 This talk is called Upset, the first open-source satellite, so I'll just fall back a little bit. Before I was born, there was a situation where a beep freaked out half of the world, and the other half of the world had a big party. Today, you could send that beep from outer space down to the world with Upset, an open-source satellite. Here I give you Piero Papadeas with this talk. Give me a warm welcome for him, please. Thank you. Thank you. Hello. My name is Piero Papadeas, and we're here to talk about Upset, or Upset, or Upset. It's going to be a brief history of the satellite itself, how we got started, what are the struggles to create an open-source satellite in the process of doing so, and why we did that and what do we want to do more around it than other projects. The project is a Libre Space Foundation project, and we are also members of Libre Space Foundation. LSF, Libre Space Foundation, is a nonprofit organization designed to create open-source technologies in space. That includes satellites, what we call upstream in the space industry, but also downstream and midstream applications. We're going to be having a talk right after ours here about Satnurks from Nikos, which is the ground station network that we built with Libre Space Foundation. How it all started. Basically it was 2012, and von Karman Institute of Fluid Dynamics in Belgium set out to create an audacious project, which was QB50. QB50, the concept is basically this, there is a central institute that is managing the science part of many satellites, and it's making a call out to all different academic institutions and research institutions, universities, to create a satellite around a sensor and help you along this process, mainly with securing you a launch, which is super critical on getting a satellite in space, but also a bit of guidance around the science itself. Because every satellite needs a reason to be up there, and science is a reason good enough actually. What we started with, and what every project of QB50 started with, was basically a sensor. This sensor, the specifics of it, is basically a MNLP, multi-needle Lagamo R probe, which practically is used to measure plasma concentrations on the thermosphere, that's a part of the atmosphere. Von Karman Institute would like to do some more science around modeling this part of the atmosphere, the thermosphere, and they built many designs and built many different sensors around it. So basically you got started with one of those. So early in 2012, they opened up applications for different universities to come in and build satellites around those sensors. And the concept was that in Greece, one of the universities, University of Patras, initially got accepted in 2013, that's four years ago, to create one of the satellites, one of the 50 satellites. And while most, if not all, QB50 satellites went in a way which was pretty traditional for CubeSats, and by the way CubeSats are small satellites, it's a form factor, and this is the form factor that you can see here. So most of the universities in the QB50 program went on a traditional way of just going out and buying off the self modules and assembling a satellite. There is a quite growing market of commercial and off the self components to buy and create your own satellite, but this is the costly road to go down, but it's also the non exploratory and non innovative road to go down. And to their credit, the University of Patras said, well, we're not going to be buying components from here and there, we're going to be designing a satellite from scratch and building from scratch. And that's actually quite a, quite an adventure to go down to, and due to various different reasons, financial, organizational reasons, and reasons within the university itself like managerial type of things that you always have, they were not really able to progress with the building of the satellite. So at some point, and that was late 2015, like December 2015, they picked up the phone and called Libre Space Foundation and they said, well, initially they wanted the ground station and then after the ground station, it came different components of the satellite itself and one thing led to another, and finally we were building the whole satellite from scratch because most of the research that was done around the satellite up until that point was not really usable. When we first got the call about building the satellite, we asked, well, how much time do we have? And they said, well, six months. And for those of you that know like the time in space projects and time frame that space for projects are involved, six months is pretty much nothing, like most of the Cube sets or even really small satellites ranging up to really large satellites take a couple of years at least in development and then verification and then actually getting a launch in space. So when they told us six months, we were like, yeah, that's like, yeah, let us get back to you on that. And the fact is that there is a hard deadline, which is the launch itself, like there's a rocket waiting for you, or almost like this. And you have to rush to meet those deadlines, right? And we sat down together with the rest of the Libre Space Foundation, like the whole team, and we said, well, we might not be getting a chance to build an open source satellite from scratch anytime soon, so yeah, what the hell, let's do it. So yeah, started with a really short time frame. What the responsible engineers do is check if there is anything out there that we can reuse or take knowledge from