Swarms of electric animals will go to the moon

Hello and welcome to Space Cowboys podcast number 22 if I'm correct. Hi Thijs. I lost count. You lost count. I lost count. I think I believe one of the previous episodes I actually miscalculated. We did. I don't know and then I think we also misskipped one. So to all our fans could you please help us count how far in we are. I'll go back to the list over at bnr.nl and I'll count. Yes exactly. And again it's not about quantity it's about quality. And I would like to thereby introduce our guest. Now would you like to introduce our guest. Our guest is Chris Verhoeven. Maybe we should say Chris Verhoeven. Chris Verhoeven. Yeah Verhoeven. Verhoeven yeah. When I say this abroad then they write my name correct. Okay great. Chris Verhoeven. You're an employee. Employed at the Technical University of Delft. Exactly. Delft University of Technology we should say I guess. Delft University of Technology. Yeah okay. What's your job title again? I'm not good at official stuff. I'm an associate professor. Associate professor. And I'm team leader of the swarm robots team of the Robotics Institute. We're talking swarm robotics today. And of space robotics in the Space Institute. Okay. Lovely. Swarm robotics and space robotics. So we're going to talk a lot about robots today. Exactly. We will. Yeah. And robots in space. We're doing our stories of the week first. Yes. What's yours Thijs? Well I love the headline that CNET gave to it. Hubble reveals soul wrenching view of the distant universe. Soul wrenching. Oh my soul is so wrenched. Yes my soul is completely wrenched. Because it's a new deep fields image by Hubble. In the 90s they already released the deep field image. Which was basically like they already made this beautiful picture. When Hubble finally worked right. The Hubble space. Of this supposedly black piece of the sky. Yes. Let's point it at something where there's nothing. Yeah. And then to everyone's amazement. You just basically saw just a plethora of galaxies. Galaxies. All those. From billions of years ago. Faint. Yeah. Not dots but. Actual galaxies. You can see the entire islands of universes just from 13 billion years ago. Thin patches of light. Yeah. Sometimes vague. Sometimes actually very recognizable as an actual galaxy. This. But sometimes. Yeah. But sometimes. Just right after. Visible light was even present in the universe. It was invented. Yeah. Yeah. It was invented. So it's just mind blowing. And after that they released another one. I believe it was the ultra deep fields. If I'm not mistaken. And now there's a new one. And they call it. I believe the. Yeah. The Hubble legacy field. Hubble legacy field. So basically what it is. Is it's a combination of all these pictures that they took over the years. And they just assembled them into something that's. How many? 265,000 galaxies. Reaching back about. Across 13.3 billion years. And a total of 7500 exposures. So it's. It's insane. And. It is sort of soul wrenching. Especially if you download the picture on. And high quality. It'll be memory wrenching too. Yeah. Exactly. Memory wrench. So funny because I tried to open it up here on like an old Chromebook. And it just crashed. The picture was too big. It says. Sorry. I ran completely out of memory. So I had to reboot the thing. But if you open it up. On a. Just a little bit better computer. Yeah. You can just endlessly sort of scroll through these old unit. Galaxies. And it's. Mega pixels or gigabytes is the picture. Well, since the whole thing just crashed right before we started the show. I can't say, but I believe it was like a gig. More than a gigabyte for just a picture. Beautiful. Yeah. Yeah. Insane. And love it. I mean, everybody's a fan of Hubble. I hope it will keep on taking pictures. Yeah. In the. In the future. I'm not sure. But I love that they've been doing these sort of like grand projects over the past few years where sometimes now these pictures come out that are more of like an assembly of a lot of work instead of just say one picture of a beautiful nebula or something. So it's less pictures nowadays, it seems, or less publicly known pictures. But when something comes out, it's mind blowing. Okay. That's my story of the week and the past 14 billion years. Chris, what's yours? Yeah. Well. My. My. My own story of the week or my team's story of the week. That counts as well. At the moment two of the scientists of the team are in Berlin with the PT scientists. They are building a moon lander. And we are doing a design review because we want to have the camera and the radio system of the moon rover on that lander, which means that we as the Netherlands would have something on the moon somewhere early next year. National pride involved here. Yeah. Yeah. After Israel. Yeah. And then the last week or the week before. China? That was also. No, no, no. Israel. Well, Israel for one, but there was another planning to go to the moon. Was it India? Was it some other country? Well, we've been talking about America. I'm sorry. India. That's correct. Yeah. India, China, America are all planning to go to the moon. Okay. All right. Yeah. But the Netherlands is lining up as well. Yeah. We are lining up as well. So it is a small thing and we are good at small things. So. We are a small country. We are a small country. We are a small country. We have a small country. We have a small country. We have a small country. We have a small country. We always did it with small things. Yeah. We always did it with small things. We beat the Spanish with small ships. We did. So small is big. Okay. Great. Small is big. And so. And so. Actually, we wrote a document that was called Be Big By Staying Small. So that's. Okay. That's good. That sounds very Dutch to me. Yeah. Yeah. We talk. Think small, do big, I guess. Next week when I won't be there, I'll be on holiday. Next week we're talking smallsats. Yeah. And the small thing is right here on the desk. You brought it. Yes. It is. Yeah. Maybe a small shoe box and it has six legs. And this is not a model. This is not a model. This is. This is actually a moon rover that is built. It's not completely finished yet, but what's in there would be fit for the moon. It's basically almost like an iPad with some wheels on the side. Yeah. It's a historic moment. We have a freaking spacecraft. Yeah. We have a spacecraft in the studio. Yes. Okay. One that might make it to the moon. Nonetheless. Sorry, Chris. Tell us about it. It's great. Thank you. Yeah. Yeah. So it is a family member of the Zebro robots we have. And that's a Dutch acronym for six legged robot. And the biggest feature of robots like this is that they are, you could say unstoppable. So you put them in. I love that. And maybe they're not as efficient in space. They're not as efficient in speed or in power consumption. But since especially on the moon, but also other harsh environments, you're on your own. There's no operator to help you. You need a robot that's able to climb out itself. I'm sure it's to do with these half wheels. It's got funny wheel legs. Yeah. We call them legs. C-shaped legs. C-shaped legs. Yeah. I mean, I don't know what to compare them to, but they almost look like hooks. Like they're half wheels. They're half wheels. Yeah. They're half wheels. And then they put some stuff on it to get grip. Can you set it in motion or is that asking too much? Well, then we would need to spend some time. I have to start it up, get contact with it and so on. That's too bad. Can I touch it? Yes, you can touch it. And I can show you the iPad like thing you see is actually the solar panel that it can put upright. And now you can put it upright and there's a little box on the net. Like a big hard drive almost. And then you can clearly see two cameras in front. Yeah. Oh. Not