levinson, [assumed spelling] also gave us a highly parallel computer, too, for theoretical computations So, I had that Now I had this and I said OK, what the heck am I going to do? I’m used to using these massive telescopes that can see very dim objects But you typically only get three nights a year on those Here, though, I suddenly have 80 nights a year on things that I can’t see real dim objects, but I can see a lot of medium — let’s say dim objects OK? And so, the issue was I had to find a research program for that, and given the amount of time I had, I figured OK, the best thing to do is look for something that varies in time There are a lot of things out there in variant time and that’s tell you what’s happening — say with a star, with an active galaxy, with quasar, something exploding known as a gamma ray burst But that was the thing we decided to go after Now, you just don’t point randomly at the sky and say OK, I’m going to look at that star because that star might be quite boring, similar to our sign, which only changes by .1 percent brightness over the 12 [phonetic] solar cycles We wanted something more than that and we wanted more than one star So, one of the things we picked were globular star clusters And globular star clusters are literally, you know, it tells you a lot Globular — it’s sort of like globular, but it’s a glob of stars is the way to think of it And when we actually do the imaging it typically produces — it’s fairly paltry — two gigabytes of data, but we have to get that across to Butler And the images are downloaded The last student who observes for the night, their task is to start the download So when I come in at say — in the summertime usually is when we observe, primarily — when I come in I said I want that data sitting there, ready to be processed That is to get rid of all the image defects, things like that, so we can begin analysis So, this is a typical globular cluster You tell me which star is variant there? If I gave you 1,000 images, can you think out of these 20,000 stars — pick out which stars were variant, how they’re changing brightness? And that’s the problem with this And that’s why we pick this, though Here’s 20,000 stars we have to choose from, but it’s not an obvious thing to see which ones are actually variant So finding the objects is very difficult In the old days I’d think how the heck did they do this without digital imaging? Think of having a photographic plate, literally [inaudible] motion on a plate, you slide it in the back of the telescope, only 5% efficiency compared to 95% — light deficiency with these And then, one photographic plate, OK? That might be 20 minutes Grab another one, put it in there, put another one, slide another, then oh, by the way, you have to develop these I did that in graduate school That was the last time I said I’m doing observational astronomy OK? Because I had to develop a photographic plate And so — this I said — and I contacted my old PhD advisor He’s not old, but PhD advisor at IU and I said, you know, we both worked on globular clusters, but more of the theory of what goes on, how do stars move around, how do they collide

with each other in these dense, stellar conditions? And we contacted one of our colleagues in San Francisco State University, and she told us — she said well, there’s this program, unfortunately it was called ISIS, OK? I-S-I-S. OK? I’m going to use the word difference semi [inaudible] OK? Because I’m sure an FBI file a mile long from all the emails I have with ISIS in them But anyway, so the thing was — what ISIS does — it is literally difference image analysis OK? Imagine the following — if we take two images taken at different times, OK? There we are At different times — here’s one image, say here’s another one maybe an hour later, and we subtract the two If everything’s equal on that, that is the background, the sky hasn’t changed, the stars are the same If one is varying — what happens if it’s a digital image? You should see a negative or positive instead of a zero in that subtraction, because you’re subtracting all the pixels across the image This assumes they’re aligned, by the way, which they typically aren’t, because the telescope can shift a bit in between images Also you can have other effects Well, if you do this naively, subtracting two images, what you find is you get — yeah, you have stars being subtracted You can’t see all of them there — it’s the light shining on, but you can see zooming into one, you get these donuts The reason you get donuts is — the fact is — in between just that two minutes between those two images, the atmosphere changes I’m sure we all know Twinkle Twinkle Little Star? Well, astronomers hate that, not the song, but they hate the twinkling That is the biggest problem with astronomy The atmosphere screws things up for us We have hot pockets of air rising, cool pockets of air sinking, and so what the issue is is trying to correct for that And so, what we do is we create a — well this shows some of the shimmering These might be one second apart — images of the moon And what happens is that causes starts to blow up, to shrink down to what we call really good seeing and poor seeing So this is what we — if you look actually on the digital image, per pixel you’ll see a star rise and have what we call a point spread function across the image And good seeing — this is a lot sharper and goes a lot higher Most astronomers want the best scene they could probably possibly find This is one reason the Hubble Space Telescope went up into space — to avoid this seeing issue So they get pinpoint images Well, we can’t put these telescopes