abstracts and best practices and how we can really kick start the effort of creating a satellite. So we went into our favorite search engine and typed in, well, open source satellite or documentation about Cube sets or how do I build a satellite? Or more complicated questions like this. And what we found out is that there were a couple of groups in the past couple of years that have tried to do things in the open, not necessarily open source in the licensing aspect of things, starting with ARDUSAT, which claimed to be open source, but unfortunately was not at all. The only thing that was open source was the part that was executing Arduino code while in orbit, like none of the hardware, none of the software underlying the mission itself. Then we had the open source satellite initiative satellite that was never really operational plus the documentation. It was not licensed as an open source in the typical sense that we've grown accustomed to, like ZPL license or whatever else it is. And there were some guidelines about what you should be careful and what you should do, but nothing more than that. And then ARMSAT groups, specifically ARMSAT US, North America. ARMSAT is the amateur satellite groups that are creating satellites for amateur radio usage, have done things in the past and published them in various different formats on papers and proceedings in conferences and congresses, but that's really far away from open source plus due to ITAR regulations, that's export regulations from US, it was really hard to actually reuse or even get a hold of those paper proceedings from conferences. And then a couple of other things like SwissCube, which is the example done right, that have published extremely well documentation about the project itself, but still the actual schematics for the hardware or the actual code of the software was completely missing. So we said, well, yeah, not really. It's nothing completely open source that we can reuse or build upon. So we started to design everything from scratch. And we had to take care of many mechanical considerations around the science payload itself. You can see the science payload on the left side, together with four probes folded on the sides. And then as you move your way up to two units cubesat like this, so 10 by 10 centimeters by 20 in height, then you can find the electrical and power system, that's the EPS, then the attitude, termination and control that understands where the satellite is and how it's positioned and tries to control that. Then the onboard computer, which is the heart of the satellite itself with the imaging unit, the IAC, it's an imaging acquisition component. Everything in space has to be in fancy acronyms, although this is just a camera. And then the communication subsystem, the GPS subsystem, the antenna deployment structure itself, as you can see here on the mock-up of the satellite. So starting to do a bit of design around the satellite itself, then went down into the, and kind of in parallel, into the electrical and electronics side of things, and this is the main diagram of how the different components communicate with each other. We try to stay as relevant and as up to the latest developments around microcontrollers and best practices around them as we can be, which is not typical for the space industry. Like the default for space industries, if it works, you never change it. And that's how you end up running code that has been written 30 years ago for microcontrollers that were designed 50 years ago. So our main microcontroller is an STM32, various different flavors of it, depending on the system. And then you would have the power lines and all the communications among the STMs themselves. And that's how you end up with a master plan about how you're going to be building your satellite. And then we had to start from something in terms of the physical format of the satellite itself. And as this was developed in our Hockey space, the first thing that you do is that you print quickly something that you can see, right? So literally that was the first week of development. Like we had to start from something and be able to communicate whether we meant the Z minus side and then left to the XY side. It's much more usable to actually have a mock-up, like a model that you can point to. And then along the way, we had to discover a lot of the processes. And reading through documentations and trying to figure out the standards and how big satellites are built. We designed many things and discovered them from scratch, starting from clean room procedures or cleanness procedures to also how you integrate and start building things that have a physical form in the sense of a satellite. And gradually over the course of weeks and weeks in and weeks out, many of the things were starting to be developed much more efficiently and trying to work each other as a team really in as parallelized way as possible, given the really short time frame that we had. And quickly, one thing led to another in terms of the design. And we were able to produce really fast and interactive approaches on the open hardware side of things, like the hardware side of things. So we used KeyCAD for the EDA, for the schematics and the components themselves. And the boards were designed there. And then FreeCAD to most extent for the structure itself and some of the analysis around it. And things moved rather quickly. And you can see the progress here within only a month or probably less in this case, where on the right side you can see what we call an