yet. Yes, sir. Yeah. Oh, over there. Yeah. That's too little. That's too little. So, it has stereo vision. Oh, I have to take a picture of this. So, it is able to have 3D vision, see what is ahead. So, it can avoid things that it cannot climb over. So, basically from Earth, it would get the command to go there or travel a certain distance. And then you just get the message, okay. And then after some time, you get the message, I did it. So, it is not a robot like you normally see in space that for every detail, every step, everybody is looking at where are the rocks, what are we doing, how many centimeters are we going here. Yeah. Like Opportunity Rover and stuff that gets steered from Earth. Every millimeter. Yeah. Actually, it's very frustrating for me that you can have a rover so long on Mars and then if you look at the distance they travel, it's hardly nothing. It's nothing. Yeah. And with robots like this, it's only one of them. Yeah. It's only one and a half kilo. We want more of them hanging around. So, a swarm. So, they spread around. How many would you send up? Yeah. Well, then the discussion is what is a swarm. But for some experiments we want to do, five would do. Okay. We want to do radio astronomy with them. Five would give a picture when they are at the far side of the moon and create a low far like radio telescope. We call it loo far. Yeah. But if you have 50, you have a better picture. And if you talk to radio astronomers to say, well, we would like to have a thousand. Yeah. If you have 5,000, it becomes a real swarm. So, you intend to build a telescope out of robots. Yes. Is that right? Yes. Wow. Okay. This is all going... I believe we're still in our story of the week segment. Yeah. We're getting ahead of ourselves. What was happening in Berlin with your colleagues? Yeah. What's the story of the... What's the news actually? Well, we are going to send this rover to the moon. We want to do that. When? We hope somewhere between 2021 and 2023. Okay. So, we're going to send it to the moon. Yeah. And then, we're going to send it to the moon in 2023. And that's a fact. That's a certainty. Well, there's always the finances, so we have to get a budget together. Yeah. Oh, okay. But you can say that the Netherlands has never been closer to a moon landing than at the moment. Nice. So, that's the story of the week, that the Netherlands has never been closer to the moon than this week. Than this week. Than today. And with this project. But because when you lose your radio, you lose your satellite, you lose your robot. so first we want to verify the radio system and also the camera it's also done so we want to take a picture send it back and actually we want to send it back to the radio telescope in dwingelo the old dish because that will be the ground station of the system you're kidding yes we so we will have our own sparks yeah in dwingelo and that's a famous old dutch telescope that's why i'm saying you're kidding it's it's i guess the most well-known decades old yeah yeah yeah it's it's old but it is alive and kicking it's doing a lot of a lot of interesting things yeah because they would send it these robots would send a signal straight to dwingelo yeah just at the right moment whenever you can see it so we are not making use of the lander so you could say the lander throws us on the on the moon surface and then we are on our own yeah does it come down a little parachute or there's no atmosphere so who's paying for this at the moment it is the universities all right and there is some some money from from from projects from and the old companies are helping a lot of the help of companies is in kind hey we have to remember that they are small companies so usually when you have projects then companies have to pay cash in the project to show their interest but of course small companies don't have that cash so that is a bit of a problem in in the space world there are a lot of small companies and you want to do a big a big project and you write a proposal save for five million euros and then you have to get 500 000 euros in cash from the companies and frequently we have had to withdraw a proposal because the companies couldn't bring the cash together all right But in this project they give a lot of help you can say on websites like this are obsession-driven yeah so so there are people of different companies that spent nights weekends and so going mind at a concert or whatever but there is hope el we all have to strive and we will get to the end of the to get this done, get this working, just to get the thing on the moon. Okay. We'll interrupt this story just for a short while to finish our stories of the week, because I have a couple. Oh, yeah. I remember. You also have a story of the week. Yeah. Two, in fact. There's one about... I'll give you the headline that I have to scroll back to. Oops. Let me see. Okay. On the verge. Don't panic. NASA is just simulating a potentially catastrophic asteroid impact this week. Oh. It turns out NASA and ESA together... This week was actually... Last week was shortly after we recorded our last podcast. Mm-hmm. And they actually simulated an asteroid impact with a rather complicated scenario. Some asteroid shows up and turns out to be on a collision course with Earth. I saw some headlines. I thought there was actually something going on. Yeah. You thought it was really... I didn't really read it. Yeah. I was like, oh, I believe something with an asteroid close. Like, oh, it's usually... Another of those asteroids. Yeah, yeah. Usually it's a false alarm. Yeah, exactly. All right. But the scenario was some asteroid shows up. Everybody starts frantically thinking, well, what can we do? They send out some nuclear device, and it's... It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. It's a nuclear device. Asteroid gets sort of deflected, but one big chunk isn't. And it lands somewhere on or close to Manhattan. Oh, like that movie Deep Impact. Yeah. Yeah. It's very similar. Okay. So we'll put this in the show notes and everybody can read up for themselves because it's a rather complicated story. Well, the idea is... Was there a happy end? Because Deep Impact... No, not at all. No. Deep Impact, I believe, it's Elijah Wood or something that then flees into the hills because I think they first nuked Asteroid and a small piece still hits. Armageddon movie, Clint Eastwood. Yeah, yeah, yeah. Lots of stories like that. I took one of my first dates ever to Deep Impact. That's why I... Was that a good choice? It was a good choice. It didn't really become anything, but at least I dared to ask a girl out. That was a big accomplishment. That's a big positive. Yeah. But this... This Asteroid Impact... No, it's not a happy ending. Oh. Of course not. I mean, if an asteroid hits Earth... But do we live? Like, will there be some survivors or does... Not in this scenario. But the idea, of course, is from the space agencies to persuade governments to give more money so they can become better at deflecting this kind of stuff and discovering them in time, of course. So in this case, a happy end would be more money, more money for asteroid defense. That's right. Yeah. And that's where I get a little... Skeptic? Skeptic. That's the right word. Thank you very much. Because if they're looking for more money, is the problem that they're sketching very real? It could be. Yeah. You wonder if they have bad science. Well, this kind of thing became in fashion, came into fashion. When the Star Wars project of Ronald Reagan ended up leading to nothing. You mean after the Cold War? Yeah. There was this Star Wars project, Ronald Reagan... Early 80s. Yeah. Had the illusion that we could shoot down nuclear devices using laser beams. Yeah. That was all bullshit. But it employed a lot of scientists