up into space because it increases the cost of a telescope by $1,000 — or not $1,000, 1,000 times OK? Yeah, $1,000 Yeah, I’ll go for that OK. But anyway, so what we do is we take our best say 15, 20 images out of our pool of 2,000 images and create a nice reference image So this is what one of our — a combination of our best images And this has been corrected for all the defects in the image And by the way, if anyone knows anything about digital imagining here, this is just an overexposed star, which in your standard iPhone camera or android, they have software to correct for things by putting — I forget They’re almost like canals between the pixels And this here, obviously we have bad columns Some of your phones have those bad column, too, the software just interpolates averages and doesn’t median across them, so you don’t see them We want to see those where they’re at so we know — we don’t want the software doing things that we aren’t sure what exactly its doing So image here, we’re trying to find out which stars are variable So, if we use this reference image now, which is our best images we have, and then say here’s an image taken at a certain time, let’s see if any of the stars have changed in that time What we do then is take this reference image and convolve it to that image That is it looks at the point spread function in the image that we’re going to look at, changes the stars so they’re the same shape, same smeared out by the atmosphere so to speak, and then do the subtraction And then it can look at all 2,000 images and go through this So you start to see it’s also a computational problem, too Also it has to align all the images, so there’s pattern recognition Luckily, stars are easier to recognize than human faces, and the reason is that stars have patterns, so it looks for a series of triangles and combinations and can easily align these images By easily I mean in about 30 minutes it can do a 1,000 images are so But then it goes through and subtracts the images after it does the deconvolution, which is still time consuming And then it sends you what’s called a residual image Where does it find all the variants in the image occurring? And this is the variance in that image Now there’s those overexposed stars right there We ignore those because they have the weird shape to them But notice all these dots here are stars that have high variants That means they are variant Everything else is pretty much black in the background here Now, I’ve circled one of the stars here to show you this, and if you want to change the reference, here’s the reference image, but I scaled the black and white for you Notice that star right there So let’s go back-and-forth And you can see the stars — wrong way — that are variant here

And those stars are very important because they are somewhat in the death throes of their life They’re almost like a beating drum They’re in this unstable phase, but the fact that they’re beating like a drum, just like when you listen to drums, the sound you hear, the frequencies it’s vibrating at, tells you a heck of a lot about that star, not just the surface, but it’s almost like stellar size myology, where you’re inspecting the interior of the star itself So first off, you can see some of the problems we deal with One, it’s getting the data downloaded to us OK? That’s not necessarily on our end, it’s typically the observatories which are fighting all this problem because astronomy is very, very data heavy In fact, one of the radio telescopes they’re building is going to be downloading XO bytes per hour, 10 to the 18 bytes per hour, and they don’t know how they’re going to take care of that They’re going to have to archive, compress, who knows what I’m sure a lot of the people who are in this room and your colleagues are people who are working on that Thank God for you people because I can survive without the people in this room But anyway, imagine what we go through here if we take a whole series of images, OK, and we follow let’s say a single star We’ll watch it vary and stuff, and let’s say this is one of the stars that we’re varying Now what if I draw a line here Imagine I have 1,500 of these images now, just for a single cluster for a single season, and I follow a line for every single one of these stars, 20,000 stars That’s a heck of a lot of data right there in itself Now, each image — these are relatively small images The raw images are eight megabytes But when we process them and do a lot of division, there’s 16-bit images — or 16-bit camera to begin with, so that means there’s 16-bits per pixel, roughly 2 million pixels, 2048 by 2048 — 4 million pixels, excuse me I had to do my math there So, but the thing is when we process it goes from 16-bit We add another factor of two It goes — it’s the 8-bit, 16-bit, and then when we analyze it it pushes it up to 32-bits Now, remember I told you you also have to align the images? Well that adds another 32-bit image that you have to take care of Then when you convolve the image to match the reference, that prototype gives you another 32 megabytes for that single image You start multiplying all that times 1,500 for a single cluster over a single season, and you see we have a big data problem and we start piling up data Now, the thing is 200 gigabytes doesn’t sound that much, but the problem is you have to multiply this by five to 10 just for a single season if you use my colleagues And that’s just for the telescopes I use, which is one of the telescopes in the consortium So it is an issue Now, the thing is — it’s OK, you’re done with the data you can get rid of it No. The reason is when we