 This is research in Manoa. I'm Jay Fiedel and that means it's one o'clock on Monday. And we'd love to find out what's going on at HIGP and SOES, the School of Ocean Earth Science and Technology at UH Manoa, because it's always a surprise. There's always things happening there that are world-class things. World-class researchers collaborating with other world-class researchers everywhere in the world right out of UH Manoa. And I think it behooves us all to understand that, to know that, to appreciate that and applaud them for their efforts. One of those researchers is a post-doc fellow at the HIGP, and he's working on satellite technologies. And we're going to find out a lot about that today. He's Dr. Sam Murphy, and he's here in our show. Welcome to the show, Sam. Thank you very much for having me on. Great to have you. So satellite technologies, that kind of implies that you're building satellites here in Hawaii. Tell us about it. Yeah, that's true. So some of my colleagues are building satellites. There's a lot of work, actually, that goes into almost like the full stack, you could say, from designing the sensors, building the satellites, putting them together, trying to launch them into space. This is all sort of part of a collaboration between SOES and the School of Engineering in what's called the Hawaii Space Flight Laboratory. What's pretty neat is I think next year they're going to try an orbital launch, hopefully that will make it into orbit. And so that would be really cool. So that's Hawaiian technology that's been designed here, built here, and then put into space. Wow, this is fabulous. So is it the first time, or is this something you do regularly? Well, so my part to play in this would be I'm more of a software guy. I'm not the hardware guy. Those are the guys I like best. Really? Well, that's good. So I deal more with once we get data from space, how do we use that to extract new information? But we sort of as part of UH, or larger, as the state of Hawaii, did try a launch around a year ago. You say try? Yeah. Yes, you could say try. Well, there was a lot of successes in that. We'd build a satellite and put it installed onto a rocket and it passed the test from NASA and so on for being space ready. And it actually did get into space, but it didn't get into orbit. So at the moment it's at the bottom of the ocean somewhere. But we tried basically, and we're going to try again next year. Let's have a case study then today, Sam. You and me today in this show, we have 20 minutes or so. We're going to build a satellite. We're going to get it into orbit and we're going to get some feedback from it. What do we have to do first? A bit of rocket science. Okay, all right. We need a rocket for it. Where do you get a rocket? Do you look up rockets on Google? Well, one of the really cool things is there's a company that's just coming out of New Zealand where you can actually pay for fried shares. They're called Rocket Labs. And that's one of the partners I think that UH is thinking of partnering up with quite a bit. So that almost makes it more simple. You just need to build a satellite and you can put it into the rocket. Further down the line though, UH, well, sorry, the HSFL, the Hawaii Space Flight Laboratory wants to do basically everything. Does that mean a machine shop? Yeah, that's a good question. They have the material to build the satellites. Again, this is slightly outside of my realm of expertise. You're a software guy. And also what's really cool is they test the satellites too. So here on Earth, the conditions are sort of things we're used to. But in space, it's quite different. So often things that you can buy in a shop or something, they won't work in space. So you need to test them to make sure they will. So for example, you might put them in a thermal vacuum chamber or on a vibration test bed to see if they can survive the launch or survive the vacuums of space, that kind of thing. So you need to do all that testing when you put everything together. And also typically you need to design it to be as compact as possible because every gram or pound counts. Then a formula for that. How much fuel you need and so forth. Yeah, absolutely. So you have to know exactly what your payload is. And your payload is the satellite itself, right? The rocket falls off. Maybe more than one stage falls off. And now you have a satellite up there. And you have to do a lot of calculation to make sure it gets into the right place. Yeah. That's another one of the really neat things actually coming out, again, of HSFL and what some of my colleagues are working on. They're working on this open source software called Cosmos. And with this software you can control different satellites. So they can communicate with each other and get into the right sort of position and stay in orbit properly and point in the right direction. So that's neat too. So they're really thinking of basically everything. Open source? Yeah. You mean is it secure? Well, also this is the perfect example. It's image right here. What is that now? So I borrowed this from a colleague and it's an example of how you might assemble some larger structure in space from smaller structures. Yeah, I say it looks like either Oreo cookies or maybe sushi. It looks like sushi, California rolls. Yeah, exactly. And it's much easier or cheaper to launch a whole bunch of small California rolls. It might not even be possible to launch an enormous sort of structure into space. So what is happening in that picture? So that's supposed to sort of represent this idea that you have these different, you could call them agents, I guess, communicating with one another so they know where each other are and they know the position they're supposed to be in and they can do it automatically. So is that like mesh technology where they're all talking to each other and they're all doing this kind of wireless group communication? I guess you could call it that. And why do you want them to talk to each other? Because if they can do it by themselves it's going to be much easier than having to communicate with them from Earth because if you lose communication or anything like that or if you want to just go further into space or something then they really need to be able to sort of have more autonomy. Does each one of those