Everybody has this confidence that, okay, if everybody do their part, we will make it. Of course, it will be hard. There will be long nights, weekends. But of course, when you get that annunciation from our CEO, everybody looks at each other and be like, oh damn, this will be interesting. For most of us, satellites are something abstract. We can see them at the nighttime as a small distant dot moving around the Earth. Something that gives us GPS, internet or weather forecasts. Something that just works. But behind every satellite are years of troubleshooting, compromises and people spending thousands of hours trying to make something survive in an environment with no technicians, no repair shop and no reset button. And right now, Denmark is part of that journey with Bifrost, a Danish satellite mission focused on monitoring the Arctic. But long before a satellite becomes a mission, it starts with a prototype, it starts with failed tests and countless small problems that certainly matter a lot. Because once a satellite leaves Earth, everything is locked into metal, vacuum and physics. If something goes wrong up there, the solution has to be built in years earlier. So in this episode of IDA Space Talks, we're not just talking about satellites as a technology, we're talking about the experience from concept to countdown. And of course, we'll talk about the launch itself, the strange moments where years of work leaves the planet. And all you can do is wait for the first signal to come back. My guests today are Rama Murali and Matthias Ernst Halluarsson from Space Inventor, who have been part of turning Danish satellite ambitions into something that actually reaches orbit. Welcome to IDA Space Talks to both of you. Thank you. Thank you. Thank you so much for hosting us and inviting. Welcome to this English edition of IDA Space Talks. IDA Space Talks is a Danish podcast focusing on everything space related. The podcast is published in collaboration with the Danish Society of Engineers IDA and the professional network IDA Space. My name is Tino Tonnesen and I'm your host in IDA Space Talks. We are here in the ETH studio and I think just introduce yourself shortly. Yeah. So my full name is Rama Murali GK. I will cut it short there. It's a bit longer. And I've been in the space industry for the last 16 years. Yeah. And I'm part of Space Inventor journey for the last almost one year now. And the professional life has mainly been working with the satellites. The space systems making things happen and yeah. Then taking up different types of mission all the way from lunar missions, Mars orbiter missions and the satellite communication systems, application missions and even the programs. And now I had the assembly integration and verification at Space Inventor. So making it happen in the sense, physically making it happen. Okay. And taking up that challenge of, we have been talking about the space missions for years. Like one mission takes four years, five years or more than that. And now we are making it happen in months. Oh yeah. When the whole world talks of months, we talk of weeks. Yeah, yeah. So that's what we are aiming at Space Inventor. So how we can make it quicker. And I'm happy to be part of that journey and represent Space Inventor today here. And then Matthias. Yeah. So my name is Matthias Ernst Talwassen. I am the head of the mechanical engineering department at Space Inventor. As a background, I'm a mechanical engineer. I've worked several years in robotics and automation and joined Space Inventor five years ago, back when it was still what was considered to be a startup company. So we were less than 20 people. We were in a normal office building without much production or test facilities. So I've been a part of that whole scale-up journey to the company where we are today, where we are building, flying, operating microsatellites up towards 200, 300 kilos. Okay. It's still a microsatellite if it's 300 kilos. Yes. Okay. I want to like rewind and talk to you about how it was to start everything up. But can you just briefly tell me what is Bifrost? Well, Bifrost is a cornerstone mission for us. Briefly, it is a surveillance mission. It is the first Danish microsatellite since Öster satellite many years ago. It is the first Danish military satellite and it is operated and run by us together with the Danish military. Okay. And it's a surveillance mission. The mission objective is to detect dark ships in the Arctic and Northern Sea. And dark ships is? Dark ship is ships that has shut off their transponder system. So the GPS system. But they still use radio communication. Yeah. So there is a radio receivers on the platform or on the satellite that can then detect if the ship is communicating over RF. And then we can say there is a ship in this location and send that location to the Danish military to investigate further. Okay. Basically, when you talk about dark ships, just to add or give more clarity. So every ship has a certain size, bigger than certain dimensions like 30 meters. Okay. They should have a transmission called AIS transmitter, automatic identification system. Transmitter fitted on board the ship. Yeah. It is basically has multiple applications. One of them is to anchor them in the harbor to navigate it. Also, it has multiple applications. Like it has to continuously transmit. Mm-hmm. So it transmits the data about the ship. So this is the characteristics of the ship. And this is the origin of the ship and everything. Yeah. And it is legally every ship is supposed to do. Yes. But then the ships, which