for a lot of time. Yeah. And when they decided... It was all a bad idea. Suddenly, astronomers and space buffs came up with the idea, well, we might have to deflect some asteroids. And that's when all these movies came out. That's in the 90s. Yeah. Yeah. And skeptics then thought, well, maybe the idea was just to keep the budgets and invent a story that will keep the money flowing. Mm-hmm. So, later on, I got the impression... that maybe there was something to the story, but I'm still not completely sure because I keep remembering that the idea in the first place was to keep the Star Wars money flowing. Yeah. Okay, so... I believe Neil deGrasse Tyson also is constantly hammering on it because he knows that when it comes to defense, then it seems like you have a point and then it seems you get more budgets. Well, the problem with this kind of event is the chances of it happening are very, very small. Mm-hmm. And the consequences... The consequences are very, very big. Yeah, yeah. So, how do you... And they're both very uncertain. Chris, do you have any robots... So, how do you do the calculation? ...that can counter an asteroid? No, but I can imagine a world with electric life that will survive it, maybe because it isn't at the Earth anyway. Digital life. Digital life. Yeah, I call it electric life. Yeah, you... Okay. Electroloid. Electroids. Electroids, that's how you call them. Yeah, electric animals, yeah. Electric animals. Can you just... Briefly, tell me what electric animals are. Well, they are a result of evolution. So, if you look for evolution from species, you get more species, you get specialization and so on. Natural selection. And now we get what I call evolution 4.0, which is because of us, evolution has started thinking. Mm-hmm. And because of the thinking, a new species just is born and it is the electric animals. So, what you are looking here... You're at the table, a little robot is, you could say, a little electric insect. That's how it starts. And gradually, it will evolve to more things. And sometimes, I think that you can look in time, say, back 65 million years, and then you see the dinosaurs. Well, we can look 65 million years the other direction. And then probably, there's electric life there that is looking back at this place. And then says... Biological life. Yeah, yeah. They say, well, yeah, in those days, there was some carbon. Yeah. They did some carbon-based life. They did something. Very primitive. And here we are. And they have already become extinct. And then in the museum, they have some fossils. And there you see a person with a smartphone fossilized in stone. Mm-hmm. And then they pour into the smartphone. They said, that's us. Oh, yeah. And we needed a host to survive those days. And we don't need them anymore. So, that's... Oh, wow. And these electric animals, they eat electricity? Yeah, they eat sunlight. They eat sunlight. But they poop. And they... No, I think... The interesting thing, they eat sunlight. And they don't need air to breathe. Okay. Which means it makes it by excellence, species that go to space. And how do they multiply? Oh, they can do that in factories. They can build factories. Okay, yeah. Now, if you go to a factory that makes cars or a factory that makes chips, then you can imagine, yeah, give us, say, a few decades. And then the factory doesn't need us anymore. The machine builds the machine. Fair enough, yeah. Yeah. And, of course, we always think with reproduction about, yeah, having sex. Having sex with each other and so on. But there are quite some animals that don't do that. Uh-huh. There are jellyfish that make a factory. They grow a little plant-like thing with branches. And from the branches, little jellyfish grow. What? So, even building factories is normal by nature. And then, of course, the interesting thing is that, yeah, species like this, they can go to other places. So, when you have a swarm on the moon, it can just stay there, live there. Yeah. As long as it has resources. Sunlight and it can reproduce if it has materials to reproduce. Yeah. And, of course, you can have scavengers. So, they can eat the remains, take a leg off the robot, put it in the grinder, 3D print a new one. So, of course, they are also doing sustainable things. And what is one of the most interesting things about these species is the way they think, the way they organize their brains. If you look at... If you look at ants and humans, then we have about the same amount of neurons in the world. Only the ants cluster them in units of 250,000 and we cluster them in units of 100 billion. And in between, there's a slow interface. For the ants, it's a bit of sound and fear amounts. And for us, it's also sound and vision. So, the interface between my brain and your brain is slow. Yeah. Even if there is fast internet in between, but the slow part is talking and I see it. It's a vibration through air. Yeah. So, this means that the collective intelligence, of an ant makes that they can build a hive, something like that. Well, we have a bit more smart units, so we can build, say, Manhattan. But actually, the collective intelligence is quite low. It's disappointing. If you see what people do, and I think a lot about the intelligence of people, actually, I'm getting more and more disappointed. But then if you look at the robots, they can connect all neurons with high speed. So, imagine that all neurons, all human brains could be connected in one single brain. So, you get a super intelligence. Yeah. So, you mean also they have radio communication with one another. Yeah, radio communication. So, the individual units, they exchanged thoughts with the speed of light. Yeah. Witches used to be able to do that in the Middle Ages, I think. What? Yeah. I don't know. Telepathy is just, I don't know who was able to do it. Game of Thrones. I don't know. I don't know who was able to do it. I don't know. It's a big question. Yeah. But that means that if you look at the swarm robot, and you put a lot of swarms together, it means that you scale up the intelligence, and you scale up the memory, so you scale up the processing power. Yeah. And that means that when you would look in evolution, that for some reason more intelligence would be the end goal, then there is no reason for robots to evolve in units with big brains. I see. So, we are talking about, so we are totally focused on humanoid robots, but actually from evolutionary principle, it makes no sense. You want to build ants. Yeah. And the ants will form a swarm, and the swarm will be the intelligence. Distributed intelligence. And it is much more robust. If you lose some of your robots, you stay alive. If you need more thinking power, you put more over there. So, the survivability of a swarm with this fast coupling of processing, I think it's the brilliant next step in evolution. It's a bit fascist, but... Yeah, but you see it already happening now in the world of computing. So, the only thing is that computers connect. We create things already that are bigger than we can imagine. We see all kinds of interesting or funny or spurious correlations in the huge amounts of data that we have. The only thing is now it doesn't have legs and hands, that system, but that will be happening. And we see already this indicated by things like Internet of Things. People say, yeah. But I always had a problem with Internet of Things because I don't know what a thing is. So, I built machines and so on. Anything that can have Internet. Yeah, and then people say, yeah, build me a thing. Maybe you know the movie of Ace Ventura, the Pet Detective. Yeah, Jim Carrey. At some point, he wants to impress people. And then he says, when you don't listen to me, I will turn into a hideous, vicious, dangerous thing. Because he cannot come to the ward. And