devices have its own propulsion? They could do. So that's just a concept but I think, yeah, they could have their own ways of moving about that might be, for example, using little jets. So it doesn't take a big jet to move a small thing around? Exactly, yeah. There are other technologies but I'm not sure I'm even allowed to share the right idea. Okay, you mean classified? If you tell me you have to kill me? All right. Let's go back to the slideshow. Okay, we're going forward from that. Okay, so this is more the sort of work that I do. I can speak more too. So when you look from space down to Earth the sky is kind of in the way. So when we look up at the sky it's blue and that's a problem when we're looking down on the Earth because it's not completely transparent. So on the left-hand side there's sort of the image beforehand if we look just straight from the satellite and on the right-hand side is sort of the same area at the same time but corrected for the effect of the atmosphere. Traditionally you had to pay quite a lot of money to get that done so the software was very expensive. That's the same place. You're enhancing the photograph. Exactly, same place at the same time. So on the left it looks kind of hazy and blue and on the right it's more green in color so you're really seeing the vegetation there. There's actually an example of the work that I'm doing. So this is a volcano and it's got some crater lakes inside of it. A volcano where? In Indonesia, on the island of Flores. So that is now an open source. So the great thing about open source is anyone can contribute obviously you don't have to pay for it. And yeah it sort of takes away that inability for many people to contribute. In many countries there are people who are capable of contributing but intellectually and with time and probably a computer or whatever. But if it's too expensive they have to pay a lot of money for the software. Open source is cheap. Exactly, open source is absolutely free. But it's also hackable isn't it? Well, so I suppose so but you can still have things that are open source and secure or you just have to design them. Design the security. So what about the software that you've written in order to enhance the photograph and all that? So what that was building on is actually the work of Robin Wilson and he wrote sort of a wrapper to another project that had this, what they call it's a slightly technical term but radiative transfer code which basically enables you to model the atmosphere and remove its effect from what you're seeing from space. What do you need to do? What data do you need to do that? Depends what wavelengths you're looking at but in the visible wavelengths you sort of need some idea of the water vapor column, the ozone and the aerosol optical thickness. Where do you get that information? You can get them from a few different sources. Traditionally from radiosomes you had to launch a balloon and that kind of thing. But that's not really, that's the best quality information but it's not really feasible. But it's not practical in other situations. So typically people use a mixture of radiosomes with satellite data and models. So basically I can go to different institutions who run these quite complex models and get their best estimate. And they were given to you because they're nice. It's free. It's free. It's scientific sharing. We love to have that. Yeah, it's great. So does this mean that your primary software effort these days is into enhancing photographs like that? Well to be honest that was just something that had to get done so that I could look at the volcano underneath the atmosphere. Which I love, don't want to have to do that. You did it because you had to. Yeah. And it was not on the shelf. You couldn't go on Google and find the same software. Not on the scale that we wanted to do it. We could have paid for the software too but even that we struggled to because we wanted to look at all the crater lakes on the Earth. So you went further. But how did you learn how to do that? We're talking about scientists doing what has to be done. Right. How do you crack a book? How did you learn that? It's not easy. Let's see. Google's pretty good. The documentation from the source code that I was working with is really excellent. So it's a way to transfer code in case anyone wants to use it called 6S. Oh, 6S. Yeah. All right. And it's very well documented. And so that was helpful too. And then just sort of, you know, cracking, tripping away at it and trying to get it done. That's fabulous that you could do that. I mean, it's what impresses me is not so much the software in this one case, this one instant, but that you could decide you needed it, wanted it and you were going to do it, period. And then you can apply that same kind of thinking and effort to lots of other kinds of software development. Yeah, I hope so. Are you trained in software development? Do you have a computer science degree? No. No. So hopefully I'm an inspiration to other people. I think so. I think you do. Because I trained as a geologist. And so we're not typically associated with having many skills in mathematics or for coding. Yeah. But I found that. I found that the people at the HIGP and SOEST, when they need to do something, they do it. Right. And it teaches them and it teaches them to open their minds and solve problems no matter what. And I admire that. We'll take a short break now. We'll come back and we'll admire more of your slideshow. All right. Aloha. This is Kaili Akina with the Weekly Ehana Kako. Let's work together program on the ThinkTek Hawaii broadcast network Mondays at 2 o'clock PM. Movers and shakers and great ideas. Join us. We'll see you then. Aloha. I'm Carol Mon Lee. And I want to welcome you to our newest series called Education Matters. We will explore education-related topics that touch everyone, not just formal programs in K-12 and higher education, but also broader issues and information that affect our community. Are you looking to get shrunk? Join us on Shrink Rap Hawaii. My name is Steven Phillip Katz. I'm a licensed marriage and family therapist. I see couples, individuals, families. You know why? Because we all have problems. And if you're curious about shrinks and what they talk about, come look at my show, Shrink Rap Hawaii. And maybe you'll