look at stars — notice this is a 14 month timespan here, and this shows what one of those stars is doing each night we observe it You can see it rising and falling Then you might say why didn’t you observe it here? Well the sun was in the way OK? And that is an issue It’s the biggest source of light pollution in the sky And so, I came back into view here and we observed it for another few months why it was easily visible, high in the sky from the Southern Hemisphere And from this you might say well, what does this tell you? Well, if you look, let’s for instance add a single night here And so, this is now — we’ve already gone through this ISIS or difference image analysis Say here’s all the stars we can see right now that are varying over this 14 month time period If we look at this single night you can see here’s a star on this very nice up-and-down curve here We keep going with that If we combine all the nights together, what we call phase the data, have a computer, do an analysis of this to see what’s the period of this variable? How’s it varying? Now you can see all the different nights here and this varying quite nicely By the way, this star here is 10 million times fainter than what your eye can see when you look in the night sky And the data is accurate to about one percent That is variance in the air So it’s amazing what you can do, and this telescope that we’re using down there has only one-half the area of our Butler telescope, if that, all right? So it’s really impressive that we can get this with different image analysis is amazing, all right? And the interesting thing is the person who wrote the code for it — and this is where I tell my students you better be a good computer scientist, whether you’re a theorist or an observational astronomer, because if you are not, you’re not going to do well in astronomy, theory or observation The person who wrote the code for this — he’s manual — he’s French And the manual he has is just like two webpages By two webpages I mean literally two pages is all It’s all these scripts and stuff you’re running and all these programs And the thing is, we basically by trial-and-error went through and just figured it out We still haven’t figured it out completely, all the bells and whistles on it, but our colleagues at Valpo and Ball State are using it now also, and have used it for their research And then I have other people calling us now because my students have helped me figure out how to use this

I mean I said here — they said “Well, how do I know what to do if there’s a manual?” They said this is the way all astrophysicists work Usually we don’t read the manual, either You know, [inaudible] for you to try it out first OK? I said actually the best way to figure things out is trial and error because you learn your mistakes that way, too But anyway, this shows one of the data files here and this shows some — from one cluster some of the things we have And notice — just to give you a little science here, this one’s beating almost like a drum If you hit a drum in the middle it goes down and up, down and up, down and up Well, if you hit it on the side you can have these points where it has these ripples now in it, instead of one big ripple it can have two ripples, and these are the ones that are sort of oscillating in what we call the second, first and second overtone And so they literally beat like drums That tells us a lot about their densities, their temperatures, their mass, their ages, and their compositions That’s why we’re interested in this But these are what we call short period variables, roughly periods from maybe a few hours to a day To investigate long-term aspects of this, and to show you some of the long-term aspects, notice this variable right here One time it’s up here, then the next cycle it drops, and that cycle can be anywhere from 30 days to two years To study the second theory, what we call amplitude modulation that means we have to study it for several years, close to five years or more, which means we have to hold onto all this data, all the analyzed data, for five years minimum to study these things And then, we got to continually — if we want to look at it, every season pull it back up again and start working with it So that’s one of the issues we’re dealing with By the way, the — when we remote observe, we typically observe with the MAC or a PC, can do either one, and then we download it to that specific laptop, desktop, whatever Then we have to bring it down to our central server where we do the difference image analysis, because it has a lot more ram for us to use and stuff Then we have to bring it back to whatever iMac our students are using to process So it’s an issue of going back-and-forth, back-and-forth, back-and-forth, back-and-forth And then when we’re in this phase and the students are analyzing these light curves that I had shown you right here and the data, when they’re doing that — and this is what I kind of know — just in the last three weeks is my student, whose — he’s constantly bringing his flash drive back-and-forth from our observational lab to my office And I said well, can you show it to me? Well, it’s on the iMac Well, bring me all 500 files over here so I can look at it So, we get the flash drive out and walk it over to my office because it’s hard to, you know, merely impossible for iMac to iMac — OK — right now That’s one thing we need a central server, basically where we can actually all access it and have high speed access to that, and access to our data So these are some of the points — what I call — John will call them the pain points that we’re dealing with OK? So, let