they don't do that for a purpose. So it is very important for from national security point of view to identify that ship. Yes, of course. Now that how do we identify when somebody is not talking to anybody? First, you have an image. From that, you know that ships are there. But you also have the data from different AIS transmitters. Okay. These are the ships existing. Now you fuse those two data together. So when you have image and when you have the AIS data, so you know which ship is not actually transmitting. So this ship, so you try to map it in a database. So this all happens in AI, the algorithms, as I understand. So this is the key part of identification of the data. Dark ships basically means. And I guess in this time and age, it's more and more relevant. Exactly. And geolocating that. Yeah. Where exactly is that ship? Yeah. And that happens through the RF system. But geolocating and identifying those ships. All right. That's the critical for any country. Okay. And that is basically what Bifrost is supposed to do. Yes. And what it is doing as we speak. As we speak. Yeah. Because it was launched in June. Yes. Last year. Via SpaceX. Yeah. Via the Vandenberg station in California. Yeah. I want to get back to that. But let me just rewind a little bit and go back to the beginning. Because it's been a long journey for you guys. Many years, I guess. Can you remember where it started? How do you come up with, now we want to build this satellite? Was it the Danish military content? You said, build a satellite. How did it start? Well, Bifrost actually started in a, it's been through multiple iterations. Yeah. So, it's a very hot potato. Today, this military surveillance of dark ships and all this stuff. We actually started the project four years ago. When nobody was talking about this. Yeah. Before the war in Ukraine. Before there was mainstream media covering dark ships. All of this stuff. There was a goal from our founder to actually make the first military satellite in Denmark and provide it. Yeah. And provide this surveillance capabilities. Because there was a huge gap in what we were able to do. And we have a lot of sea territory. Yeah, yeah, yeah. And there's no way of covering that with ships or with drones or planes. So, the only reasonable choice would be satellites. And our founder, Karl, knew that many, many years ago. So, we actually started pitching this to the Danish military five, six years ago. And they were not really buying it. So, at some point, four years ago. Because it was a different world then. Very different world. Very different world. Very different military budgets. We were going down the military budgets. Nobody was really prioritizing that route of spending. No. But he had a goal. So, we started internally in the company developing and designing this micro satellite. One day in the afternoon, he actually went into our office and announced to the people that now he has bought a launch for a micro satellite. Really? Yes. A hard deadline. It was a hard deadline. I think it was in 17 months or something. Yeah. So, we were still on the drawing board. Nobody has designed anything. Nobody has tested anything. We didn't even have a customer or mission. No. But he bought the launch. Okay. Because then everybody had a deadline. We know what we are working towards. Yeah. And then the company started working towards this goal. Was there like a silence in the room for one second? Yeah. People were looking around each other. Saying, okay, building a micro satellite, the first micro satellite since Ørsted was a big deal. A very big deal. Yeah. Yeah. It is. Definitely. Yes. So, everybody waited for a second. But then I think everybody caught on and said, okay, this is necessary. This is what's moving the company, the future. Yeah. So, everybody was on board. And then ByFast actually started at it being a very small satellite. A CubeSat. A 6U CubeSat. Then it grew to a 12U CubeSat. Then a 16U CubeSat. And then finally, we convinced the Danish military that in order to get the technology that was giving them the most amount of value for the data they could pull down, they needed to fly our new platform that we in the meantime have designed. We already have bought the launch. So, we could also give them a deadline saying, this is when it will fly. Yeah. Do you want to be the first customer for this? And they agreed. It was like plug and play for them. Said, okay, here it is. This is what we will deliver. Yep. And when you say micro-satellite again, the ByFast, I read it's around 50 kilos. Yep. Around 50 kilos. It's, say, it's half a meter times half a meter times half a meter. What are the challenges in building a satellite? Where to start? Yeah. I have never built a satellite. So, there are multiple challenges for the space inventor as such. We are a fully vertically integrated company. Yeah. Now, what it means is starting from the module, the harnesses, everything is built in-house. Yeah. Yeah. So, we source components. We build the systems, the subsystems in-house. Yeah. And that's what makes it more challenging. At the same time, it's more efficient in terms of delivery durations and flexibilities. So, but that comes with a lot of challenges because you need to make the subsystem level work and then the satellite level work. And then you need to launch it and make it operational in the orbit. So, challenges are everywhere. Yeah. Okay. Now, that is more challenging because if you are just looking for a subsystem, the units to be delivered from different suppliers and integrate and launch. Then