Bullaki Science Podcast

16. Swarm Robots for Moon Exploration and Holography | Dr. Peter J. Christopher

February 08, 2022 Bullaki Season 1 Episode 16
Bullaki Science Podcast
16. Swarm Robots for Moon Exploration and Holography | Dr. Peter J. Christopher
Show Notes Transcript

Dr. Peter J Christopher is a Research Fellow at the University of Cambridge working on embedded 3D holographic displays and additive manufacturing, and Chief Scientist at at Exobotics.

In this Bullaki Science Podcast we discuss the future of moon exploration by new mini rovers that could be deployed as a swarm and holographic 3D printing.

The video is available here: https://vimeo.com/manage/peterchristopher

Recorded: 03/02/2022
Published: 08/02/2022 

COPYRIGHT BULLAKI ltd

******
CONNECT: 

- Subscribe to our YouTube channel: https://www.youtube.com/bullaki
- Support on Patreon: https://www.patreon.com/bullaki 
- Spotify: https://open.spotify.com/show/1U2Tnvo1PZY4Fu4QLHURJV 
- Apple Podcast: https://podcasts.apple.com/gb/podcast/bullaki-science-podcast/id1538487175
- LinkedIn: https://www.linkedin.com/in/samuele-lilliu/ 
- Website: www.bullaki.com 
- Minds: https://www.minds.com/bullaki/
- Vimeo: https://vimeo.com/bullaki 
- Odysee: https://odysee.com/@bullaki:0

*****

Peter Christopher (PC): I get enjoyment out of simple things. Possibly the saddest moment in my life was when I once subscribed to Concrete Quarterly. Which feels like I should be on a "Have I got news for you". 

PG: Welcome to Electrical Engineering here at the University of Cambridge. I'm Peter Christopher. I'm a postdoctoral research fellow here. This is part of the Cambridge University's grand plan to move all the engineering provision out of the center.

PG: This is the device that made our research group famous, from one of Cambridge University's largest spinouts. This is a holographic projector in here as opposed to a classic on focus projector. So one of the features of a holographic projector is that it's got infinite depth of focus.

Samuele Lilliu (SL): Thank you very much, Peter, thank you for being here.

PG: Thank you for having me.

SL: I think we had a great trip in your lab and that was exciting.

PG: It was nice to be there.

SL: So today we will be talking about two things. Moon exploration and holograms. I’d like to mention one thing… that the two things don't come together because you might be aware of that conspiracy that the Moon is a hologram...

PG: It isn't. I wear two hats. I work at University of Cambridge, which is the holograms bit and I'm also chief scientist for a company called Exobotics that does satellite and Moon exploration.

SL: Moon exploration started with the Russians who managed to land the first probe on the Moon in 1959. That was when there was the space war. Then Kennedy gave a famous speech. He said we do things because they are complicated, but he was also doing things because of the space race. So they eventually managed to send a probe on the Moon in 1969, I believe. The program lasted for like three years, I think. So at the beginning of the '70s they stopped sending people on the Moon. 

PG: Yeah, so I think '69 was when they landed Neil Armstrong. Don't quote me on that. That's about right.

SL: Do you think it's real? 

PG: Yes. 

SL: Yeah, me too. Why do you think there’re all these conspiracies around?

PG: As someone trying to do what they did without computers or calculators, really, 50 years ago, and I'm trying to do it on significantly better kit, I realize how hard it is, I can understand why people think it's made up. It's not easy. I think it feels unrealistic. I don't know. I'm not an expert on conspiracy theories. But I think that people like to have an understanding of things that are beyond... 

SL: So now there are there new programs going on. The Chinese managed to send a probe to the dark side of the Moon for the first time in history with the Chang'e 4. That was in early 2019. Now we're expecting the Artemis [NASA] Program, which probably is planning to send humans to the Moon. Right?

PG: I'll have to check. Sorry, I'm not the Space expert, but I do the software. 

SL: Why do you think we should go to the Moon right now?

PG: Because we can, is the obvious answer, but I think it's also a testing platform. There's a difference about us going back this time as we're not going back to prove the point, because we can, we can do that. But in real terms, in today's money, the NASA program to send Neil Armstrong to the Moon cost a trillion or something...

SL: A trillion? Wow.

PG: I think, if you scale for inflation, I'd have to check the exact numbers, but it's a lot. That's obviously very expensive. The big difference is, over the years, we've moved towards commercial space. It used to be satellites were just made by governments with government level funding. Nowadays you can buy a satellite. It will be expensive, it will cost you millions, but it's not going to cost you billions, unless you want something huge. That change is pretty significant nowadays. There's an entire ecosystem, a commercial sector built around this. If you're a telecoms company, you can buy a telecom satellite. The idea is that could we do to go to the Moon for that? Now, the difficulty is there's not as necessarily as obvious commercial advantages in the Moon. Telecoms is an important reason why you'd want to be in orbit around the Earth. I think, if I'm honest, there's an element of exploration, there's an element of people doing it because they can, like the Elon Musks of this World, and then there's an element of people applying for very high risk, high reward of "Could we have a mine asteroid?". 

There are things you can do on the Moon that you can't do on Earth like this manufacturing techniques. [The Moon is] non tectonic. So even though it's not necessarily obvious to you and I, there are vibrations...

SL: You have Moonquakes [on the Moon], very little tiny Moonquakes... but compared to the Earth...

PG: And that's a major issue, if you want to go down from your five nanometer lithography process or something. One of the major issues is the vibration. Also, you can get a higher vacuum on the Moon than you can on Earth. But I think the thing people are really playing for is the slice of the trillion dollar pie that would be asteroid mining. The rare materials we use in all the devices around us are rare, extremely rare. They're very hard to get and some supplies are running out. So moving towards being able to mine an asteroid would be a fantastic situation. Now, obviously, that's 50 to 100 years in the offing.

SL: And the problem with these rare earth materials is that they're not really abundant in the West, they tend to be elsewhere.

PG: Yeah, there's also a geopolitical element. Many of them are just rare, full stop, and some of them are only abundant in certain countries.

SL: And there is exploitation involved and all these things...

PG: You get the same issues with, say, batteries and electric cars. There's ethical sourcing requirements when you buy your t-shirt, there's also ethical sourcing requirements on when you buy materials. But whereas in theory, you can make a t-shirt anywhere, just you need to pay your workers enough. You can't mine materials anywhere.

SL: If we think in terms of Moon exploration, Moon mining, what do you think we can extract from the Moon? What's in there that might be useful for other missions or bringing it back to the Earth?

PG: I guess there's three elements here. First is, this isn’t my area of expertise. The second is that there's the materials that will be useful here on Earth and bringing back for which I don't think there's a lot the Moon has. The Moon's main element now would be as a staging base for future exploration. But there's things that we could be mining on the Moon that would be useful for building a Moon village or Moon system. For example, I was in a presentation six months ago by a guy talking about building reactors for taking lunar regolith to extract oxygen from it, where you ended up being in interesting situation where metals were the byproduct.

SL: You can build stuff with the regolith. Regolith is the moon dust basically.