I always think about that. But of course, if you follow the reasoning I just did, then you can imagine that we are not talking about Internet of Things, but about Internet of Robots. So, we should stop talking about Internet of Things because that's dead end. We have to think about Internet of Robots. The robots can be tiny, can be like little insects, can be drones, can be little beetle-like robots. We have them the size of a matchbox. They can crawl around. They could go to the places where you want to sense. If you say, well, I have a thousand sensors in this building, who is going to put the sensors there? Who is going to clean them? Who is going to clean them up when you don't need them anymore? What you want is you empty a box of sensors in the entrance and you wait for two days and then they're everywhere. So, finally, we're talking swarm robots. Yeah, Chris, I wanted to stop you also. Herbert, I was just eyeing you. Do you still have another story of the week? Yeah, but I'm skipping it. Okay, great. I'm not going to interrupt Chris anymore. Any further. Except for to structure the conversation a little bit. Exactly, yeah. Because before we go on, I need a definition of swarm robotics. What is it? Yeah, swarm robotics is working with swarm robots. And a swarm robot is a really appealing, really autonomous robot. So, although it is in a swarm, in principle, it's on its own. So, it's able to survive on itself, to get its food on itself. But it is aware of its neighbors. It can be the distance, can be the direction, things like that. And based on this awareness, it can adapt its own behavior. Yeah, they can cooperate somehow. Yeah, so there's no coordination to the group of robots that you have. But the robots might have a similar purpose. And we feed them that purpose, right? If you have a swarm, then we can still, or something gives them a purpose. I hope it's us. I hope we can. Yeah, and the purpose could be, for example, find fire. Well, and then you put the robots... Find water, okay? Or water, whatever you want. And then fire to extinguish over there. And then you empty the box with robots that want to find water. So, they spread around in the building because they don't like to be too close to each other. But they also don't want to be too far away. So, it's like foam that you push in an object with funny dimensions. Then it will spread everywhere. So, it's like some kind of a pressure in the swarm that pushes them in all places. So, even a room that you forgot, robots will go there just because they are pushed in there by the other robots. And then they are hanging around there, minding their own business. And suddenly one finds water. So, he puts up a semaphore, I found water. Then the neighbors of the robot see a robot that sees water. So, they also put up the semaphore, but they give themselves a higher rank number. So, the robot that finds it says, I'm zero. So, they all say, we are one. And then next to it are robots that see the number ones and say, hey, I'm becoming number two. And this ripples through the building. And then at the entrance, a robot says, we found water and I'm number 30. Yeah. And then the only thing you have to do is walk from 30 to 29. So, they create a trace like ants. It's a pheromone analog. Yeah. And we say, so they create a potential field, we call it. And along the potential field, you find the shortest way to where you want to be. And the little matchbox size robots we have even have given them light emitting diodes on the back. So, they can create a running light. So, this means that they can create a running light just so follow the light. But it could also be that you have a festival in a park. So, you deploy the swarm and they are hanging around there. And suddenly, there's panic or something like that and the people have to go to the exit. Well, then they can form a potential field to the exit. So, you will get light traces to the exit. And people also turn into insects when they're panicked. So, they just follow the traces and you can do crowd control. And the interesting thing is that the robot itself knows nothing more than, oh, I'm number 30. Yeah. And I know nothing more than I became a three because I saw a two. Yeah. So, you can scale it beyond limits. Yeah. That's beautiful. This is really beautiful. It's an amazing thought. It's just that this looks quite far into the future. Well, actually, there's a spin-off company that is already busy trying to put this to the market. Because what could you do with this form of… I've seen drones. Like I've seen drones as an art piece. Are there any real… Yeah. Well, you have to be careful with that because… Yeah. You see these fireworks in the air. Yeah. You see a lot of drones, but that is not a swarm. A swarm cannot do that. A swarm cannot create the Olympic rings. Oh. Maybe they could fly in one ring if you give them the right things. Yeah. But you would more see something like the sparrows in Rome. So, those clouds. That's what they would do. What Thijs is describing is something in which every drone is steered from outside. Yeah. And where every drone… Every drone has a place where it is supposed to be. Yeah. And one of the essential things of a swarm is the swarm is somewhere, but you don't care where the individual robots are. Yeah. And they have almost an individual intelligence on knowing where to go. Yeah, because… It's very small and it's just the real intelligence emerges from the swarm itself. Yeah. Yeah. So, the property that in a complex building like this building, they can find in a hidden room, they can find water and lead you to it. It's very intelligent behavior. It's very intelligent behavior. Yeah. But they don't need the navigation system for it and they don't need the intelligence for it. So, are there any real life examples now of swarm robots or is it just a scientific prototype thing? Yeah. Well, you could say we are that far that at the moment we have robots that can be swarm robots. And with that, I mean they can live longer reliably so they don't crash, they don't catch fire and so on. They are safe for children, things like that. And we want to have one of those swarms alive in the garden of the science center of the TU Delft, which will be like a little children's farm and you can just go in there and then those robots are hanging around there and you can play with them. We plan to give them a camera so they can recognize these things so you can… You can feed them sunlight. Yeah. Yeah. They live with the sunlight. Yeah. And they are there. Yeah. But… How do you do maintenance for a swarm? Like one fails, you just add a new one and done? Well, actually the idea is that we do not do maintenance. Okay. So, the robot stays healthy for its economic life and then it dies. And then like an ant, there's no ant hospital, there are no ant nurses. When there's something wrong with you as an ant, you die. Yeah. Ant is disposable. Yeah. And it's the same for the swarm robot. Forehand. So, this means, for example, you have those matchbox-like robots in this building and then one dies. Well, then the cleaner comes and just swipes it up and puts it in the garbage. Or the robots do that themselves. Or… And actually, this was a suggestion by children. I talked to schoolchildren about this and they said, so if these are animals and you have an electric ecosystem, there also must be scavengers and predators. Yeah. Actually… Yeah, there will be. If there's a limping zebra… Smart kid. Yeah, let it be eaten. And of course, we already had examples where one of our flying drones was attacked by a crow. So, also animals see them as other animals that you can hunt. Yeah, hunt. They won't be able to eat them, but you know… Yeah, they will discover that. Well, and you already said that space is like the