you can lean on their subsystems and ask them, okay, you provide me the perfect system, I will build and integrate. But when I need to do it, everything, then the challenge starts from the scratch. Yeah. Yeah. And then, again, these subsystems are not just taken and built. So, for each mission, there are set of goals. So, let's say one mission needs a specific set of goals. Those goals are further divided into different requirements and those requirements are further broken down into different specifications. And those specifications goes to each module. Then you pick the modules. Okay, this is my mission requirement for to meet this. This is the module I need. You're breaking everything down. Exactly. So, you break down the requirements and start building from and realizing from the bottom. Yeah. And this whole process, now where there is no challenge, tell me, because breaking down the requirements, if you miss one requirement, you are missing something in the bottom, which you can never track. No. Now, while building up, a module may be working as a module, but when you are putting multiple modules together, then, you know, it's just like people. So, you just need to make the whole system work. Yeah. And to add to that complexity. So, we have a very modular, flexible kind of architecture in our satellites, wherein each module, like let's say you need higher power, you can stack the modules up and increase the power. Yeah. And all the modules have a very nice architecture in such a way that it's highly software integrated and it's highly flexible. Yeah. Adding flexibility to a system makes it more complex because you need to test for all the cases. You cannot have even a single part missed in the testing because if that missing element gives you problem on the space, you cannot fix it. No, no. So, you need to ensure. So, adding software, making it more flexible and making it work and delivering what is the mission requirement. So, these are all some of the elements. I'm sure along the way, something went wrong. Have you got like a favorite failure moment or something? And what happened? For me, I think there are plenty of moments. Nothing will go perfectly as we designed on paper. There will be challenges. There are failures and there are lessons learned. I wouldn't put it as a failure in a space system. It's more like a lesson. Learnings. We always will have learning. Failure was the wrong word. It's definitely like, so when we move along, we learn, okay, one mission, we launched one mission. We had some observations or we had some challenges. We fix it and move forward. The major or the aim is not to repeat the mistake and to fix the processes such that the mistakes doesn't repeat. So, maybe specific incidents I would request Matthias to think about something. Yeah. There must have been at some point something behaved differently. You said first you started putting CubeSats together and then you changed the whole thing. In that process, what behaved differently? Was there something that you thought, how do we solve this? Because you had a very hard deadline, I guess. Yes, we did. There was multiple things that went wrong in that scale up. But I think one of the very key things that was a vision from the beginning, actually from all the way from the beginning of the company's start, was the goal was to end up building microsatellites. So, even though we spent the first five, six years building CubeSats, the modules that was developed through that time period was always with the long-term perspectives looking forward, saying we need to be able to scale this to bigger platforms. So, when we did the Bifrost mission, all the modules we have been designing for the past five years, they were already ready. Yeah. They were flying in multiple CubeSats. We knew the telemetry we got from other missions, from orbits. We know how to work with the modules. We just needed to work with them in a bigger scale, adding more modules, this flexible modularity we have as our concept for our modules. Of course, there was challenges because as you just add more modules, you add complexity. You add more test scenarios, things can go wrong. I think one of the new things we learned a lot with Bifrost and testing microsatellites as well was when things get to that size, it's hard to test something. You want to test everything on ground, but here we have gravity. Yeah, yeah. One of the things that's very hard to test in gravity is solar deployable wings. Yeah, yeah, yeah. On Bifrost, we have double deployable wings, so the wingspan of Bifrost is 4.6 meters, I think. Okay. So the satellite is only 500 millimeters, 50 centimeters, but the wingspan is almost 5 meters. And that is deployed by just simple hinges and spring mechanisms. And that, of course, works in the zero gravity. Yeah, that's the sign for zero gravity. If we were to design it for ground, it would be totally over-engineered and the wings will almost explode when they come into space. So one of the challenges we had was to actually do the deployment test in the integration phase and verify that, okay, we can deploy in zero G, but we also need to verify it in 1G, which is very different. So that was a matter of building deployment rigs that would then support the panels without over-supporting them. So we are simulating a zero G environment. But that for sure involved a lot of back and forth because that was... I can imagine. That was difficult. And how did you solve that? We built a rig that would support the wings, basically. And then a string and a wire system that then