PG: One of NASA's big programs is the idea of building this Moon Village. The big issue is the Moon does make a good starting point. It's a solid base on which to build. Because gravity is one sixth [of what’s] on earth, it's very easy to get off of it. So you could very well imagine a scenario where we captured resources from an asteroid, we then did manufacturing on the Moon, and then we shipped things down well, to Earth.

SL: Okay, so you're talking in terms of electronics manufacturing…

PG: I mean, that's the obvious one, but basically very high value to weight ratio.

SL: I think we discussed this thing back in July when we had the first chat and I was saying that I don't think... when you think in terms of, let's say… I'm familiar with scanning probe microscopy where you have a probe and then you look at matter at atomic level. So you can get for example scans of gold lattice, with atomic resolution, you can see the gold atoms, you can even see silicon atoms. One of the things they do is that, of course, they do it in vacuum, but they also use a system of springs or even magnetic levitation to damp vibrations. Then they can even put the thing in an anechoic chamber, which is a sound proof chamber. So, do you think it's worth? I mean, if we have all this kind of procedures that we can do here, do you think it's worth doing that on the Moon? Is it still relevant?

PG: Yes, I think it's about costs and I think it's somewhat of an open question. Because I'm the engineer here, I'm not the big picture economist. But my understanding is that there's a potential, it could be very valuable. The issue is about basic cost away. So we could put some numbers on this, if you want to put something into lower Earth orbit, a rough rule of thumb, it's about £100,000 a kilo. Obviously, it's significantly cheaper if you do it in bulk. If you want to put something into orbit around the Moon, it's roughly a million pounds a kilo. Again, this is to an order of magnitude only. You may be able to get it much cheaper, it may cost you much more depending on scenario, you want to land on the Moon, you have to double it again. If you want to drive around on the Moon rover, you have to double it again. It can be very, very expensive.

SL: Maybe you need to manufacture electronics for usage in the Moon itself. Maybe you're going to have a base and you need to have that electronics instead of waiting for a cargo to be shipped from the Earth.

PG: Yes. So one of the questions I've been asked a couple of times I've been a panelist is always like, so when are we going to send humans back to the Moon? And my argument would be, I'd actually like to have our company tagline as "Exobotics beyond human", because I don't see any point in humans. Obviously, not in general. But in terms of space exploration it's cool to have a human in space. The reality is I don't think in the near future, there's ever going to be an economic demand for humans in space. There's nothing that you can do, that we can't do with a robot, bigger, faster, better, cheaper, lighter, you name it.

SL: But then you can always find someone that wants to go there for the experience. There are things that robots cannot do and having a human there is always good. Plus there is the excitement aspect that people would like to...

PG: So there's massive economic advantages for society having humans [on the Moon], and it's cool, it captures the mind and that's the reason why Elon Musk's wants to send people to Mars. That's fantastic and it would be a good thing to do. I'm thinking from the engineering point of view of in terms of doing something.

SL: Yeah, I know what you mean.

PG: A human is a 60 to 100 kilogram, organic vehicle that excretes material and requires mass quantities of material. That's a lot of stuff that needs to be looked after. There happens to be a lot of stuff to look after, we know what humans eat. There's always gonna be unknowns on the Moon or things that we may not have the robot prepared for. I think we're a long way from, I guess the quote unquote from Star Trek’s fabricated technology. If we go to the Moon with a robot, and we suddenly discovered we need a slightly different robot, we aren't gonna be able to make it on demand. But we're not too far from that. 3D printing is only a first step. You could imagine, in 50 years, that we'd be able to design, simulate, VR thing on Earth. But again, this is obviously a concept as all space stuff is. The engineer in me always wants to sort of solve the real world problems in front of us now. People wanting to go to space is a big part of that. There's a lot of the motivations for commercial space that is either underwritten by the taxpayer, particularly the US taxpayer, or there's been people with lots of money who want to do cool things.

SL: Yeah. And I think another... before I forgot to mention that another thing that can be extracted from the Moon is helium-3, because that's rare in our planet and that might be a fuel for future fusion reactors.

PG: Yes. Helium-3 is one, there are a couple other things as well. But I think the primary motivator of the Moon is that it's the nearest body is a stepping stone for places beyond. And the big one would be...

SL: Like Mars, for example. 

PG: Yeah, the big one would be that. The reason why SpaceX can carry a much larger payload than, say, the Apollo missions could, is that once you burn all the fuel to get into orbit, the amount of space you've got left for payload is very, very light. So even very minor weight savings in your rocket or efficiency savings can make massive difference in your payload. The issue is [that] there isn't a lot of efficiency to be made. 

[With] my car engine, you can actually burn fuel a lot more efficiently now, when you could 50 years ago, 50 plus miles to the gallon as opposed to 10. Whereas in space stuff, you're burning the same fuels we burned before. It's not like we discovered a revolutionary new fuel and you're just shooting out the back of your rocket. So maybe you can be a little bit more efficient in that. There's massive savings to be made in the way to your rocket and in the efficiency of your payload. But the breaker would be if we could send the payload to the Moon and then we could refuel or build a new rocket and then go from the Moon beyond.

SL: Or build extra stuff that we might need for Mars and build it... So can you explain why it is convenient to ship stuff from the Moon instead of from the Earth?

PG: The big term people use in the space sector is what's called Delta-v. And the idea is heavier objects warp space time to give you your gravity well and the amount of Delta- v you need to get out of Earth is a lot. It's obviously not as much as the Sun, but it's a lot. Whereas the Moon, because its gravity is about 1/6 of Earth, […] needs a lot less energy to get out of its gravity well. The result it that you've got a lot weight left. And there's some quite unintended consequences. 

I'm relatively new to the space sector, I joined when I started Exobotics with my two co-founders. So there's a lot of things that took me by surprise. One of the ones was that actually it's easier to go to Mars and is to go to Mercury, in many ways. 

SL: Mercury is too hot. 

PG: Not in about the planet, just in terms of purely in terms of the orbital mechanics. Going to Mars gives you... you have to burn fuel to get there, more than if you wanted to go to the Moon. Mercury is down well of Earth. So actually, your issue is you have far too much speed you can't get rid of. So all of that, Mercury is closer and would be, on paper, easy to get to. It's actually much harder to get to [Mercury] because you're moving way too fast that you can't lose that velocity. That was one thing that actually in hindsight makes perfect sense. But I certainly didn't expect it when I came into the industry. That's why there'd be more much more rovers on Mars and Mercury.

SL: I don't know what's the temperature on Mercury? Let's me check. It's gonna be bloody hot.

PG: I think it is pretty hot. 

SL: 400 degrees. I think Venus is even hotter than Mercury. That's because of the atmosphere. Yeah, no, it's about 460. It's hotter than Mercury [wrong!]. That's because of the atmosphere. I don't think there is such a dense atmosphere on Mercury. But one thing I wanted to ask you about… so using the Moon as a platform for exploration, and other places in our solar system, could be Mars, could even be the Moons of Saturn or Jupiter, whatever. How do we get the fuel?

PG: I believe you can make it. But again, this is not definitely not my area, but I believe you can make fuel from the regolith. But yes, feel free to fact check me.