engineering model. So an engineering model would be a board that is not meant to be in space, but it's meant to be staying here around your engineers and actually quickly verifying what you do. And some of the components changed, some of the manufacturing processes changed, and are less restrictive than the actual flight model, which you can see on the left side. And this is the exact same board with the same functionality. That's the electrical and power system, the EPS. And then the satellite started to take form by itself. And we need to build and find out how to even build a carrying case, like a case that we can carry the satellite around for different testing purposes, or even for doing vibrational tests and everything else. And that involved a lot of back and forth. So we built something, and then we didn't work over the weekend. We rebuilt it and tried to make use of the Hikerspace that we are operating out of. That's the hikerspace.gr in Athens, Greece. And tried to get access to as much materials and tools that we could in order to progress as fast as possible. One of those things was that during various different phases of the satellite integration, you actually need a clean box, like something that is clean enough in order to assemble your satellite. So basically, here you can see the clean box that we designed and built. We were looking around for different best practices around that. Again, zero open source or documented designs around it, but we were lucky enough to have enough people to be able to concentrate on that and create something that was actually usable. So the clean box and then some thermal verifications, when we first booted up the components just to make sure that nothing is off the chart. You can see the power amplifier on the coms platform really hitting up there when it first got operated. And then one critical component in order to do a really fast-paced and as open as possible procedure of creating a satellite is to be able to give your engineers, especially on the software side of things, give them the ability to work on the hardware as fast and as direct as possible. So one thing we set out to do was that we created multiple engineering models that we distributed to all the core members of the software engineering team so that they were able to quickly verify their code in real hardware. So you can see here one of the engineering stacks that was given to the engineering team. And while doing that, keeping everything in the open. Like we defaulted to open and publishing everything from scratch. From the day one, we just had the public repositories and we were pushing code there and taking everything there because one thing that we found out is that many people say that they're going to be publishing things along the way. And maybe they do sometime down the road. But it's completely different in terms of the mentality that it creates for the engineers and the teams that are working together to get everything public from day zero until the project is delivered. It provides visibility and it gives a much better history of the project so you can see where the mistakes were made and whatnot. In the process of almost completing the satellite, we had to create the telecommand and control capability itself. And that's built on top of SATNOX. You're going to be hearing about SATNOX in the next talk. That's the ground station network that we have. And again, trying to stay as modular as possible so that other missions can reuse this. So here are a couple of photos about the last integration steps of the satellite in the local hacker space. Here we are on the vibrational testing chamber. Unfortunately, we couldn't build the vibrational table ourselves, although this is a challenge that we should accept for the next satellite. And you can see the satellite in there in the vibration box ready to be tested on one of the axes. And here is a vacuum chamber that does the thermal testing. So basically, you put something in there. There's vacuum generated inside the box itself. And then there are thermal cycles that you go through. So plus 40 degrees Celsius, minus 20 degrees Celsius. And that, again and again, is simulating as much as you can the space environment. And we built it also from scratch as we couldn't have access to a facility like this. That's the satellite right after the thermal and vacuum testing. And then, unfortunately, we couldn't build an echo chamber like this, because that's quite huge. This is a military facility in Greece that we got access. Actually, it's demilitarized now, so that's good. And we got access to do some testing around the electromagnetic compatibility and the electromagnetic interference among the components. So this is Manolis Arhapi, RF engineer that the RF components actually worked as predicted. And those are the final pictures of the satellite literally a day before it was delivered to the launch coordinator that then it can be integrated into the deployer as we're going to be seeing. So this is a photo of us basically delivering the satellite to the big clean room of the integrator. Those are some final touches on the satellite itself. One of the science components had to be integrated at the end of it. And this is the official picture of delivery of the satellite. And that happened August 2016. And from that point, you just wait. There's nothing you can do. You have delivered your satellite. You're just waiting for other satellites to come in, get delivered, be integrated inside the deployer box, which is basically