best application for this, for swarm robotics. Yes, because if you… Why? If you look at, say, the adaptation of animals to their… They call it umwelt, so the environment. Environment, yeah. And then you see that the electric animals, they are perfectly fit for a vacuum atmosphere when you have light and you don't need air and so on. So, actually, you could say the moon would be, in the end, a better place to be for the robots than Earth. Mars would be a better place because there are no plants where you can get stuck and so they could live there. So, if you say… And free roaming in space itself for exploration? Well, you could say we are thinking about having a nano-satellite swarm hanging around the moon to actually also form a radio telescope, Olphar. So, you could also dare the swarm ideas there and that has the similar drive as this. It's very small, not so much resources, so you cannot make it a very… And I can imagine many advantages like it being sturdy, you have a less chance of… It's redundant. The parts can fail, but the whole can't fail, is that right? The swarm is very robust, it's redundant, it is self-repairing. Yeah. So, if you miss a sensor node, then another sensor node… But are there any disadvantages compared to one big satellite telescope, for instance? Well, it depends on what you want to see. Sometimes there's an instrument. If you want to detect X-rays, it's a big thing. So, it would fit in a small satellite. Okay. So, sometimes you need something… Okay. …something big. But if you want, for example, to catch a field and then maybe you want something bigger, you were talking in the beginning with your story about looking deep space… Yeah. …when the stars were ignited. But of course, between the Big Bang and the ignition of the stars, also things happened. And we can't see them at the moment, but there are signals coming from there. Intensely redshifted hydrogen lines, we expect. And there was one satellite, Radio Astronomy Explorer 2 of NASA in the 70s that… …has seen some of those signals. Really? …could have been signals from an exoplanet, could be the structure of the early universe. And we want a radio telescope that could actually observe these signals, but these are very low frequencies. So, at the surface of the Earth, you cannot receive them. The atmosphere is opaque. Okay. And the Earth itself generates too much disturbance. This experiment of the United States showed that only at the far side of the Moon you could receive those signals. So, the far side of the Moon, you could receive those signals. So, the far side of the Moon is very interesting for radio astronomy to really look deep space beyond the ignition of the stars. So, see the early universe. And then you need another big thing. Then you need a big telescope, say a telescope like LOFAR. Well, that covers half Western Europe. That we talked about two weeks ago, I think, because I happened to visit it. Yeah. Yeah, in the Northern Netherlands. So, LOFAR on the Moon would be heaven. Yeah. But it… Very low frequency radio telescope. Yeah. But it is… It always looked like science fiction to have LOFAR at the far side of the Moon. That was called LUFAR. It was one of the… It is one of the holy grails of people like Heino Falke. They say, have a radio telescope. The black hole picture guy. Yeah, exactly. Exactly. And one of the things you can do is have satellites in Moon orbit. And we call that radio telescope OLFAR, an anagram of LOFAR, the orbiting low frequency array. When the satellites are at the far side of the Moon, they receive the radio signals. And when they see the Earth, they relay it back. Yeah. The correlation and the interferometrics must be done on the satellites itself. Otherwise, there's too much information going down. But I already told, they connect with each other. You have distributed processing and so on. So, they can do something there. But there is differential speed between the satellites. So, maybe they will not be able to integrate long enough to see the signals. But if you put an LOFAR antenna on the back of these little robots… And you tell them, walk to the far side of the Moon, which means you don't see the Earth. That's easy. And the other one is spread like you would always spread as a swarm. So, if you have enough of them, then they spread over an area of maybe 100 kilometers. But how do they do propulsion? Well… Just walking. Hook like legs. Oh, just walking. Okay. So, they're all walking. So, it takes some time. So, you deploy them. And maybe you have to wait a year before the radio telescope really becomes saleable. Yeah. And maybe 10 or so will disappear on the way. Some will die. How many would you need for this telescope, you think? Well, with five or six, you could make a picture that could give you the idea that you should make a better picture. Yeah. So, like proof of concept. With 50, you really get something clear. Yeah. And the astronomers… I'm also an astronomer. So, I don't know that… I hear them talk about a thousand or something like that. Yeah. Okay. Yeah. So, then cost becomes interesting. And then you have to… And then you have to… And then you have to… And then you have to… And then you have to… And then you have to… And then you have to… And then you have to… I think they talk about a thousand or something like that. Yeah. So, then cost becomes interesting because… Well, one big satellite costs a lot of money. Tens or hundreds of small robots would maybe also cost a lot of money. But I can do the math. Can you? Well, it's all about mass. So, this robot is only one and a half kilo. Yup. A bigger rover is maybe, if it is small, 60 kilos. So, you can… calculate how many of these you can have for one big rover. So for the mass of Curiosity you could have quite a big swarm over there. And it is interesting to have a swarm there even for rovers like Curiosity because the swarm can go ahead and see all the dangerous things and then the big rover just has to follow the trace of the swarm robots so then the big rover can also speed up a bit. Now you're talking like mass is all that matters but building a big satellite, it's a complex task force, is that much more expensive than building turning the same mass into 50 little robots? That is because you're looking at other technology. So sometimes I say you have big space and small space. Big space is the big satellites we build, the complex things. Things like New Horizons and so on and that's totally different species. But here you're looking at smartphone technology, smartphone and automotive technology. So actually it is the smartphone and automotive technology that goes into space. So that is the world of short times to market rapid prototyping and of course an immense amount of research to get things okay. I bet that on the gyroscope in this rover more money is spent than what ESA and NASA could afford together to spend on their sensors. Because if this is a gyroscope in a smartphone and you put a smartphone on the table and the smartphone says I'm slowly rotating, you will start a Facebook page that don't buy this phone and the designer should die and whatever. So this means that the manufacturer of that gyroscope with a market of billions can spend a thousand engineers and hundreds of millions in getting that gyroscope better. Yeah. And it's easier to fail. I mean not so much that it will fail quicker but it's like if something like this fails, it's not that big of a deal whereas a billions costing rover fails then all hope is lost. So the only thing you have to take care of is that there is not a common cost that you put a thousand robots at the moon and then they all fail because their battery cannot deal with vacuum. So that's also the risk of swarms. For example, you have 999 robots in the swarm and it works