SL: I think the Artemis is going to land on one of the poles of the Moon, right? Where there is a high concentration of water. If you do electrolysis with water, you can get oxygen and hydrogen. But I don't know… what's the fuel for normal rockets? I think there is oxygen, right?

PG: Again, I'm really sorry. I don't know. It is. I'm still playing catch up.

SL: I think there is like… probably there is methane or something, liquid methane. I don't know. Another thing I wanted to ask is that one of the things they say is that dark side of the Moon would be a great space, a great area for doing astronomical observations…

PG: It would be yes. It's also cold. Very cold. 

SL: Yeah. 40 Kelvin or something. 

PG: Yeah, Exobotics has looked a couple of times at doing stuff in... So the there's two things here, the dark side of the Moon isn't dark. Now that's another thing that surprised me is the dark side of the Moon is a bit that is facing away from Earth, but it has the same exposure to the sun as our side. So a day on the Moon is 14 Earth days, and then there's a 14 day long night to give you a 28 day Moon cycle. But the dark side of the Moon, if you look at a crescent Moon, The dark side of the Moon can also be the bit that's facing the Earth and vice versa. So dark side of the Moon refers to the bit that's away from Earth. 

SL: So opposite to the Earth. 

PG: So the key thing about the Chinese rover [is that it] wasn't landing on the far side of the Moon. At least from the engineering point of view, the key thing was it could operate during lunar night, which is incredibly cold. That's one of the biggest issues about lunar exploration. It is not that it's cold, because cold we can handle. It's that there's no sunlight, there's no power inputs. You have to run on batteries or basically tiny little nuclear radiation things that keep the thing warm. Part of the issue is because it's a big rover. Square cube law means that it loses less heat per unit of volume as a small craft. That's one of the things we've really had to work on for Exobotics.

SL: Okay, now I'm curious of one thing... now, let's talk about the technical challenges on the Moon. Certainly [moon] dust is a big problem. I mean, if you're sending humans, that dust is apparently toxic, you might get silicosis, which is what workers get in very dusty environments. There is also another problem, that [the moon dust] is electrostatically charged. It gets everywhere. How do they fix the issue with the rovers... because that thing can go anywhere... it can go in between things...

PG: You try not to kick it up, is ironically, the biggest thing you do. You also try and elevate your craft a lot. It's one of the reasons why... So, there's basically two big reasons why lunar rovers are so far from the surface as far as... the big ones. One is that rocks on the Moon are awkward. They're not like in the pictures, all these giant rocks with and basically smooth sand around it. That's how I was always done in artist’s impressions. Actually, the reality is that rocks on the Moon don't really allow you to drive around them, you typically have to be able to drive over. There's quite a wide variety and sizes. Although obviously, previous missions have tried to go to places as smooth as possible. And the other one is dust. So you want to keep your metallic instruments away from it.

SL: So you need to go slowly. So that didn't bring it up.

PG: You go slowly. You also have wheels that try not to disturb it. So they won't have like a standard tire grip. They have like veins that slot in and out. So they smoothly go in and out while still providing good grip. Those are two big things. Also, you're trying keep all electronics self-contained.

 

SL: Yeah. Because [the Moon dust] is electrostatically charged.

PG: Yes. But that's not... The electronics as a situation isn't so much the problem, obviously, you could shield it. It's more just dust getting where it shouldn't be. So you wrap everything up. I would say the big issues, the biggest issues that we've been facing, obviously there are lots, there's temperature, there's mass, there's radiation, and then there's magnetic field. So the biggest one is temperature. If you're operating in space, you've got either constant exposure to the sun or you've got only short periods of eclipse. And you're exposing cross section to the sun is consistent over time. And that's pretty nice, because we can build a craft that operates at, you name a temperature, between -70 and +100 or so, we can build a craft that operates at roughly that temperature.

SL: Celsius degrees. 

PG: Yeah, sorry. And we can change albedo with the craft, the amount of heat it rejects. So as long as we got constant heat input, we've got a lot of design choices to be able to work for that. The issue is when you've got variable heat input, say you go into eclipse and you're hidden behind another body or you have something like a lunar rover, which isn't a sphere and over the course of a lunar day, the sides of it, though exposed to the sun changes the cross sectional area and also the amount of heat it can reject. So thermal is, the big, big one. 

Mass is another very big one, because it's very expensive to get things out. So rough rule of thumb, and again, it's only a rule of thumb, but the cost for things in low Earth orbit is primarily dominated by volume. So you can afford things to be heavy. Now, if you sent something might have lead up, people are gonna complain, but it's primarily volume. 

So for the Moon, we've had to look at entirely new ways of making things are lighter, stronger and things... There'll be a whole bunch of composite materials and new architectures for deployment [that] are not needed for low Earth orbit. 

Radiation is another one, whereas on lower Earth orbit, the two key things are (a) you're primarily protected by the Van Allen belt, so there's a lot less of a radiation issue. (b)...

SL: So, can you explain what's the Van Allen belt?

PG: Oh, it's this... it's effectively a shield around [the Earth]... let's see if we can get a picture... but yeah, it's effectively a shield around the Earth, which, which captures a lot of the radiation.

SL: And that's due to the magnetic properties of the Earth...

PG: Yeah, so basically, it captures radiation, and it almost forms a blanket around the outside. That's my understanding, again, not my area.

SL: And this radiation basically causes problems towards electronics...

PG: Yeah. High energy particle coming from your solar system, cosmic background radiation or something in it... a classic example [is that] it may flip a bit in your computer. And if that bit is in the “which direction do I want to be firing my thrusters command”, then that causes problems, though my biggest role within robotics is on the software side. So how do you write software that doesn't crash?

SL: So redundancies and all these things?

PG: And sometimes it's as simple as okay, you run the program three times and you get the vote. But what if the voting procedures gets broken or something like that. So it's answering questions like that. And there are a bunch of tricks and things you can use and it's more of an issue on the Moon because (a) the radiation is more abundant, (b) radiation is very difficult to shield for. Also, shielding is heavy, because shielding is basically... you basically can't have light radiation shielding.

SL: So shielding… how would do it With metal and Faraday cages and things like that, and with thick materials like lead or something...

PG: In terms of materials straightforward to use, aluminum has the best weight shielding. Yeah, so for volume, lead is best, but obviously lead is really heavy, whereas aluminum is quite light.

SL: Can you use electromagnetic fields, like magnets or something, similarly to what happens with the Earth with the Van Allen Belt?

PG: Faraday cage will pick up some things. But that's not typically the major issue. It's the high energy particles that Faraday cage is not going to stop. That's the problem.

SL: So using a magnet wouldn't work...

PG: Not really, no...

SL: It would be too heavy...

PG: Also, yeah, whoever's taking your mission up to space is going to complain of as a giant magnet.

SL: That's right. Yeah, it's gonna be very heavy, very, very heavy or require a lot of power. 

PG: So, there's a major issue or at least, it could be a major issue. As technology has moved on. It always used to be people trying to avoid ever having bit flips and ever having problems. And as technology moves has moved on, people have gone towards a model of being able to handle the problems when they arise. I personally much prefer that model. It keeps me in the job. But also, I think it's a much more sustainable model in future...