will try to track and keep the same tension on the wings while they were deploying without pulling in them. Because we wanted to be... The deployment would be guided by the deployment mechanisms, but only supported. So there's a big string and pulley system in one of our clean rooms that was used to suspend the wings on the deployment. It sounds also like a complicated setup. It was just to tick one box in the test schedule. Then you still have 300 plus points you need to check off. Yeah, yeah. Yes. Wow. Yeah. Again, there could be challenges in terms of technology, like something we miss to understand or which is mission specific. We realize during the phase of the mission. Or there could be even human errors because we are all humans. We make mistakes. Yeah, yeah. But the best part what I had seen, for example, is I think within a few couple of weeks I joined, we fried a module in the clean room. We just connected the harness wrongly. Okay. And it just smokes. Now, the thing is, it's not about that failure. No. The next day we bought a cake and celebrated that failure. And we ensured that we don't repeat the mistake. Yeah, yeah. I think that culture of accepting the mistakes, learning from it and growing and not repeating it, that's what takes us forward. Of course, yeah. So, yeah. And these are some of the good things in the space industry, as I've seen. Like, we all do mistakes. There are challenges. Yeah, yeah. But how we cannot repeat the same thing. I guess that's the only way to learn. Yes. At least it's the fastest way to learn. Yeah. Yeah. Yeah. It is. Isn't it also like a very expensive way of learning? It depends. I think then again. Yeah. Like, today you have plenty of simulators to simulate and see things before doing the hardware. Yes, we can lean on them to a certain extent. Yeah. But how accurate they are, how close they are to the reality. And some of the things without really putting the system in the extreme conditions and test it, you can never. Like, how do you simulate a rocket? How much you can simulate without actually launching it? So, these are some things you learn as you do it. Yeah. I think it's also, it's where we really benefit from being vertically integrated, building everything in-house. It's cheaper for us to build something and test it instead of spend 200 hours simulating every kind of event. If we build it in our own factory and try it out, if it smokes, if it fails, we go back to production and we can have a new module to test within 24 hours. Yeah. And we try again. Yeah. We have all the capabilities to do it in-house, so we can do this rapid test and prototyping. And if we were a company who was sourcing other stuff, it would be totally different. Then we need to make sure that we were getting it right every time in the first try. Yeah. But it would be a lot slower than what we are doing. It sounds like a luxury, actually, to go to the workshop and tell them, okay, I just blew this up. I just burned that fuse. Can you make another one and do that? That is the thing. But then again, it's up to everyone as well in the company that even if we have the luxury, do we do that every time? We just try not to do that, not to get in there. But then if something happens, it is okay to do the mistake. That's the thing. Of course. Yeah. Cool. It depends on where. If it's in Morales areas in the clean rooms in AIV, it's sort of a big deal. Yes. It's a lot of non-conformance reports, root cause analysis and everything. If it's in the R&D department, it's more fun. It's not as serious. Fun in what way? Then you get to point fingers at your colleagues and tell them that you need to bring cake tomorrow. And then you go down to production and ask for a new module and then. And then they demand cake as well, I guess. Yes, they do. What should you bring if you did something in the clean room? A very big cake. I got to ask, what could mess things up in the clean room? Was just opening the door. No, the ultimate priority in the clean room is ensure that no mistakes are done. Ensuring highest quality and highest reliability throughout the AIV phase, starting from collecting the modules from the production into the clean room till delivering the satellite to the launcher, integrating the satellite to the launch pad. That's the last room that the satellite enters. That is the whole AIV process where AIV team is part of it. Okay. So that is the most critical part here. And that's the most, I would say. So that means doing things with very thorough inspections, thorough the non-conformances. If you see something, report it, discuss about it, come out with the solution and then implement it and move on. Okay. Don't just ignore and move on. That needs to be discussed. Even if it is a small observation, it could be the observation in the process of doing it or it could be a technical issue or it could be a personal issue. Like somebody would have moved some things. It could be even that. Yeah. Recording every activity in the clean room is also key because after a couple of weeks, we would forget what we did two weeks back. So then you should be able to go back and see, okay, this is what I did two weeks back and this is what it is today. So doing those kinds of analysis, everything becomes key there. And ensuring that we don't do the mistakes. There is procedure need to be followed. We need to be inspecting. We need to clear the route. We need to have a plan in place. Everything needs to be thoroughly discussed before we do it. Yeah. You're moving quite expensive, delicate equipment. Exactly. I keep