me just move forward a bit here One other curious thing about these clusters — there’s a tiny little globular cluster, actually one we observe Notice this dark streak here That’s a cloud of gas and dust One nice thing about this is we can determine how much dust is between us and the cluster In our work, we’ve suddenly gone from being theorists to being the premier group in discovering and finding these variable stars and globular clusters, which I never thought I’d be that kind of observer Doesn’t mean I’m doing it well, but I’m doing it, and the fact is we have 80 nights a year That means we can do this continuously, basically throughout the year, and study things And, as I said, remember good telescopes like the Keck telescopes, things like that; you get three nights a year on them if you’re lucky That’s not enough time to do all the work we do And also, that work would not — the Keck’s would not work on projects like this because the stars would be too bright You would have to take too short of images and it would just be a waste of telescope time But here the fact that we have dust and it can vary; we can use our variable stars to determine the age of this cluster, the distance to it, things like that And this tells us a lot about the history of the early universe How did stars form? What were their compositions? How did our disk of the Milky Way form where we live today? So really, it’s going back in time and telling us the early star formation in our universe OK. And this just shows — I didn’t mention this, but each point here is a single image equivalent, the star on that single image observed And this is infrared, visual went to the green, and blue filter So we produce a heck of a lot of data for each star When I plot out these plots that look something like this, there can be — well, if you imagine 1,500 points on each plot, and then I might have — sometimes I put 30 variables on one chart here, well that’s 30 times 1,500 The printer takes a while to get through all that, so I’ve noticed that as an issue, too, just printing things out So, one of the other things that we have done — our donor — and we did this —

he has a place up in Lake [inaudible], Indiana, and he invited me up there back in 2009 or 2010, and he said well, come on up, Brian I want you to see my house He has this extremely green house It has solar heating It has everything on it OK? And he is one of these — he might — this guy has plenty of money, but what he does at night if it’s cool — he opens up all his windows has the air-conditioning off, then the minute the sun comes up he closes all the windows OK? All of a sudden he starts running around the house, this two story house I was like wow OK. You are serious about this But you really want to see it But, by way, he said lets go mud walking in the wake Mud walking involved, basically, getting your swim trunks on and walking like this in the water and just mud in the offshores — about like this, and just walking along these He said well, so-and-so lives there, so-and-so lives there, OK, there’s so-and-so the disgrace of so-and-so or whatever, and then he’s going on and then he said — the provost had come to him So we were talking and he goes “Well, what should I say?” Well, here’s what I want to sort of push him toward I want him to say, you know, and by the way he used to be a tour guide at our observatory We get about 10,000 people a year to our observatory to look through our scope, so he knew what I was dealing with And I said well, we really should take our 1954 telescope, upgrade it, and by that I mean change the optics of it, make it remote observing, just like the Sera telescopes And so we are almost done with that now, and because it’s our observatory that means we have a minimum, taking into account cloudy days and things like that, probably about 180 nights a year There’s another 180 times 2-gig, times all the other multiplication of data analysis that we’re going to have here on our university, creating data So — and some of the things we’ve done now — we’re going to large formats, CCD’s, and the typical — the biggest thing with this telescope, we change the optics So it’s gone from a very long focal length telescope to a short focal length telescope, which means our field of view has gone from this to 10 times larger now in our camera, which means we can do a heck of a lot more science with it, even here in the City of Indianapolis Luckily we are not in downtown Indianapolis We’re in a residential area We have the White River and a lot of trees on one side of us, so we do have half our sky is relatively dark And we have tested the system already And, by the way, that person there, Peter Mack, who Andrew here has met many times, a jack of all trades, another PhD astrophysicists, but I’d say he’s an electrical engineer, a mechanical engineer, a computer scientist, network — well, you’re better now working person than he is, but anyway, it’s amazing the stuff I’ve learned just from working with this guy And he’s the only one who refurbishes old telescopes in the world And he is totally overbooked, and so when you said I can get it done in two years, you at least multiplied by two with him But anyways, this telescope is — our telescopes going to be starting probably coming into research use here in the next few weeks And that means any clear night we’re going to be using it And my colleague [inaudible] Hahn, he goes to China every year and the interesting thing about that — its going to be daytime over there for him, so he doesn’t have to stay up at night He’s going to be using our telescope from China and doing all his