SL: How do you detect a bit flip?

PG: You typically don't. I mean, you can, for example, when you communicate over the internet, and you send someone a message, you would normally send them with a checksum as well. So for example, if I send you a message, "Hi, Samuele", I'd probably get hash of that, somewhere, a mathematical function of the message I'm sending and send that. Send the checksum. Then you, at the other end, can do the same piece of math and if there's an error, you'll know. But then there's more advanced code systems where you can error correct. It would be a bit like you have like a RAID hard drive or system. 

SL: Yes. I was thinking about this. I wrote RAID 5.

PG: Basically, it's that. That's the message equivalent, and there's memory equivalent, there's computer operating system equivalent.

SL: Can you use a Blockchain system for that? 

PG: Yes. 

SL: Because if you're sending the wrong thing is not gonna be accepted, right? There is a check that needs to be performed for the following block, which could be a message.

PG: So as a lot of people that are very interested in using Blockchain or crypto in space-esque stuff. There's several reasons why that one is cool. Another one is that the openness of the ledger makes life a lot easier security wise.

SL: Would you be able to explain how this thing works? Because it's, I find hard time explaining it.

PG: Do you want the one minute over two minutes?

SL: Whatever you want? Go ahead. 

PG: So the one minute introduction to internet security is that one of the biggest inventions of all time in the in the Internet security world is the idea of a one way function or a one way lock. The idea is that you can encrypt something using one code, but use a different code to decrypt. The idea is a public-private key. Imagine I published my public key on a website. Now if you want to send me a message, you can use that public key to encrypt it. But you couldn't then unencrypted a message again. If you like, it's like a padlock, which if give you an open padlock, you can close it, you can't then open it again. And then you send me a suitcase and you padlock it with my padlock on only I can unlock it. 

Then a lot of security problems boil down to... Okay, so let's say you want to send me a message for the first time, how would you do it? For example, you could take your suitcase, you could put your padlock on it. And then send me the suitcase and I could padlock it with my padlock. And I could send it back to you and you could unpack, take your padlock off and send it to me. And I couldn't unpadlock it. And at no point in the middle, could someone unlock it. So that's, I guess, the 101 on internet security. And Blockchain is not a dissimilar approach of public-private key relationships and on a ledger system. But yeah, I'm not sure... I'd probably I'd probably have to go and do some prep. To find a good way of explaining it, that doesn't become too convoluted.

SL: So how did you become interested in space exploration? When did it start?

PG: So I'm going to be blunt, my interest isn't in space and expression. Genuinely isn't.

SL: In building rovers...

PG: It's not even that. There's two other engineers in the company Nadeem Gabbani and Maxime Burgonse, and they're amazing. My interest is the engineering challenges and space has a lot of them. They're really, really interesting. But if I found those same problems somewhere else, I'm just as interested in something else. 

SL: So it’s the approach that matters? 

PG: Yeah, it's what problems do you get to solve today? And there's one thing I've always, always tried to do for jobs and work is to find something that is going to be fun to solve, may not be easy and may not be lucrative. But it'll be fun.

SL: But if you tell people "I work in the space industry, and we're doing all these things". 

PG: I mean, that is true. It certainly is certainly good for starting conversation in the bar. So Exobotics started in 2018 and Nadeem Gabbani and Maxime and I we wanted to join. And we said yes, we've been through a few iterations since then. Our first contracts fell through because one of the launch provider went bankrupt. So we had to reinvent ourselves building cots [?]. And now we're back I guess where we were before. So we got four missions lined up for this year. That means customers have paid for four things. I'm very limited on what I can say I had to get my crib notes handed to me by Nadeem, but one of them will be going into a lunar orbit and it's on a CubeSat. So that'll be quite a big thing for us. We'll get to test a lot of our kit in lunar orbit, which has never been done before. If that works, which obviously, we're doing our best to make sure it well, that'd be a fantastic, it will be major breakthrough for us and also for the community as a whole.

SL: So you started collaborating with a company called the PTScientists which is a big German company.

PG: Yes.

SL: But they are backing businesses as far as I'm aware.

PG: Yeah, they are. They got bailed out by the German government. And then now known as PTS... by the time that had all gone through...

SL: It’s a big company, a big space company, right?

PG: Yeah, we are decent size. And we went out there, I think in early 2019, for a PDR. I was just after...

SL: For a what?

PG: Preliminary Design Review. So you would typically take the, I should say the basic idea, but the key requirements, it's a preliminary design...

SL: And that was for the rover. 

PG: Yes. And then you then go back, and you do come back for CDR, which is like the technical review. And that's when it's approved, and you can start going building it.

SL: And so what's the rover like? Can you describe it? How heavy is it? 

PG: I can show you a picture from the website if you want.

SL: Shall I open the pictures that you sent me?

PG: Oh, yeah, there's one there as well. I think there should be a picture of the actual rover. 

SL: So this one is the size of a CubeSat. What's a CubeSat?

PG: So CubeSat us we're a an idea, I believe, concocted by some students that wanted... from an American university, I forget which one, in the ‘90s. And the idea was to basically modularize space in a similar way to say, with a computer, now, when an ATX system, you can buy a case and a motherboard and a GPU separately, and there are standards to make sure they all work together. It's a good idea, because up until that point, satellites were very, very large, very, very heavy. Also, they weren't very standardized. Having a standard system meant that space was a lot more accessible for a lot of people. So that's the popularity of CubeSats. 

I forget the exact numbers, but a vast percentage of all satellite missions are now based on CubeSats. This means you have a standard deployer, basically standard everything. In theory, if you had the money, and you were willing to wait for the lead times, you could buy everything you'd need for CubeSat, right now. It wouldn't take you more than a few hours’ worth of quote, gathering to buy a bit, we could now screw it all together in a couple of days and then we could send it off to the fly. Obviously, we'd need to test it and meet lots of criteria and the design bits not that straightforward. It's not as easy as building your own computer. But it's not that far from it. 

SL: This one, what kind of sensors does it have?

PG: So this is a platform. This has cameras and the things we need to drive around. If you go on to the next slide, you'll see that the legs fold out like this. So this is the CAD model, obviously we haven't… that's not the real thing. The idea is that it comes with about 0.5U of payload space, as refer[red] to [1U corresponding] to 10x10x10 centimeter cubes. So that's a 2U version. And of that one quarter, so 0.5U 5cm tall by 10 by 10, is up for the customer... and whether that [?] system.

SL: Those are solar cells and I have seen that you also manufacture solar cells and you sell them separately as well…

PG: Yeah. So our aim is eventually to be to control the full stack to use...

SL: So you sell different parts. 

PG: Yeah. Currently, if you buy one of our systems will be about 50/50 stuff we've built ourselves and stuff you can buy from other companies. Our aim is eventually to sell the whole thing. I think the future of satellites and any small satellites, in my opinion, is going to be something called Software Defined Satellite. Because the actual hardware that goes into a satellite is not that expensive. Well, the manufacturing cost isn't expensive, it's very expensive to buy because they're very low volume. And it's very expensive to buy because there's a lot of test involved. I'm very keen for us to move to the point where you can buy almost the entire satellite, the board, there's a minimum of connectors and many more things to go wrong. Yeah, and then the software is what defines it.