thinking about you getting this hot deadline. Was there a point in the beginning where you thought, are we going to make it? And then was there a point when you had confidence enough to say, yes, we're going to make it? Or you believed in it from the beginning? No. No. I mean, it's very much part of the DNA of our culture in the workplace. That we believe in the overall objective and the end mission. Yeah. And I think there is a very keen culture that has been carried out from back when we were 5, 10, 20 people of this united workmanship. That everybody can move together, focus on a objective and really move things very, very, very fast. So, everybody has this confidence that, okay, if everybody do their part, we will make it. Of course, it is a fast stretch. It will be hard. There will be long nights, weekends. But if we move it together, we'll get there in the end. Okay. But, of course, when you get that annunciation from our CEO, everybody looks at each other and be like, oh, damn, this will be interesting. I think that's the fun in it, right? Yeah. If there are no challenges, why are we there? Anybody can do it. That's true. Yeah. And I think it's, I mean, working in the space industry, for many, including myself, it's a lifelong dream. It's a thing that you strive to work in that industry. It's something you really believe is unique and something you can be proud of. And that most everybody in the company has the same feeling. So, it's a mix of your work, but also your passions. I had even instances where I had to push my teammates to their home from the office. Just go home now. Enough for today. Really? That's the passion with everyone, like the passion we have at work. I guess you made the deadline, of course. Tell me about, you followed the launch of the satellite. Tell me about that experience. I mean, you've been working on it for years. I have been part of many missions where the satellite I have worked, I've seen it from the mission control center going up. Yeah. Like, for example, the Mars Orbiter mission I worked in the past was one such thing. Okay. And that sound of launching, I would say if anyone appreciates, you wouldn't get that much satisfaction. Okay. That is the feeling. And I'm sure even for us here, every satellite we ship out, we all get the same feeling when we send the ship, like satellite out. Yeah, yeah, yeah. When we go to the launcher or launch integration, we integrate the satellite and come back. We get the similar satisfaction every time. And the final moment when you let the satellite go, it is definitely the moment worth that effort. Okay. And when you say let the satellite go, in space? It could be in space. That is definitely an amazing feeling, even from sending the satellite from the clean room out. Because we would have toiled like days and days, nights, everything in the clean room for that satellite. And we see that go successfully test, completing the testing. Yeah, yeah. That's really like the most happiness we get from there. I can imagine. How does it go from the clean room into the rocket? How do you transport it? It's like a mobile clean room. Yes, for the big satellite it is. For Bifos, for instance, there is a big transport case. So it's kind of like a big metallic box where you can lift the entire top off. And then there is a big base on wheels. Then there is a fixture in the middle that is suspended by wires to all corners. Okay. And then you mount the satellite on that through the interface ring. That is also what you will mount it on on the rocket. Okay. There it's protected against shock and vibration. So it's suspended in a wire system. Yeah. And then you put on a top that is then purged with nitrogen to seal everything off. So it's a mobile clean room. And then this big aluminum top on top of that. And then you ship that. And you ship it with UPS or for snow or what? Or Dow or yes. No. You buy a dedicated freight systems. They also need to be insured. Yeah. Yes. There's a procedure. And we normally, it's one off. Like we cannot just come by and put it in. No, no, no, no, no, no, no, no, no. And then usually there's a team, a mix between the mechanical engineers and AIV engineers who go receive it at launch site. Yeah. Okay. Do some final checkouts of the satellite to check the health, check battery levels, check that all of the computers' autonomous systems is set up correctly for launch. And then assist in integrating it on the rocket. And put it on the right shelf. Just also to ensure that the shipment has not disturbed the satellite. So it's still fine. It's a last minute check, I guess. It's more like our babies. Then the satellite goes up in the rocket and then it's launched in space. And what happens? And then you get like the first beep and then you realize that things worked. Yeah, basically. I mean, we do a lot internally to celebrate each launch for every satellite. So usually there is a launch event for all the satellites we build and launch. So it doesn't matter if it's three in the morning on a Sunday or 12 in the afternoon for Monday. Yeah, you have to celebrate. Yes. So the entire company gathers at our main offices and then there is cake, beer, champagne, all that stuff. Then we watch the rocket. Then we see the separation from the rocket. Yeah. And then everybody, at least a group of people go down to our mission control center and wait for the first pass to be live. Sometimes it's straight after deployment for the rocket. Sometimes it's 30 minutes, sometimes 40 minutes. But then they wait and everybody watch our group chat to see