an aluminum box with a spring on one side and then a door on the other side. And that gets in space. The door opens. The spring pushes the satellites out. And that's the deployment. There's nothing more fancy around it. So you can see UP-SAT on the left side. And then barely see the next cubes next to us. That's a France satellite FR04 from QB50 program also. So that happened August and September 2016. And then once everything is integrated, they have to be exported to US, get into the delivery spacecraft to the International Space Station. That was our launch procedure. We would be delivered to the International Space Station. And once we are up there, we would be deployed from the ISS. And we were waiting. And we were waiting. And for those of you that have been following space exploration and space time schedules, it can be really frustrating. Like one thing leads another. And there's always delays on the launches and everything else. But the magical moment came. And that happened on the 17th of May, 2017. So this year, basically. So we were waiting for almost eight months, basically, to get to see our satellites in space. And once you are in space, then you have to wait another month in order to be deployed out of the ISS. And that's the magical moment of deployment. You can see the people opening and pushing the three satellites outwards. And the first one would be our satellite, and then two others following that. And you can see the glorious solar panels of ISS, too. It's a really nice picture. Thank you. And then you wait again. But this time, you wait for a short time. And you have to wait for basically 30 minutes. The regulations state that once you're deployed from ISS, you have to shut down everything for 30 minutes and then start pulling everything in operational mode, because you have to be far away, enough far away from the ISS itself, in order to not interfere with anything which is a human-arrated environment, because there are astronauts actually in ISS, and you wouldn't want to upset that part. So we have the Global Ground Station Network. So we were scheduling passes of the predicted position of the UPSAT, and rightfully so, 30 minutes after the satellite, we got word that the satellite was deployed from the ISS. There was one pass of the UPSAT above one of our ground stations in Bloomington, Indiana. And that was 3 AM in the morning from Bloomington, and we were all actually traveling to US for a different conference. And we're anxiously waiting to see if UPSAT is alive. And you can see in the middle, I cannot even point to that, but really in the middle of the waterfall that you see here, you can see the CW beacon of the UPSAT, some anxious minutes of decoding and trying to figure out if it's actually our satellite or someone else because we have multiple deployments of satellites, and then more clear passes come along from the ground station network, and we can clearly see that this is us beaconing. So super happy times for the team that it actually worked. And then we started monitoring the process and how the satellite is performing. And you can see some of the initial diagrams. That's the charging cycles of the satellite itself. We're pretty happy that most of the components were actually working OK. Not everything worked as expected. And that's unfortunate to some extent, but really that's to be expected, especially first timers in space missions like us. And although we could see most of the things working really well, one of the things was really not accounted for. And that was the consumption of the heating around the battery component itself. So we had to hit the batteries in order to make sure that they're not degraded if they fall below 0 degrees Celsius. But at the same time, the consumption of those heaters was actually large enough that we expected more thermal insulation than we actually saw in space. So we had to overuse that. And right now, we are in the phase that the satellite is operational, but at the same time, most of the time is actually in safe mode in terms of power itself. So it senses that it's going below the power that is safe enough to use all the different components. And it protects itself by cutting down the various different components. And then once it charges again, it beacons and says that it's OK. And then same thing goes on and on. That means, unfortunately, we cannot perform the science mission to its entirety. But still, if we can communicate and actually have these beacons, that verifies that many of the things that we built were designed correctly. So what have we learned from that? That it's really not rocket science? Well, it is to some extent. But it's really not as hard as people think it is. And the critical thing for us would be to make sure that no one has to go through the same ordeal that we did in terms of lack of documentation and lack of open code or hardware or best practices in order to create the mission for themselves. So what we set out to do after UPSat was basically to try to find who else is interested in open source satellites. So we co-organized an open source CubeSat workshop, the first of its kind, in the European Space Agency, Facilities in Darmstadt. And that happened almost a month ago. And that was pretty successful for the first gathering of people around open source, especially if you consider that open source was not really a word shared in space industry up until really recently. So we would like to extend the call to all of you and all of you watching the streaming right