 

SL: So the hard part is what actually goes in the [software] system basically.

PG: In terms of failures of satellites, things go wrong in space, sometimes for reasons we know, sometimes for reasons we don't. There's lots of unexpected things. One of the classic examples being a phenomenon is cold welding, where you have two metals with a very, very clean, very, very crisp edge in a good vacuum, they will often weld themselves. Deployable things like deployable panels or something can have problems. Now we've learned to deal with that, we have dissimilar materials and stuff, but there's the unexpected can happen. Also, sending satellite up on a rocket, it's pretty, it's pretty high intensity, it's not as likely gentle...

SL: Vibrations and stuff. 

PG: The inside of a CubeSat may have 20 different PCBs all wired together and any one of those connections could go wrong. Ideally, I'd like to move to having satellite on a chip where everything, except for batteries and solar panels, is in a tiny little box...

SL: All in the same chip.

PG: Yeah, that is the dream, but we're away from that yet. But Exobotics' big vision is that we want to make the whole systems. So if you want to be a customer, you come to us and you say you want to fly this mission, we build your satellite, optionally, we'd integrate it for you, optionally, we'd operate it for you once it was launched. Our aim a couple of years’ time will also build over kit, because lead times in space industry have very high. Like, we've currently got one part that we're expecting to take 32 weeks to be delivered, but supposedly, commercially available off the shelf.

SL: I understood that the plan is to, if you want to do some exploration is to send a swarm of these kind of little robots that can cooperate and work together to achieve a certain target. Is that the kind of idea?

PG: So actually, communications on the Moon are actually quite tough. The reason for this is that comms on Earth are easy, because radio bounces off the atmosphere. That's not true on the Moon. If you lose a line of sight, it's actually very hard to talk to something. And then when 10-20 years time, there's probably going to be very reliable ways of getting around this.

SL: Can't do just have satellites?

PG: You've actually got to do that and the number of satellites you've got to have for coverage isn't not one.

SL: And line of sight wouldn't work because the curvature of the Moon is different from the one in the earth. I don't know, what's the line of sight that you can get on the Moon? Maybe a few kilometres…

PG: It depends on the terrain. I mean, we'd like it to be able to go down inside a crater on the Moon or a lava tube or something. So a big part of the swarm theory is that it allows robots to work together to keep effectively a chain of communication while covering a reasonable amount of ground.

SL: And maybe one breaks and the other one can take over and so on... redundancy.

PG: Yeah. So that particular one there minus the payload weighs only just over a kilo.

 

SL: One of those is one kilo?

PG: So it's quite the size of a loaf of bread and it weighs 1.1 kilos. 

SL: I thought it was much bigger. That's small okay.

PG: Yes, it's, it's pretty small.

SL: And what can you mount on it? What sort of equipment do you think can be mounted on it? 

PG: I don't think you would want to put anything in it disturbing. I don't think with this particular one, you won't want to drill. I'm sure we could build you on for drill or something. So this one would probably be primarily mapping and exploration. So there's little holes in the side there where we mount our cameras, but we could also mount a much more complex set of cameras and light, maybe...

SL: Maybe something for analyzing the terrain, spectroscopy or ground penetrating radar to see what's underneath...

PG: Possibly, not the radar because of power requirements. But the others definitely yeah.

SL: So it's basically exploration and seeing what's in there, the terrain and...

PG: Yes, I think in the short term, but I think we want to eventually do it all as a company. But yeah, baby steps. The issue is in the market. There's a really strong market for orbital stuff. There will be a market for Moon exploration. It's just in its infancy and we're very keen to be first movers, but nothing in the space sector moves fast. So we're having to do a twin strategy where we, we build stuff for commercial space in orbit and also prepare stuff for lunar space.

SL: I was wondering what kind of solar panels are you using there, what kind of materials are you using? I think I've seen that it's over 30% efficiency in your website. Maybe 36% or something like that...

PG: No, it's about 31%. But yeah, if you go on our website, you can see a technical details.

SL: Another thing I've seen that you guys do is you have system that can be used to test stuff that you're going to send space, you got a vacuum chamber and you also got some vibration stage. What's this vacuum chamber for?

PG: So again, so the other two co-founders on the company, with space background. I think they got frustrated with the way the space sector worked, which is that the idea of flight heritage is incredibly important. If you're trying to sell a computer, flight computer or something, people want to see that it's actually flown before successfully.

SL: So you need to fly the flight computer on an actual airplane...

PG: And no in space... for people to build want to buy. My concern is that doesn't make sense in today's manufacturing world, where you shouldn't have to try something to have a good idea whether it's gonna work.

SL: If you have a good model...

PG: Yeah, basically, and we should be able to test to space standards as much as possible on Earth. So everyone will do thermal vacuum system testing, but it will be in something the size of this room, it'll be very, very expensive...

SL: Like the ones that we see at NASA, massive chambers where they test things. I think, remember that, when they tested that drone that they sent to Mars, they were testing it in one of those chambers.

PG: So that's really nice. It's really cool. But it's very, very expensive, which means for small satellites, people will just do that at the final stages of the testing process. That's not how the rest of the engineering world is moving towards. Nowadays, you can go to your 3D printer and print something to try something out. So our idea is to basically build a vacuum test system. This one does vacuum and temperature. But later ones will also includes solar simulation and stuff.

SL: So radiation as well...

PG: Well radiation is more tricky, because the big issue of radiation testing is you then need to block it from going everywhere else.

SL: No, but you can use the sort of solar simulators that we use for solar cell testing would be a lamp that is...

PG: So that sort of radiation is fine...

SL: Ah, you have that.

PG: Not in this model, it will be in a later model.

SL: And that provides heaters as well, not the sort of heat, you know, what kind of temperatures can you get with that system, with the system that you have?

PG: I mean, high temperatures are easy, because it's just resistive heating, low temperatures, we use a heat pipe system to get it out. And we'll get down to -50, -80. But again, it depends on the model, because it's very much dependent on the amount of back end basically. So the cheap ones are relatively high, low, lowest temperatures [?]. 

But the aim is eventually that you'd be able to sit down between your espresso machine and your 3D printer and your laser cutter, you'd also have one of these and a test kit. 

Because it makes much more sense to how to do it that and you combine it with your digital twin you draw your modeling, in a program like in our case, [Autodesk] Fusion 360, you then run a simulation in your you can go to your printer when you print the test parts, and then you see whether it works. And that rapid iteration does is much better for design and manufacture.

PG: Can you simulate the electronics or just the mechanical parts with the [Autodesk] Fusion?

PG: So with Fusion, you can do both, but it depends again on what you're trying to do. Fusion is certainly aiming to be an all in one package, but it's not made, not made by us, and we definitely go outside. For the optics for example we would use OpticStudio by Zemax and things. We use a wide range of tools. But fusion is typically the starting point because it's definitely been designed with (a) usability and (b) breadth in mind.