observations, saving the data here at Butler why he’s doing that And that is one of the beautiful things about remote observing That’s why I like the Canary Island telescope I can start observing at three in the afternoon and be in bed by midnight because it’s to the east of us quite a ways So just to go review some of our issues that we have with this research, and we do find inventive ways around it Flash drives are a big problem, you know, we have to use, but downloading data from remote observatories, transferring data from I-Macs to process, moving [inaudible] images to the campus server for difference [inaudible] analysis, moving [inaudible] files on each star back-and-forth from iMacs OK? And one of the things is sometimes when we download the data, we carry it just across the room, almost like 20 feet across the room, with our flash drives to put in the iMacs because it’s just easier to do that half the time than it is to — frankly, try to push it through the network, get on a server and do this-and-that So it is one of the issues where having a central server where all this stuff can lie and having fast access to it, to move the stuff back-and-forth, because we do have to move it back-and-forth, and that’s probably our biggest issue And then, transferring — like I said, between faculty and students And when students come to my office — I’ll say well go get your flash drive It’s just easier right now to do that So you see some of the issues we deal with Astronomy is one of the most data driven sciences there is I was talking to — who was it? Shaun [assumed spelling], whose in our biology department and, you know, I was reading this article, I said well, Shaun — he said here, this telescope over here they’re creating is going to have XO bytes of data an hour And the human genome is only 1 gigabyte OK? But essentially it’s just there’s so much things Because remember, in astronomy we just don’t have X, Y, Z, we also have time

We also have luminosity So we have multiple dimensions we’re working at We also have colors we’re working with So we’re going in all these different parameter spaces and dimensions that we have to follow, which creates a lot of data for us Yes, the initial imaging it’s not all that large, but what happens is once we get to the final product, we have produced a lot of stuff that we need to hold onto So anyway, I’m going to hand it back to John and he can comment on anything here, if you would like Do you want to go any further with it? >> I guess I do have a question >> Sure >> When you talk about variance with the stars, does that include — this is going back to how the universe — you know, we have a sun and planets revolving around the sun, dimming the light, does that variance also include those types of things? >> You’re thinking of planetary transits they’re called OK? Typically the chance of having a planet in the plane [phonetic] — absorbed is one in 10,000 of having it at any time pass through OK? Most of the time the planes of the orbits are tilted so they never pass in front of the stars Now, if anyone’s ever heard of the Kepler space telescope, which was launched specifically to look for earthlike planets, it was looking for dimming’s as earthlike planets would pass in front of the star there would be 100 of one percent dimming of that star We’re looking more in the 300% value down to maybe the five percent value of dimming So not down to the real small dimming OK? >> OK >> Yeah. If you don’t mind I’d like to talk a little bit about the infrastructure side of this — wrap it up >> By the way, if anyone has questions — >> After he’s done I can answer any question you have, too >> OK. So I’m just going to skim through a lot of this as quickly as we can We have 15 minutes I’m going to try to add what’s most relevant I added Brian’s process here and we’re going to flip through these pretty quick Step one in Brian’s process observes something Step two — so step one is to identify something to observe Step two is to observe And, of course, my telescopes aren’t as good as Brian’s — >> That was like my first telescope >> Oh. So what happens is after they observe, they move the data from the telescope to a PC in the observatory From there, they move it from the PC to iMacs with students using thumb drives Now, look at those thumb drives That’s a lot of work But you see why that’s kind of a problem, right? Imagine the time that goes into making sure the right student has the right thumb drive I mean, I just picture this stuff getting confusing and, you know, wasting a lot of time And in any given semester when you’re trying to teach students, you really have to be on point, and you don’t have time to be confused in any way And this is just one piece that I think is agitating to Brian So then they process the data and images on the iMacs They transfer the data from the iMacs over the network to a server we have, and I believe that’s where your ISIS application is They do some preliminary analysis on the server They compare the differences, which Brian showed us Then they transfer the data back to the iMacs over the network for final analysis And from there, they move it to Google Docs, which is where they do a lot of their collaboration So there’s a lot of emphasis on moving data around in the network here After that the papers are written and submitted to a research group So, as Brian mentioned, part of the problems we have are lack of essential storage, which is probably the most important thing, so we’re sort of grateful that one of the grant opportunities emphasize storage for research And since that just came out a couple weeks ago, we’re kind of re-pivoting and rethinking of how to do this, even though we’ve been working on this for a year Now some of the fiber infrastructure limitations can be pretty significant You know, we have Brian — Brian’s research, but we also have Shaun [assumed spelling] who works in genomics and biology We call him dirt guy, you know, along with other researchers And Butler, a little while back, they got approval to build a new science building