fine and the emergent behavior is fine and the robot number thousand comes and then suddenly it locks up. But if I understand you correctly, then it's basically the micro technology, just smartphones and just everything becoming smaller that makes it possible for these swarms to actually exist in the first place. Yeah, so there is an enormous amount of experience in building things like this, which is not space based. So there are many companies in the world that have become space companies without knowing it because the technology can go to space. So that's one thing. Any examples? What sort of companies are we talking about? Smartphone manufacturers, for example. Of course, the display has a problem with the vacuum, but for the rest, smartphones have been sent into space and work. And our satellite, another satellite, Delta C3, you can just order it from Farnell. And it is in space now for 11 years. April 28th, it was the 11th birthday. And every day I get three emails of the satellite that it is still alive. So to go back to my question, can you build like 100 small robots for the price of one big satellite? In all probability, you can. Yeah, well, my guess would be if you look at the technology and its mass production, and you are parasiting on smartphone and automotive industry, it will be less expensive. Yeah. Which brings me to this question of deployment. How would you deploy these? Because it might sound easy, like, oh yeah, you just put them up in the space, but if you want to put something small like this on the moon, how would you go about doing that? You need a cargo lander. Yeah, you need a cargo lander, and you deploy them all. Yeah, you need a stupid box that can land, and then it opens up, and then the robots walk out. And so it will, with a rocket or whatever, the box just dies. It would be something like the sky crane that puts Curiosity on Mars. It puts a box on Mars. It opens up. Yeah. We would need some sort of standard deployment technique for that. How about this? If you make them small enough, you can just throw them. A mouse, if you throw a mouse out of a plane, it survives. Why? Because it's small. Help me out. If you throw an elephant, it explodes on impact. The mouse doesn't. Why? And if you think a mouse does, imagine a small animal. I didn't study physics. Help me out. Has to do with the surface area of the mouse decreasing compared to the elephant with a second power, and the mass which defines the impact decreases by the third power. Will it bounce? Will the mouse bounce? I suppose it will, yeah. Help me out, Chris. You're right. I am. So you can shoot them up and basically let them... There's no atmosphere on the moon, so you can glide. Normally when you ask me a question like this, I try to find a team of students and they do the calculations. But I had a team of students that looked at the way Curiosity was put on Mars and they changed Curiosity for a container. That takes Zebras to Mars and that was a very viable scenario. That was just a one-time deployment though. Curiosity was the sky crane, right? There was a sky crane. The sky crane put down Curiosity. It went down and then this crane lowered it and then the crane itself flew off and died. I think it was a one-time thing. So you need some sort of standardization I would guess for... So if anybody listening wants a field, deployment of swarms, it itself can be a little niche. For example, I forgot when it was, not so long ago, more than 100 nanosatellites were deployed in one launch. That's true. SpaceX did that, I believe. No, ISRO, Indian Space Research Organization did that. And the launch broker was Isis Space. So it was actually a Dutch achievement. Isis Space in Space Cowboys podcast next week. Yeah, next week. Definitely. So we'll ask them. Let me write that down. Yeah, we know each other very well because they were my former students. Oh yeah? Great. Good. Good for you. But from that you see that the same deployer would also work in moon orbit. So if we would stuff the deployer with all of our satellites for the radio telescope and bring the deployer in moon orbit, you deploy all of them. So a radio telescope in moon orbit is not science fiction. Can be done. Hey, to what extent is like the GPS network, navigation network, a swarm? No, that's not a swarm. It's not a swarm? No, that's because that's a... I told somebody today if I open up my navigation app on my smartphone, I don't care how many satellites I see. I don't care where they are. I don't care which individual satellite I see. No, you don't care. But for the navigation solution, it's important. Your receiver knows which satellite it's listening to. And it is also where the satellite is supposed to be. And from that it can calculate where you are. All right. It's more fuel of plants. They're not a swarm. They're more plants. They move, but we find them and they don't find us. It's like lighthouses at sea. So they move, but you know exactly where they are. And they also take... Floating lighthouses. Yeah, something like that. And everybody takes care to know exactly where they are. And so there are formations. And they are also very interesting. And you can also have formations of robots and drones, but they're not swarms. Okay. Now, swarm anything, let alone swarm robotics, hasn't been viable until about now. Is that right? Yeah. At the moment you see it happening everywhere. Okay. So it's picking up steam. But what I want to know is, why now? Well, that has to do with processing power. So intelligence, that's one thing. It has to do with communication. That is at high speed possible. And it has to do with batteries. Okay. So the first drones that... So for example, there were a lot of fuel-propelled RC airplanes. And at some point you saw them disappear and everybody started flying electric. And that's because of the batteries. So the batteries... More power per unit of work. Per kilo. Weight, yeah. Or lithium-ion? Lithium polymer. And then the other thing is the MEMS, so the micro-electronical mechanical systems. So these are the gyroscopes in your phone. So normally a gyroscope was a big mass rotating, very expensive and so on. And now for 80 cents you buy a gyroscope. And when those gyroscopes were put on the drones, the drones could actually fly with those gyroscopes. Yeah, they fly themselves basically. So nowadays with your joystick, you tell the drone where to go. But actually it's doing a lot of things itself. Yeah, it just... And that is what I always say, animals. So the animals are autonomous. And with an animal, you can command the animal, you don't control the animal. Which is fairly normal. When you sit on a horse, you command the horse, you don't control it. You don't drill a hole in the head of the horse. No, no joystick there. Put connectors. A joystick to move his muscles. It is so in the same way, the cars we have now are silly things. We drilled holes in the brains of the cars and with the steering wheel and the pedals, we are controlling the muscles. So an autonomous car is logic. It is normal. And if you look in history, then you see we always had autonomy in our environment. Cats, dogs, shepherd dogs. Can you imagine that the shepherd would actually control, the dogs with a joystick where they have to go. Now the dog understands what it is supposed to do. And with this understanding, it listens to the shepherd. Where should sheep go? And then the dog knows. So we want that. We have always wanted that. When you sit on a horse and you're drunk, you get home safely. That's what I want. I want to sit in my car, be drunk and get home safely. So my car should be a horse. And then I always think about this anecdote about Henry Ford that he said, well, Henry Ford, he was a horse. He said, well, I was building a car. And I asked