SL: So this [Fusion] is a 3d CAD software that you use to define and design all the parts and then you assign materials like, this is a metal, this is a plastic piece and so on. I remember that with [Autodesk] Inventor, as we discussed before in the lab, it was possible to simulate the heat transfer and simulate if things might reach a point where the break...

PG: So Fusion does a lot of that. There are some tests and more advanced tests that you have to do somewhere else. 

SL: Software wise...

PG: Software wise, like, obviously thermal models are very complex beasts and depending on the setup, it can make very big differences to what you do. But in terms of the, I guess the scoping up test, Fusion does everything. Most of the advanced test Fusion does and has a couple of things we take outside for... Fusion doesn't have an optics plugin... Obvious it's something we need quite a lot for say cameras.

SL: Could you use COMSOL Multiphysics?

PG: We have done... we have a license. Most of our stuff works well with ray tracing. So we've been using Zemax quite heavily. But yeah, I think, yeah...

SL: So basically, what you do is that you have some project in mind, you define it with the software with this AutoCAD software, Autodesk software, or Zemax, whatever and you get some conclusions, some results, and then eventually you need to test that on the on the actual, testing platform for real, right? That's, that's something you need to validate all these things.

PG: Yeah, I'm just, I think the dream, and this is a long term dream is that you should be able to understand and emulate the space environment well enough that you wouldn't need to go there to test it. Now, obviously, flight heritage is going to be really important for a long time to come. But it shouldn't be something holding us back, which it is, in many cases, because a product will have flight heritage, therefore you're buying that one instead of maybe the newer, better one, which doesn't have that flight heritage yet. We'd like to have a system where we can have flight heritage, but also there's a history to design. For example, if you're going to buy a boat or something, it will probably not be the same as the previous boat built by that boat builders, but it'll have a lot of history in the builders will be part of the trust.

SL: Yeah, an example would be the semiconductor industry, you design a semiconductor processor, a chip. You are going to do everything via software, you need to test in a computer. Because the manufacturing itself is a very expensive process. So you need to get it right. Then when it gets into manufacturing, it's not just that you make one. If just to make one, it's very expensive. You need to make lots of them. But that's a mature technology and they’ve got very good models. 

PG: I think that's basically what we're looking for in Exobotics is that the space sector has been around a long time, but it's been chronically under invested in by governments. Now, there's a growing commercial interest in it. There's a need for bringing the more advanced manufacturing approaches and design approaches that we've developed every other industry over the last 50 years into the space sector. Some companies are doing a fantastic job with this already. This is where Exobotics, that's one of the things we want to do as well, we want to change the world. But I'd rather change the world from my bedroom than by having to spend five years waiting for something to fly and discover I made a mistake.

SL: Another thing I wanted to ask, in terms of power, other than solar cells, does it make sense to use nuclear batteries?

PG: Not on something that size.

SL: Do they [nuclear batteries] only work for tiny chips or they would also work for big things?

PG: There's a lot of licensing issues, obviously, at the US government. The US government isn't that keen on people sending radioactive things to space on their rockets or on SpaceX rockets for that matter. 

Power is cheap in space, ironically enough particularly on the Moon. If you want to go into deep space, it's different. On the Moon is not that difficult to send lots of solar panels. Solar panels are really light. And it doesn't matter if they look ugly and you can deploy them. It is not like there's infinite power and you can boil a kettle on demand, but there is quite a lot of power available for small craft. The square cube law means as you scale up if your craft is now a meter cubed rather than 10 centimeters cubed, your surface area has gone up by 100, but your mass has gone up by... sorry... your surface has gone up by 10... I'll try one more time... if your craft is a meter cubed, instead of 10 centimeters cubed, then your surface area is gonna weigh 100 Instead of up, but your mass is gonna be 1000 and therefore your power demands have probably also gone up by roughly 1000. 

SL: What are those holes over there?

PG: That's the wheel stoves.

SL: So where you store the wheels.

PG: So the key thing is this particular one is designed to fit inside a standard CubeSat deployer, which you can basically think of as a square Pringles tube. The silver rods on the four corners of the craft rails, and that's what slots out of the tube. So it has to fit inside a deployer in a folded configuration. 

One of the things that's unique about this design as the other lunar rovers are being made and designed for commercial missions coming up for next few years, they're all deployed via a ramp. And they've all looked basically like small cars. That's a good design, but you don't have to design the lander around it to be configured was this will fit in a standard CubeSat deployment system.

SL: How do you land this thing?

PG: XXX

SL: I guess you have a simulation software for these things? Right? 

PG: Yes.

SL: Okay. What can you do with this thing? Can you move objects in space and then test and try different things? 

PG: So primarily… Yes, I feel Yeah, possibly... varying towards what I can't really say. 

SL: But that's something you made, right? 

PG: Primarily, yeah. So it's duly rigged on top of commercial software...

SL: It's like a video game or something like that? 

PG: Yes. A very, very ugly video game. I'm an engineer here. Function over form. So the key thing to note is that, this is a product you can buy. But what it does on the Moon is very much defined by the users. We like to consider ourselves a bus. And we're the last mile system. But the customer decides what you want to do with it.

SL: But there must be common components, for I mean, the navigation system must common... what do you have a LIDAR or something like that?

PG: I can neither confirm nor deny. It depends on configuration?

SL: Well, I mean, you can navigate just with the cameras, if you want, if you look at the most recent DJI drones, they do a very good job with the just camera navigation.

PG: Yes, the issue is obviously getting enough processing power on board. If I were to do that navigation or the delay in any kind of comms link. But yeah, is not it's I want to be clear, this is not something we've yet sold one off, we want to and it's something we're working towards. You could buy one of those of us, but we wouldn't be able to ship it to you tomorrow. Whereas like many of the bus systems on that we are actively working on our website. Yeah, we could send you tomorrow and it will be well...

SL: Do you mean the testing systems and the panels and so on?

PG: Yes, but we could also send you a satellite. So those are the four missions we're working on at the moment, they are all do to launch before the end of this year.

SL: That means coordinating the satellite thing...paperwork...

PG: So we both work. So for example, one of these missions will be going to lunar orbit, it's going to fly payload from a customer and will fly around the Moon and we're just providing the satellite for that. And that will go into a lunar orbit. Now the operation will be done by a separate company. So we just provide the satellite and the hardware and the onboard software, the customer puts their payload in, but like a CD into tray into a computer case or something. And then we ship it to the integrators who put it in the rocket. And then it gets launched. So we're, we're basically, we're building the satellite but not operating it.

SL: What's the most difficult thing that you find when designing software for these projects?

PG: Trying to reverse engineer what other engineers have done on their bits, with badly written manuals, you know, it's like, if the actual software itself is not too bad, it's trying to work at what someone else did on the previous piece of software or piece of piece of previous hardware.

SL: Trying to understand code written by other people...

PG: Yes, code written by other people also, because you're operating with lots of different components, all of which work together in slightly different ways and making absolutely confident that it's going to work every single time means you've got to have a really good feel for how something works and all the edge cases on where it could go wrong. We'll take some serious effort.