when they display. Usually what they say first is the battery voltage to say, what is the current state of charge? And are we power positive? Are we generating power or are we power negative? Are we losing power? Yeah. That is the first thing we are interested in. Yes, of course. Of course. If we know we are power positive, everybody is good. Yeah. We are getting power and we have communication. That's the first sip of champagne, I guess. Yes. And after that, what would be like the second and third sip of champagne? What else do you check? Then you check all the systems to see if everything is live. You check all the deployments. Do we deploy solar panels, antennas, whatever it could be? And then pretty much prepare to go into some of the ADCS modes. So we have autonomous modes that could be sun pointing mode. So pointing a satellite in a direction where it generates the most amount of power. Could be a ground mode where it points the antenna towards a certain ground station to ensure to have a good pass to send down data. Yeah. And all these things. And then we go into what's called the free axis control mode. So that's where all the operation modes, the ADCS, the magnetorges, the reaction wheels, the fine sun sensors, star trackers, all that stuff is working together to ensure that the satellite is stable. How do you try to figure out what could go wrong and how could we fix it if it goes wrong? You do a lot of testing of different scenarios in the clean room. Yeah. Okay. So we do a lot of testing on a systems level on a satellite, preparing to see, trying to configure the satellite to be as safe as possible. So even if a error occurred, if it were power negative for some point, at some point, it will go into a power safe mode. It will shut off all the non-critical systems. It will only keep on the radio for ground transmission of telemetry. So we can communicate with it. And then slowly, it will start turning on, for instance, the magnetorger. That would stabilize it if it was spinning. Okay. And then de-spin it over a certain amount of time. Yeah. And then we would try to get this stable communication again. And then we can slowly start powering up each system, powering up the reaction wheels, powering up the star tracker, the radios, the OPCs, everything to actually go one module at a time to get it stable. So even if pretty critical errors occurs, there is a lot of safety in our systems, in our software architecture that protects the satellites so we can still recover it. Okay. So you can restart each module basically? Yes. So there's also automatic restarters. Every 24 hours, every module will power cycle. So if there is a module that goes into a weird state where it's keeping communicating over our communication bus, so it's jamming all the communication for other modules, every 24 hours, it will restart. So hopefully that would fix if there is an internal block. Yeah. Also, there are something called as FDIR, the false detection and isolation logics built in in the system, which can actually detect the faults within the system. And then it can isolate it. There are logics built in in the satellite which can do that. So that will try to fix the issues at the first level. Still it is. And there are redundancies on board. Like let's say one system fails. There is another system which can, till you fix the first system, the second system can take over and do it. There are redundancies and even beyond that it triggers. Then we always have the ground control where we, through the telemetry, we can see, okay, this system is not behaving. Then we need to fix it in the next pass or we plan it and do it. So that's how normally the... Redundancy is key in space. So every module we make and produce, so it could be a power distribution unit or a computer unit at OVC. If you look, if you were to take the lid off the module and look at the PCB, it would be mirrored. So it would be a full module on one side and full module on the other side. So it would be internal redundant. So if one side of the module fails, it would automatically switch over to the other one. To the other side. Yes. So that's, we do that. And then on top of that, we also add extra modules on top. So more redundancy. So for instance, if we are flying a mission, we could maybe use two computers of our computer system, but there would be eight available. Okay. Yeah. So you have just the next in line that will take over if one fails. Yeah. It's like a spare computer and a satellite. Yeah. It must cost extra then because it adds weight or... Not much. It's part of the internal, how we build our subsystems and modules. So within our computer, we have four cores. We have four computers in one. Maybe we will use one or two of those. And of course, then if you add one more computer, but our modules is 160 to 200 grams and eight millimeters tall. Nothing. So it's nothing. It's better to have it and not use it. Of course. So I think it's something to do with the architecture that we have as well. So we have more like a distributed computing kind of an architecture where it is not focused on OBC controlling the whole things, where the systems can take care of itself. It has its own processes. They can talk to each other in a different way. So that flexibility and modularity adds to and helps the faults or fixing the issues. Yeah. So now the satellite is up there and it's working. Do you hand over the control to Danish military or what happens now? It really depends on the customer and on the mission. Yeah. So we have our own operations team. We have our own