people, what do you want? And he would say, I want a better horse. And he thought it was silly, but actually people wanted a better horse. And I still can't imagine how come that when he was thinking about the better horse, he was thinking about faster and stronger. And he forgot about autonomy. Henry Ford should immediately also have thought about autonomy. Ford could have started with working on autonomy, say right at the emergence of cars. So why did he wait so long? But there was no technology like that anywhere near at the time. No, but autonomy, people had it. You mean thinking about it would have already helped us? Yeah, if you would talk to an Egyptian and you would say, well, I have an autonomous truck. He would say, me too. There's a horse with a cart. And Chris, I would assume that you were saying this before Elon Musk had a press conference about this a week or two ago. Oh, I've been telling this for years. So you would agree with him. He said, well, it's actually really weird to think about cars and not think about autonomy. Yeah, but it's about time. So you would say he's 100 years late. Yeah. He should have thought of this before. And I also am convinced that in 30 years, we have forgotten about non-autonomous cars. And in 300 years at school, people will say, well, we had horses and then we replaced the horses by autonomous cars. And then one of these kids will say, yeah, but Mr. Teacher, we had one century and they didn't have autonomous cars. Yeah, that was a very silly middle-aged thing. We don't do that anymore. Very short and immediate. Yeah, exactly. And it's so funny because I have to think about certain science fiction movies, whether it's Star Wars or Independence Day, where a swarm of the enemy comes out of the mothership and you see them swarming. But they all have a little individual, a biological individual with a sort of steering wheel, do-do-do-do-do, do-do-do-do-do, shooting each other. But actually, it's really weird. You would think that whoever, you know, you're going to be a little kid, you're going to be a little kid, you're going to be a little kid, you're going to be a little kid, you're going to be a little kid, you're going to be a little kid, whoever writes the better algorithm would win instead of who has individual biological skill. Thinking a lot about this, because in the beginning, I told you about the super brain. So in principle, all those robots together, all those autonomous cars and all those drones with the Internet of Robots will be one creature. Oh, yeah. And then we're toast. And yeah, well, we call that, we call that, we even gave it a name, the robotic trinity. So big data, Internet of Things and robots together will form, God. This, this, yeah, sometimes I made, sometimes I made the joke that I said for the first time we can prove God exists. But we are creating something that goes beyond our imagination. And then suddenly the word autonomy also becomes very important again. Because if you have an autonomous car, then of course it is part of the community. But it also should have its own autonomy. Like, for example, it doesn't want to have collisions and so on. So suppose, somebody hacks the big system and says the minimum speed in town is 150 kilometers per hour, that the car itself says, but I'm not going to do that because it's dangerous. And maybe in the autonomy you could have a personal relation. You could have a personal relation with your car in the sense that you know the car and the car knows you. Of course, it is behavior in the cloud and you enter an autonomous car and it becomes your car. But the car knows you. So this means that you can have a trust relation with the car. So although the car is part of the big system, you can have a trust relation with the car. So although the car is part of the big system, you have a community in your partisan system. It's also like a firewall between you and the community. And that could help us a lot. Against something that I also have a big problem with the information overload. You get so much information that it is very difficult to select what am I going to read, what I'm going to see, and what not. You're human right now. And now at the moment we are helped by computers. They built a cocoon around you and say, oh, you like through green things, so at no time every picture you see from the Internet is green. And in no time you think that the world is green. is green and and then if somebody tells you well there are also red things uh but if you have your robot companion with you then the robot companion could tell you well you're looking at all kinds of green things but how about looking at something red and actually that that is also something that that already happened before uh a british guy once told me that the british aristocracy was was not so smart because they don't have to work and they have enough money so there's no reason to be smart but they had a smart butler oh yeah so if they were about to do something really stupid and the butler would say maybe you should do that and maybe we end up in the same situation again so you have your robot companion your friend you talk with it it sees all but it is your friend and loyal to you where can i get this robot soon working on it can you call me when you have one i really need one i'd like to get back to reality now oh sorry i need somebody that sent texts back to all my friends for all the dinner invites and my to-do list is endless and i and i i'm really bothered by the algorithms on youtube that keep on showing me the same types of videos and i would love to see more of the world in my inside my algorithm so i'm really looking for this robo buddy so call me yeah i think i think the robots will call you let the robot call me when it's there thank you i would like to know because we're talking a lot of science fiction here um what will be the first yeah um the first robot swarm that's really out there be it on the moon or well i think that i think the first robot swarm you will see will probably be a little primitive robots hanging around in for example the departure hall of schiphol with an electronic nose just sniffing for drugs and explosives okay fair enough and that is and that is i love the dogs and that is and that is so close that it might be in the coming years that you're going to see that all right uh and uh for somebody with drugs or explosives there's no escape for this form because the form is everywhere so they always see you see this sounds and then they create an effect and a potential field in your direction yeah so at some point you will see somebody in the departure hall with lead lines just all pointing to him so it's going to be a really interesting thing to see what happens when you're in amsterdam airport i'm just saying there's going to be a really busy swarm not explosives but the other thing yeah but there can be many of them yeah yeah you can you can put the whole hall it's gonna be funny yeah that's the other field of experience yeah exactly whenever you pass a swarm robot it lights up red oh yeah exactly yeah just just when finally the border police wasn't so strong strict with that anymore and uh then you will have the swarm is there anything in particular that still needs to be done in the future i think it's going to be a really interesting thing to see what happens when you're in the future i think it's going to be a really interesting thing i mean i've just seen a lot of all fates and in fact what we've seen here is it will change overall but this time you know we're not actually making directly issue with swarms so we're the first ones to видео less so this is going to be kind holding fan feels but who's going to beat this issue i'm just just talking about it uh plus the truth of course the risk factor