SL: Alright, so let's change the topic a little bit. I want to ask you about your intrapreneurship experience. When did you start? What was the driving force? Because I've read that you were a co-founder of something like 10 startups, you worked with Fortune 500 companies and all these things. So what's the story?

PG: So I've been on the founding team of 10 startups now? Okay. Many in the software side of things. In my consultancy business, I've worked with a lot of big companies over the years. My first startup was a security applications thing. You mentioned blockchain earlier. It's not too dissimilar. My original undergraduate degree was in civil engineering. I basically got bored and during the summer after my second year, I went and spent a weekend writing an idea I had, which eventually sold for enough. Fortunately, much of the consultancy I've done and also the companies have done have been very much on the security side of things. So I'm very limited what I can say. That's why, yeah, so it's possibly not the easiest place to have an interview.

SL: Security, like, what, like, financial security, things like that?

PG: Can we turn the mics off?

SL: Don't tell me it's alright. If it's military stuff, don't tell me anything. I don't want to hear that.

PG: So the stuff I'm working on now I can talk about. So the big one is Exobotics, we are doing reasonably well. We're cash positive and we've got a few big projects on the go. We're going places. I'm also a consultant for a couple of other companies. And then I've got two other startups. Now, Prospectral and Nanomation. They are both very young.

SL: Prospectral... you started in November 2021.

PG: That yes. So Prospectral is very, very young and we've got a novel filtering technique for multispectral/hyperspectral Imaging.

SL: Okay, hyperspectral... I use the hyperspectral camera ones. We you can basically get multiple images at different frequencies, they use it for agriculture and all these things.

SL: There is archaeology as well...

 

PG: So so like an RGB camera has effectively three wavelengths it detects but those wavelengths quite broad, like, yeah, anywhere in 100 nanometre range. Hyperspectral captures a lot more accuracy on that color information. multispectral does that but it only does it for some parts of the spectrum. Prospectral's got a better filter system than is commercially available. But the big advantage is what we're using it for. So an example at the moment is we're very interested in geology and mining, but an exploration there and also, we'll also be agriculture and recycling, which is can you tell materials apart and this is a relatively common thing using hyperspectral. The problem is the systems are large and bulky. Whereas what we want to do is build something that's basically a GoPro. I see you've got a GoPro in the corner and then basically want to do one of those but it's going to be hyperspectral. And the idea is that this some industries have got massive use for this straight off the bat and I say mining but some industries they're going to be very... it's going to take a bit more market building. 

PG: There's definitely archaeology. The one we're really excited by is actually recycling. So there's some stats and I forget the exact numbers on this. But like you send plastic to recycling and only a quite small percentage of that plastic gets recycled. And there's a whole bunch of different types of plastic that look the same to you and I, they're both transparent.

SL: So you want to do inspection when it arrives, from when they after they collected, you're going to inspect it...

PG: Recycling facilities already do that. Recycling facility? Well, it's more than once do, I want to, I want to put it in your phone, I want you to take your phone and point it at your dinner and see what it is or pointed to your, your bottle and see what the material is, and get some kind of eventually some kind of, I guess, score and how ethical it is. But also, like, is this HDPE [High-density Polyethylene]? Or is this a different type of plastic? 

SL: There are already things written in the plastic...

PG: There are labels on plastic, but they're not obviously not as easy as pointing your phone at it. And most people probably don't know what the labels mean. But that's, again, a long term vision. 

SL: But then the way they dispose it and recycle it, I mean, it's out of your control, you don't really know what's gonna happen afterwards. So it's a system problem, I think.

PG: Um, yeah, I'm just trying to give examples, all we want to do. It's like, there were plenty digital cameras before GoPro. The reason everyone has a GoPro is that anyone can take a GoPro around the box and get stunning pictures. It just wasn't works. And that's what I want to do for hyperspectral. Now, yeah, we'll see, that may not work. But that's that's our aim.

SL: What would be the kind of frequency range frequency band that you can find much dependent on the visible is from what 380nm to 7[5]0nm...

PG: Yeah, I tend the rough rule of thumbs, it's 400 to 750...

SL: They remove the filters for certain DSLR cameras to take nice pictures.

PG: And so sensors and standard cameras are sensitive over a wide range of you take the filters off...

SL: In the infrared. 

PG: Yeah, now there's two types of filters, I guess, that you're going to get on digital camera. The key ones are if it's a color cameras and have a Bayer filter built directly into the camera sensor, so straight on top of the pixels. Our filter technology is somewhat independent of the sensor used. So in theory, we could build it on top of a long wave infrared sensor just as well, we've never done that yet.

SL: How many channels can you get?

PG: Somewhat arbitrary, but double digits is the obvious place to go.

SL: double digits, okay.

PG: I mean, we could do 1000s, but you'd lose resolution. If you imagine on a, I don't even take a 4k camera, there's actually four times and you're getting let's say you've got a million pixel camera, on your computer image on your computer, and then the camera to take that would be 4 million pixels, because it'd be red, green, blue, and a normally a second green channel. Which is obviously quite a lot. So our system is basically we pick the colors you want, let's say it's red, green, and blue. And I know... special purple channel, that would be then four you'd need. So we'd be fine. If you could have 16 colors, you can have 1000 colors but and only have very few sampling points on any one of them.

SL: Do you do design the sensors yourself or you just put new types of filters.

PG: No the sensors off the shelf, you buy the sensor. And again, I'm a software guy. And I guess to some degree, the business development guy here. The key, I think the best piece of advice I've ever heard from a friend of mine, who's a far better entrepreneur than I'll ever be and also far cleverer than I am. Her line was always want, you always want to be the worst player in the band. And I think she got it from her management book. But she's absolutely right. Is that, you're not gonna learn anything if you're the person who knows the most in a team. So I always tried to put myself in teams where I'm going to learn things and people around me. Yeah, like, basically, I want to get carried. And obviously wasn't going to do my best. I want to learn things and I've learned a lot and everything I've done. 

PG: How do you manage your time? You got so many things going on. You are working at the University of Cambridge, you are a Fellow, right?

PG: Yes, I tried to delegate as much as possible and control. The nice thing about software is you can get a lot of value for time expended. Because there's a very large component of it goes on experience. And I've got quite a bit of software experience, which works out well. But yeah, I also don't sleep much and I worked very long hours. Last time I was in an interview for a job a couple of times ago. I was asked the classic interview question, what's your biggest weakness? And I think the line is, for me at least is that my biggest weakness spoils [?] my biggest strength is I get bored easily. And that's driven me to try new things. And but it's also meant that I hadn't stayed still and carried on doing something possibly when I could have done more. I've always struggled to stick at projects long term. And that's just me.

SL: So when you write code, do you get help from other people or you do everything yourself?

PG: And the moment I do it myself, like

SL: Because getting someone that can write code like you is a bit hard... right?

PG: Yeah, my, I like, I'd like to position myself in the architectural... For example, Exobotics is currently hiring for software engineers. If anyone you know, wants a job. And I feel the most value I can add is in the architect role of understanding the big picture. But to be blunt, I think it's much harder to make the hardware for something than to program it typically. There's exceptions to every rule. But certainly in satellite business, is much easier for me to program the hardware than is for Nadeem or Maxime to build it, I definitely have the easiest job of the three of us.