With that, we’re hoping to sort of consolidate a lot of our research people in there and our systems, and hopefully increase our research footprint as a whole at Butler But to do that, our fiber infrastructure needs to be upgraded So surely you guys are already noticing multiple things we kind of have to work on, we just have to pick the best one that satisfies the requirements of this grant Some of our buildings are only connected through one gig links and we have to ether channel those links Our secondary Internet is only 100 MB to 500, which is a serious problem Just imagine a situation where — now it’s all ready These guys already have to worry about cloudy days and other things that could prevent them from getting a good observance So imagine some of their collaborators in China or what have you are connecting to our telescope and, you know, that same day maybe we didn’t have time to coordinate with Brian or all the other people You know, maybe we have a vulnerability or a very bad Dee-Dos attack happening on our primary Internet circuit Maybe we need to upgrade the adjacent directly connected switch for whatever reason There are many reasons that our primary Internet circuit could go down and when all that traffic transfers over to our secondary circuit, that’s going to be a problem You know, people, you know, imagine if you miss the observance of Halley’s Comet or something like that I mean that doesn’t come every day So we need to emphasize and look forward to strengthening our infrastructure as a whole, not only for Brian, but to lay the foundation to increase our research footprint as a whole So some of the process that we did when we heard about this, you know, we really didn’t know where to start And thankfully we had organizational restructuring before I started at Butler, about a year, two years before I started They developed a partnership team, which bridges between IT and all the other departments Thankfully we were able to engage some of our account managers to go out and talk to the faculty and kind of get a broad spectrum of who we think could potentially be doing significant research So that resulted in having Brian and several others, you know, there are a couple other people We all got into a room We invited I-Light Internet2, which by the way; if you guys go through this process, definitely involve them They’re so unbelievably helpful with this We all sat in a room that started to get way too hot We had each person discuss their pain points and we took notes And we were able to shave off 80% of them right off the bat You know, they just — they weren’t science and they weren’t certain things that we needed And at het same time we got value because we wrote down their needs, and we now know that we can come back to those later So in that process we — once we knew what we wanted, we were able to develop a grant team if you will It included IT people It included our researchers It included the grant office We got a new director of grant something or other We knew we wanted to capitalize on her new motivation, you know, being a new employee Got to get her now before that dies down So we got I-Light and Internet2 and its been a great team So in summary, you know, we talked about Brian and his research You know, along with other researchers doing stuff on campus and things that they want to do, but they just can’t do You know, Brian, I think some of the things we didn’t talk about because we wanted to save time is, you know, Brian and I have talked a little bit about potential e-learning with astronomy, and kind of engaging community for these things And, you know, if we can’t get enough collaboration and people able to login to these servers or, you know, maybe even just streaming astronomical events, how do we really reach out to the community and get people interested in science, especially astronomy? So we talked about some of the pain points, submission process

That’s it Does anybody have any questions for Brian or myself? >> I got one Do you guys actually [inaudible] Yeah, this is Mike I’m with IDSolutions Did you guys actually get funded then with the NSF or is that still in process ongoing to get the final version kind of polished or are you still trying to add things and whatever? >> We haven’t been funded yet We — in anticipation for this, we kind of thought that the 2016 updates to the grant would come around January, I believe So we have prepped all the way up to that point so that once it was released we could hit the ground running You know, at that point our idea was that we would already have everybody engaged, everybody’s aware of who needs to do what when the time comes Because we have other jobs, right? You know, we don’t want to stay up until three o’clock in the morning make sure we’re getting a letter of recommendation from a campus leader, what have you And we just needed to make sure that we had something solid to work with So now that we have all that, you know, of course, the new information came out a couple weeks ago, and, you know, some things have changed, but most of them are the same So 80% of the work we’ve done is still very relevant You know, now we just need to