ground station on the top of our building. There's a big powerbore antenna. So we can keep communicating ourselves. We also have partnerships with different suppliers. We can keep the entire control and just forward the data to whoever they want. It depends what the customer is. Some of the customers are technology companies or satellite manufacturers themselves. So they know how to operate a satellite. Then they will take over the entire satellite operations after commissioning in space. For other customers, they don't know anything about satellites. No. They know something about the data we are getting down from satellites. Yeah. Okay. So we would just forward them the data. Basically, we support all types of customers. Yeah. If some customer says, we want to build the satellite ourselves, why don't you give the subsystems, the modules? We support them. We provide the modules. Or some customers may say, okay, why don't you build the satellite and give us? We will put the payload in. Yeah. We support them as well. We build the satellite and provide it. Some customers may say, we will provide you the payload. Why don't you build a whole satellite and help us to launch? Yeah. We will support them as well. Or some may say, okay, I don't know anything about the satellite, but I need this solution. Can you give us? We can even provide that. Yeah. Yeah. And yeah, so that way we are a bit flexible and based on the mission requirements, the customer requirements we support. That sounds very flexible. Definitely. Definitely. I read that you also use AI as part of the Biforst mission. Can you elaborate on that? Because AI is a big thing now and everyone's talking about AI. How do you use it? Well, we use it a lot internally for design, automation, scripting, all that stuff. For Biforst specifically, there is a computer that also uses AI and AI algorithms. It does cloud removal, for instance. So if you take a picture from space, of course there will be clouds. Yeah. So it detects clouds. If it's an image that is filled with clouds, it will not download it to Earth because that takes up all the bandwidth available. So it detects, okay, I am 80% of this image is just clouds. It will just be discarded. Yeah. If it's maybe 60% clouds, it will isolate the part of the picture that's without clouds and then only forward that down. Everything is about making that link budget. Close that link budget to see how much data can you actually pull down every pass. So the satellite is orbiting the Earth every 90 minutes. You have maybe 10, 15 minutes to communicate with it. You want to take advantage of every second of that 15 minutes. Yes. To pull down data. If you are downloading 100 images, that's just clouds. Yeah. Okay. It's not really worth it. Yeah. In the conventional satellites normally, as soon as the satellite takes the image, it doesn't have the intelligence. It just dumps to the ground stations. That means whenever the ground station is available, it just starts bumping. And then after downloading, you realize, okay, this was not a good data. It had full of clouds. Oh, so the AI is actually ensuring the quality of the data. So you send down only useful data, I guess. Exactly. It's making it more efficient. So we use the bandwidth we have available to the best extent there is. Yeah. Otherwise, we would need a lot more ground stations, a lot more expensive antennas, a lot more expensive receivers to actually pull down enough data. And then we need to sort it in ground. Yeah. Yeah. But that's way too inefficient. Also, when you have a technology on your hands, it's better to use it efficiently for the mission. You could also do ship detection, automatic correlation with the AIS database, all that stuff. Yeah. That is something that could be implemented and worked on over time. What are you doing now? Now Bifrost is up there and what is the next thing for you guys? Is it the new heart deadline of a bigger satellite? Yeah. We are working on the next generation of the Space Inventor platforms. So now we are working on platforms that will be 300 plus kilos. Okay. So something that is five to six times the size of Bifrost. And why does it need to be that bigger? Does it have more equipment or more sensors and cameras, whatever? We can accommodate payload, mainly optical missions for big optical missions with optical payloads, telescopes for a certain size. If you are flying in an orbit of 600, 800 kilometers and want to take very detailed pictures of Earth or out of space, you need a telescope of a certain size. In order to fit those, they are large, they are heavy. You need a platform to support that. Yes. Yes. So that is it. And then we are also working on our Marni mission, the moon mapping mission. Yes. Which is very exciting and would be a huge project for us and for Denmark, I think. When is the heart deadline for that? Well, we are still in negotiation with ESA. Okay. ESA wants to do it. I think they are a little bit concerned if we can do it as fast as we are telling them we can do it. So we are trying to work on the middle ground to see when the heart deadline is, but we would like to fly it sometime in 28. So every time your CEO comes through the door, you're all, what's now? It's a heart deadline coming through. Yes. Great. We are looking forward to that definitely and some new data and everything from the, especially the Marni mission. I mean, it's like the Danish moon mission. Thank you very much to both of you for joining my podcast. Thank you so much for hosting us and it was very nice catching up. Yeah. Thank you.