find your shrink. That's a picture of a satellite. And here we are on research in Manoa, talking to Sam Brown, the postdoctoral fellow at HIGP. And we're talking about satellite technologies. And he's directly involved in that. And he's involved in conceiving of and executing really creative plans for the software that they use on these satellites. So before the show began, you told me there were three things to focus on. Opportunities, applications, and technologies. And I can't say that I understand exactly what the relationship's of it. Maybe you could describe it to us, and we could also see some of your slides. Yeah, I would be happy to. So I guess one of the main opportunities is here in Hawaii, we have a unique geographic location which gives us a sort of unfair advantage for launching satellites to space. Because on the entire planet, there are roughly 100 spaceports. So that's a location that's launched at least one rocket into space. Those rockets are most often satellite rockets. Yeah, almost always. There's only about four spaceports, well exactly four spaceports, that have launched humans into space. So it's in Pyongyang, for example, on this discussion. There should be clear about that. Yeah, so that's a couple in America, one in Russia and one in China. But at least in America, which has been the powerhouse of launching things into space, especially people, but just in general, they have two main spaceports. One on the east coast and one on the west coast. The thing is, you can't launch rockets about human populations. So the one on the west coast can only go west, and the one on the east coast can only go east. Exactly. But in Hawaii, we can go east, west, north, south. Sure, great location for that. Yeah, it's much more flexible. And so that's one of the sort of creators of opportunity. Another is there's a move towards launching lots of small satellites, moving from the era of unique, huge missions to many small missions. And that's something I think Hawaii can contribute to and has been working towards for about a decade now. What's a huge mission as opposed to a small mission? A huge mission would be something like Saturn V or the Apollo missions. Okay, okay, yeah. Man's spacecraft. Big, big rockets. But you know, aren't rockets and satellites going, you know, autonomous now like everything else? That's safer. You get less trauma, but you get more payload. Well, I heard that. So you know SpaceX? I heard that they're going to have a pretty much autonomous mission for those two guys that are going to the moon. You know, the two tourists? Yes, right, right, right. So apparently they'll be the first people to go back to the moon for about 40 years. That's going to be autonomous. Yeah. They make some money in the process though because they're going to charge them a lot of money. I'm not doing that. I've got to save up a lot. But science, like the Mars trip, the Mars trip is completely autonomous, right? There's no people involved in that. Yeah, and I think that's definitely a trend. It makes a lot of sense for space to be autonomous because it's very expensive hardware. It's really far away and it's difficult to communicate with. So you want to make sure it can just look after itself at home safe. It takes so long to actually get there that I'm only going about what I know from the Martian movie. Yeah, it was true though. It takes a long time. So the software then becomes more important when you're making a trip like that because you want to scour every piece of data you can get. So what kind of data do you want to get as the application and the technology part of your trilogy of events here? Okay. Looking at the earth, the kind of data that's really useful to basically anyone is data that's in the visible wavelength because it's intuitive. And we can just look at it and say, ah, yeah, I get it. It's a river and it's some buildings or whatever. So that's always useful and it's the most popular. And the whole thing about photography becomes much more important. So you're looking at sensors, you're looking at cameras which are just another kind of sensor. Exactly. So you bring all that, what do you call it, graphic imagery back and data imagery back. And that can tell you so much. So it's in getting it, seizing it so to speak, capturing it and then analyzing it somewhere along the way. Are you involved in all of those things? So when it comes to that, when you're analyzing it, basically my job, I guess, starts once the engineers have basically got information at that point I can start and it's sort of an image, then I can start looking at it. But you're going to tell them what bytes you want. My job is to try and extract information from the bytes. So it might be like, what color is this volcano or what temperature is this lava flow or whatever, or when do you think it might erupt? Ah, okay, okay. But you tell them, look engineer, this is what I want. I want you to design the sensors in such a way so that this is the data I get and this is how, the format I want and so forth. Or do they just give it to you any way they want and then it's your job to figure it out from that. Yeah, it's more the second. It would be neat if it was the first. But to be honest, as someone that studies volcanoes or has been studying volcanoes and also wildfires, it's more like I have to be kind of resourceful with what already exists. So if it exists like, you know what, I could use that to do such and such a thing. Maybe in the future, NASA for example are designing a mission that is specifically designed to look at volcanoes. But it hasn't historically been sort of top priority, to be honest. So I give you the data. I give it to you on multiple sensors in multiple ways. I suppose it's a huge, big spreadsheet of data and it's from a volcano, maybe on another planet maybe. What can you do with that and how do you do it? So the first thing, once you've got it into something that makes a physical sensor as opposed to just bits, then often you will run some kind of analysis on it. So there might be some time series analysis, or these days more and more people are using pattern recognition and machine learning to try and bring these data together. Is that like AI? I guess a little, I said like a subdivision of AI, I guess you can say. It's really useful