reassess, tie things together, and really think about how we’re going to submit this We were thinking about doing a couple Brian has talked about, you know, maybe engaging other universities involved in this research to setup sort of a shared collaboration and storage, maybe through some type of MPLS circuitry or something like that And those are the bigger grants, but, you know, we still need to reach out to those people to see if they’re interested, you know, not only the faculty, but the IT staff of those other universities to see if they could support it So, at this point, we’re kind of taking a multi-pronged approach then we’re going to eliminate, which of those doesn’t make sense and move from there Anyone else? Come on guys, shoot for the stars Don’t shoot me I had to get one pun in there All right There’s one [ Inaudible Audience Question ] >> Actually what I was thinking about was a question got asked about transit >> Yeah? >> So when you were looking at the globular cluster I thought well how many times or how likely is it that there’s a star behind the star that’s contributing to the light that you see, and if that is the case, how do you tell the difference? >> Well, what we do — we can actually — we call them planetary You know, since the atmosphere is blurring the stars, if we get two stars close together, what we can do is we see its amplitude and variation isn’t where we think it is And also the color appears different than what that star should be for its amplitude and period And so we just go to the Hubble space telescope image and look at that You might say well why don’t we just the Hubble space telescope to do this research Cause its booked 24/7, 365, and they would not give us 80 nights a year for the project OK? So we just use a single snapshot that they took to look at high resolution of what’s sitting there where the star looks to be suspect And referring to planetary transits, people have looked in the last 10 years, just stared at globular clusters to see if they can pick out these minute planetary transits And they found a few Jupiter’s but that’s about it OK? So part of the problem is there’s just so many stars and you have to look for really, really tiny variations in the star to find those We’re looking for somewhat larger of that >> So where do you fall on the Pluto issue? >> Well, actually after I teach my class and my students are sent, I have one person in a summer class I’m teaching says “Pluto’s a planet.” And then she goes “Yeah, it belongs with its own kind so to speak.” OK? It’s brother in the Kuiper belt Pluto has the advantage — and just to let you guys know, it’s sort of an interesting historical fact, it was the first object in what’s known as the Kuiper belt discovered When it was discovered it was thought it was 20 times the mass of the Earth It turns out — – now we know it’s only 2,000ths the max of the Earth So there’s been a drop of 10,000 in what we thought the mass of Pluto was from its discovery to today

And we now know of 10,000 or more other objects that sit out there where Pluto is, so it’s in the belt of objects It’s the brightest of them all, only because it’s one of the closest There’s actually Eris, and that’s the one that sort of knocked Pluto out of the planet club, because it’s actually larger So, most people don’t know in 1849 we had 22 planets That because as they’re counting asteroids they’re including those in the planet list And they realized — let’s call them minor planets OK? And so, we went through the same phase in 2006 >> Is there a common standard for the definition? >> The definition is it has to be round, which means it has enough gravity to pull it into a sphere Number two, it has to orbit the sun And it has to have enough gravity to pull it clear out its neighborhood around it, which all the major planets have done, including the terrestrial planets like Mercury, Earth, Venus and Mars >> In regards to the observatory in Chile, are you working with the government there to get — >> Yeah, well we — actually — Cerro Tololo is this joint operation between the U.S. Government and Chile And there’s numerous observatories all up-and-down the Chilean Andes There are observatories all across that from Europe, from China, from everywhere, going up and across there And their main site, Cerro Tololo, is one of them East [inaudible] European Southern Observatory, another spectacular site And they have the best food of any observatory I’ve been to, because it’s a 12 nation Europe, and so they have to have all the cuisines there The American is typical American fare But anyway, we are funded — I didn’t mention this, but the observatories themselves, when you saw that Cerro Tololo Observatory, that was partially funded by our dues, our buy-in I think part of Valpo’s dues and Ball State they came in shortly before us That dues help do it But also the National Science Foundation gave us substantial money Any one university, small university like us, wouldn’t have got the money to refurbish one of those But as a group, we have all these students that it’s going to help, so it’s like two or three our one university put together, as far as astronomy goes And the same thing for the Canary Islands go And the NSF has gave us money for that to go through the refurbishing Also, they have given us high-end cameras worth about 100K each We had them built for us in a spectrogram [phonetic] So we have great equipment on them and we have totally refurbished optics on the telescopes And that’s why they are so good to use, and the fact that the software allows us to remote observe >> All right Well thanks guys [ Applause ]