of tele 배우 때 situation but the problem here is we want to be successful in each aspect of the showstopper so the technology we're looking for otherwise we don't set up space for tele vulkano seconds do you know because profitable teleечap points and but we have a very farming group that is actually showing tele-lovers in this sensibtly as well that consultation is going to be<|sk|> out ree कน PRO reporter here in its own right-hand work team that still but cloud showstoppers here but the showstoppers happen to be legislation and insurance fees oh oh yeah another interesting sideline uh near the end of the show tell us about it yeah it happens more often than not that you that you want to do something and then it happens that for example you cannot insure the thing can you give me an example yeah i heard about an uh a company that that puts concrete in those big deep holes in london and then you have those concrete pumps and those concrete pumps have piping and the pipings have joints and you want to know if they are leaking yeah and then two people in an in an in a little basket with a crane they go past past the pump line and they look for for this and then the company thought why not do it with a drone with a high resolution camera you just you fly past the pipes take a picture bleep so they did it and then they thought that the insurance at the building site would cover this but an insurance company said no this is not covered the only thing we can find actually that would fit is uh flying uh low above london with an aircraft well you can imagine what the insurance fee is there so you have to stop the project yeah yeah exactly so it's basically imagination or basics or existing laws or rules well this is a lack of structure i guess this is a category problem yeah but it's very difficult people have to imagine so what's the risk like for delphi c3 we pay an insurance fee of 8 000 euros for what what what what can it do what can happen so it's not the technology it's more the context in which is being used that still needs to be updated and then of course people get afraid and then just to make sure government will say well let's make it illegal yeah so for example now the with the drone legislation they say well you need to basically two people one looking at the drone the other doing the control yeah and then if you look at our reaction time and you know what the reaction time of a drone would be imagine that you are trying to chase birds on on a landing strip and because of a gust of wind your drone ends up in front of an aircraft that is taking off yeah well in the time that you see it happen the response time of your brain doing something with the joystick if you calculate the delay there then the aircraft did three four hundred meters oh yeah so it will hit the drone but if you look at the reaction time of the drone it's not the same as the reaction time of the drone but if you look at the reaction time of the drone it's not the same as the reaction time of the drone itself that would see it and that knows i'm not supposed to be there then it's gone in milliseconds so so it is a way before the aircraft can do anything and then the other real rule is uh it should not be above 50 meters but if there's a gust of wind that goes above the 50 meters please use it or as in my how birds do this yeah or as in my case two weeks ago if there's a tree please just know that there's a tree and don't fly the drone into a tree took me 100 euros to get it back out of the tree it's a real story um so i'm really waiting for this technology to come do present-day drones really have this kind of self-steering capability that they can on their own stay out of the way of a plane well we're working very hard on on the exact things but you but for example a drone could know when it is not supposed to be on the landing strip because it could look in the schedule so it just knows yeah that an aircraft is coming yeah so so even before the aircraft is there the drone knows it's going to be there so if you say to the drones autonomously chase away birds uh then knowing the wind direction and so on it knows where it's not supposed to be it it it could calculate what are the hazardous areas yeah so uh stay out of there when an airplane is coming so the system will be much more safe than with the person doing the joystick i want to just before we before we end for listeners yeah this is now cradling the moon lander i'm padding it also you can rotate the leg a bit then you can see uh this is the solar panel going up and down yep let's have the the legs rotate oh yeah so that's how it's going to sound when it's walking on the moon no there's no atmosphere on the moon you won't you won't hear anything no sorry still the body will vibrate this way oh still it's interesting to listen at contact sound yeah it will tell you something about the bearings i suppose yeah and so this would go around and herbert i think your your uh your question about like practicality and when can we actually see this is super valid you talked about skiphole um you talked about the maybe the garden at the university maybe having having something um but what's the time frame on all this because we're saying science fiction but some of these things are not so scientific it's not science fiction anymore and i'm and i'm and i'm holding one of these robots right now so yeah yeah and uh and uh suppose i get the the the the funding okay and probably i will and the law and the insurance then in three years time it's on the moon and it's performing in three years time we have a dutch presence on the moon yeah very small have you told have you told the prime minister or anyone here up up high he's an historian he looks back and with that well with that okay but let's let's just um stress this yeah uh because people in politics always tell us that we're behind in something and uh and we should uh make make uh we should get back with the crowd and make sure we're not behind anymore but we can also um be ahead yeah definitely yeah and in this case we have a chance to get ahead of the pack yeah by being small by being small by staying and we could tell the prime minister this yeah because he thinks very small usually so he doesn't think big our prime minister thinks think small okay so so maybe he's very interested but if you look at let's make sure we're ahead if you look at the amount of money needed and this is say doable also for the netherlands we couldn't do an apollo project in the netherlands but we can do a project like this yeah okay amazing thank you for the for the wonderful conversation sometimes it's just mind-blowing to see what's coming and this little cutie is coming does it have a name uh it's it's uh probably we're busy with it going to name uh belka belka you have belka strelka and laika the three dogs that went to space from the russians yeah and belka and strelka will stay here because belka is they say the engineering model strelka is the qualification model and laika in honor to the first dog in space uh will go to the moon beautiful beautiful yeah we are we at awesome thank you very much thank you very much chris chris verhoeven assistant professor at the university so sorry associate professor associate important distinction associate professor delft university of technology yeah thank you thank you guys thank you herbert and like uh like we said next week jeroen rotterville from isis space isis space uh dutch space company and you will not and you will not be here herbert you're gonna be on vacation yeah so see you in two weeks i'll make sure i'll listen yeah yeah great and Joeri is gonna be my co-host Joeri our in-house space nerd yeah will fill in for me awesome thank you okay thanks so much everybody see you next time thank you all bye