because when you have two variables or three variables maybe you can plot it and look at it on a screen. I mean, three is already stretching it a little bit. But when you have four or five, you can't really even visualize it properly. So it starts to become really useful to have machines to help you find patterns in data that has many dimensions. So it's very sophisticated software when it does that machine learning and artificial intelligence. And the great thing is it's open source and free. Yeah. There you go. Okay, so now you get it, you use the machine learning and the artificial intelligence and what have you. What are you going to learn? I'm your client, okay? I'm NASA or whoever may be funding it or asking for the information. What are you going to be able to tell me about this volcano on this distant planet which has a funny atmosphere and so forth? Right. Well, volcanoes on distance planets, that's tricky. Almost anything is kind of new, I suppose. But if I could speak more to volcanoes on Earth. Okay, let's start there. Then what you would, you're always hoping to try and mitigate the hazard that they pose to people. So if in any way you can understand better when they might erupt then that's what people are just typically striving for. Yeah, and you wouldn't be the only one at HIGP who's looking into that. No, we have. A lot of people are trying to save lives over there. Great team working on that. So what kinds of things would you spot that would give you concern about the lives of people here on Earth? Well, the way I see it is it's a bit like like when a doctor goes to a patient and you have lots of different possible symptoms. They might be a bit, you know, they might have a shake, they might get hot or they might, you know, whatever. The color of the skin might look a bit different or something. If it was any one of those you might not be able to diagnose it. But if it was all of them you could hope to extract some kind of information that otherwise you wouldn't have been able to. You know, a number of interactive variables all acting differently. Yeah. And then you say there's something funny here. Yeah. But is that because you know that that combination is funny or because you're just concluding this is out of the norm? Yeah, often you're sort of trying to look for stuff that's not normal because each volcano is kind of a little bit idiosyncratic. It behaves in its own particular way. So you're often instead of saying you know what every time a volcano does shakes a little bit like this it's going to explode. It never does it just like that. You're trying to look for lots of odd things happening at the same time. And then, yeah. Is this intrinsic or do you say let me think now we have some odd things going through the photography through these graphic analysis and all that clarified atmosphere here on this particular what in Indonesia for example. But we also have information coming from other sensors that are not satellite necessarily. Right. In the ocean, in the atmosphere who knows what. Yeah. Are you going to be in position to crank that data into your model also? Yeah, absolutely. So in the example of that volcano in Indonesia we also have the chemistry of the lakes and some weather data. So we're hoping to put that together and see what comes out of that. So have you published? Are you publishing? Will you publish on this? How important is it to publish on this? Oh it's extremely important. I probably shouldn't say much more about this volcano in Indonesia because we are going to submit the manuscript this week in the next few days. Okay, so what will you say? Can you reveal it in advance? Nobody will know. I don't know. Well yeah, it's hot off the press I guess. What can I say about that? The color and temperature of volcanic lakes can show something that we might have missed otherwise is what I can say. Okay, that's it. We'll have to read. We'll have to watch it wait. So where can I read your paper? Well we have to submit it to an article and if they accept then it will. But we don't know which article it's going to be in. Oh, so how long will it take? Typically it takes a couple of months for the reviews to look through it. So I'll give you a minute now to tell our audience why they should read it. Why it's important for us to know what you're doing. Why it's important for us to make these efforts to save lives and understand about how our Earth, our planet works. Sure, okay. That red light right there. There we go. It's there. There it is. That one. So I think it's important to study volcanoes because they and also wildfires because they are hazardous to people. They, people can lose property and their lives and or just have poor health afterwards from for example inhaling smoke or ashes. Oh my god. Exactly. So I'd say it's critical to look into those things and fortunately we have a great team at HIGP that work on looking at volcanoes using satellites and I think it's an excellent endeavor and we should continue on with it. Okay. So this is this is the way your career is shaping up. Can you give us a trajectory and where it's going to go from here, Sam Murphy? Oh, wow, that's a tough question. So there are a couple there's actually two potential directions here. One of them is to continue in sort of trying to extract information from satellites to look at volcanoes and or wildfires and another is to and sort of these two competing ideas have to see which way to go. Maybe I'm being too honest right now. I don't know. Never. The second would be to try and potentially once you know if I get all the work done that I need to get done now to try and help HSFL and Hawaii launch satellites into space by trying to bring in sort of a bridging the gap between the private sector and the public sector. I might be able to play a role there possibly. I think that that would be a great collaboration. So something a bit more like in the Silicon Valley you have Stanford and so on excellent collaboration between public and private sectors. I think that's something that we might be able to do here in Hawaii. Yeah. And in this administration you're going to need to monetize these efforts. You're going to need to find connections with business to fund your research. Well, thank you so much Sam. I really appreciate you coming down. Thanks.