now that if you have in mind something which is related to open source, it's open source and it's related to space technologies, feel free to reach out. We are creating a much diverse and global community of people, which we named the Libre Space Community, and we'd like to claim space the Libre way and make sure that no one has to reinvent the wheel and we can really work on innovative projects at the edge of today's available things in technology. So thank you a lot for your presentation. So that was Piero's paparazzi. We still have a little bit of time for Q&A. If you have questions, please line up at the mics. We have four in the whole area, so just line up and you will be called up. We'll just start up with mic one. Yeah, thanks for the interesting presentation. I actually wondered what you would guess, how long it would take actually to get some open source satellites with inverted band pipe transponders up into low orbit. Like, you know, we have SELSAT2, which is going to be up there hopefully in 2018, but it's on a commercial satellite from the Qatar Satellite Company, they're in cooperation with AMSAT and so on. But what would you guess like getting such a transponder only as an open source project with no commercial and no secondary status of actually the project and payload itself? So that has to come from close collaboration with the AMSAT communities around the world. I'm aware of phase for ground and phase for space, which are AMSAT-backed projects for such specific technology aspect of the transponder itself. Unfortunately, I think that the pace that this is going on right now is really slow, so I wouldn't expect anything within 2018 as it stands right now at least. Mic two, please. How do you cool the electronics in vacuum? Is just the radiation enough to dissipate it or do you have some sort of active cooling as well? So no active cooling, you have to rely on radiation and actually touching of thermal transfer through actual physical contact between the different components. Okay, mic four, please. Hi, thanks so much for the information. I was wondering about the GPS chips. As far as I know, consumer available are locked, so what did you do about it? I'm wondering, thanks. Yeah, so that's one of the black spots of the open source side of things. And the concept is that in order to use a GPS in such altitude and such velocity, you have to have a chip which has a license to bypass the COCOM restrictions, which are the restrictions put in place for GPS. So you have to buy that and this is not open source when you buy the actual GPS receiver. Consumer available? Yeah, although it's pricey, like you can get to the couple of thousand. Mic one, please. Hi, so first of all, thanks for doing the hard work for us. I think you did some cost evaluations, so is there any numbers you can give us? Yeah, so the total cost for the hardware of the satellite amounted around $30,000, and almost half of it was actually the solar panels, which we should invest on creating open source solutions about actually doing that. And then for the whole project itself, like together with the engineers that worked on it and everything else, like I would say that the total was around 120,000, like together with the launch and everything, like the entirety of the project. Um, Mike Four. Two little questions. The first question, can I as a random person try to communicate with your satellite? Okay, I see, he's nodding, okay, great. Then there's the next question, is this alone enough to postulate kind of open source global satellite battle mesh, like taking a fly funk to another level? Yeah, all another level. We would be really interested in talking with anyone who's developing networks like terrestrial networks right now, or actually how we can move this to space. And yes, we're super interested about that. There's been a couple of ideas, so feel free to reach out or actually do reach out, not feel free. You or anyone else that has interest on creating networks of satellites, we actually have a table on the fly funk guys. There's a little bit of space foundation table and there's a ground station you can find it for sure. So, yeah. Mike Two, please. So the Cube set ideas are already a little bit older and stuff and so what I was wondering as it is also like used commercial by companies like Planet, we are doing these imaging solutions and stuff. Have you had any experience working with these startup companies who are also working with basically the same technology as you are? And do you think that there's a future of maybe learning from each other? Yeah, I think that the answer to that lies on the difference on the business models around those initiatives. And as long as the business models of the current conversion of the self suppliers of satellite components is based around actually licensing the code and well, paying for that license and paying for getting access to any kind of documentation or even the designs itself. Like most of the times you don't even get access to design even if you pay, then that's completely different from our approach in terms of how we create sustainability around that. I think that there are ways of collaboration but really that has to come through hard work among each other in order to find the sweet spots on that. But for us, not working in the open is not an option. So that step has to come to some sense from the other side too. So I think we're through, we don't have any time anymore. That was Piros Papadeas, thank you very much. Thank you. Awesome talk.