PG: You have you ever experience with hardware construction as well, right? I mean, with this project that you have at Cambridge, you've been working with holography since I guess your PhD...

PG: I spend a reasonable amount of time in the hardware world, like when you came around I showed you a vacuum chamber was designing. So yeah, my PhD project was on computer generated holography. So it's not very easy to explain. But broadly speaking, if you don't know... if you're not a camera, and an apple, then there's a one to one correspondence between the part of the apple that's being shown, and the other camera. So, I don't know, if you looked at the top of the apple, all the light that's being scattered off that top of the Apple gets captured by a lens and then directed to one point on the camera, and you look at the bottom of the apple, that's all captured by the same lens and then redirected to a different part of the camera. Holography basically adds a dynamic lens to your system. There's a couple of advantages and disadvantages of this. The big advantage, from my PhD project point of view, is that you can make a lens effect, give you a projected image. So you can shine a laser, what we call an SLM spatial light modulator, we can think of it as basically like a little TV screen. And if you delay parts of the light beam, you get interference patterns. And if you get delay the right bits in the right way, you can get very complex interference pattern that looks like images. And you can have like a full projector built on a system like that. 

SL: Is it like the one you showed me? 

PG: If you imagine a little like a cinema projector or something and you look inside, than you take one of them apart, you'll have a little LCD screen or digital micromirror device or something else that the light reflects off of, and then you you turn bits of it on or off effectively to get reflection. Eventually that for a lens and that's shown on the cinema screen in front of you. And you can see on the slide that the what you're showing on your display is not too dissimilar from what you're seeing. Basically, the hologram slides do a few things. So for my PhD, my interest was primarily on additive manufacturing, 3D printing. So there's roughly three types of 3D printing, obviously a sweeping generalization of it. But there's extrusion based printing. So that's the ones which you can get very cheap from home 3D printers. Most common example is plastics that is melted, where you heat the nozzle, but I always like to think of it is it squeezing toothpaste out of a tube. The second one is stereo lithography systems, the most common of which is basically an LCD screen or TV screen under neath a tank of resin.

SL: So you have liquid and then you move around this laser, and then everything grows...

PG: Well... for the stereolithography ones, you can move a laser, but typically just have a TV screen, and then the bits that are switched on cure the resin. So you're doing a whole layer at once. So there are laser scanning ones. But so the third type is the powder based ones, which typically are laser scanning. When you move a very small heat spot over, say, a bed of plastic or metal powder very, very fast. Obviously, that takes longer. But the big issue also means you've got tiny little melt pool, which is incredibly hot. Any kind of issues on the powder being slightly wrong consistency, or you're not predicting it correctly causes massive issues.

SL: So with the first two, you can make plastics and with the third one, you can make metal printed parts.

PG: Yes. So for the third one is the one you need if you're going to put enough power into melt metal. And my PhD project was on could we use holograms to melt powders. And basically, to melt a whole area at once...

SL: Instead of sweeping...

PG: So I did the first half of this as a PhD was over runs, which is I basically built a holographic stereo lithography system, which was fun. But the powder based one is ongoing. But the aim is... yeah, we'd be able to mount an entire layer of metal powder in one go. 

PG: [In the lab] So this is in pieces, I'm afraid. But this is the the holographic 3D printer I built for my PhD. I was like I think I printed on it, for example...

SL: So the parts that you showed me in the lab...

PG: They're plastic, so they're made on the stereo lithography systems, the type two, so it's a vat of resin. And instead of having an LCD screen under that I'm using damp switching on and off to cure it. I'm sharing a hologram on top to layer it up that way. 

PG: [In the lab] One of the things that interests me a lot about 3D printing is the printing mechanisms in place. And this one didn't work exceptionally well. But in theory, you can turn the bottom gear on a top gear turn and it's printed. 

SL: This one was printed on in one. 

PG: Yeah. So stereo lithography is a lot more forgiving of having poor supports. 

SL: So all those little tiny parts, the ones in red. Those are all made using stereolithography system using holograms instead of the LCD screen.

PG: And then the blue ones are made on a Prusa SL one printer for comparison, which is standard LCD screen storage boxes. 

PG: How about the piston. 

SL: That's just an extrusion printer. That was just me playing around and...

SL: And that big one, the red one that was on top of your [drawer]...

PG: Oh, yes, that's, that's another personal project, I've started working on for an [?] it's not finished. I never have the time to get around to it.

SL: And so yeah, tell me about your PhD then you you finish this part. But and did you managed to... has anyone managed to finish, complete the part, where you have the powder and the Holograms...

PG: [In the lab] We're still working on it. The reason it's all in pieces is I've got a 200 watt CW laser in the corner, and very expensive SANTECH SLM in the corner as well. So I'm designing a new system to integrate them all. 

PG: We actually were having interviews for my replacement last week. I only technically finished my PhD in September. So only a few months ago.

SL: Last year? 

SL: Yeah. So I'm only just done with my PhD. Someone says Dr. Christopher, and I still look around for my dad. Nowadays my pitch is more on the one, in my opinion, one of more exciting things about holograms. So holograms allow you to do true 3D. That's the real magic. Holograms come with a very high computational overhead that you wouldn't need to do with a classical image and they have a few advantages, like the infinite depth of focus thing I was showing you earlier. But the big one is you can do true 3D. Instead of like a VR headset when you've got stereoscopic vision. I don't know if you're at school, but you're probably at some point during that school, your teacher would have told you to put a hand over one eye or close one eye and look at the world around you and see how everything was flatten out. You probably did it and you probably thought the teacher was wrong. That's my story. The reality is that your brain can see 3D with just one eye, it's not as good because the stereoscopic vision thing, which is having two eyes looking at same thing, like different angles is the dominant feature of 3D vision. But there's a lot of other cues as well as the one thing holography can do is it can mimic all the cues rather than just the one. So if you look inside of VR headset, if something's in the display here and here... your eye is actually focused on that point just in front of your face, not in the distance. Even if I'm looking at a hologram view, my eyes are focused here and that can be very confusing to the brain. That's one reason why lots of people feel sick using VR headsets. And it doesn't mean it's not real one accurate, you wouldn't want to say train for a sport or to do brain surgery or to fly plane using a VR headset because you're actually teaching your brain bad.

SL: Do you get the same bad effect If you use augmented reality? Is it the same thing?

PG: So augmented reality is just a VR headset that's overlaid on the world around you in effect. It's not as bad because you for reference, but it is still an issue and a non trivial one. Holograms allow you to fix this. In theory, at least, you could... So the picture on the right there is a screenshot from Star Trek. And the idea being is that you could overlay the world around you with this virtual world with projectors. What the dream is, at some point in the future, that's what we'd be able to do. We'd put our grid projectors all around the room and they'd all show you different slightly different sides of the same image and you've got to walk around the room, we could have his interview, and you'd be wouldn't be here you'd be in somewhere else.

SL: So in terms of... because I think I've seen it somewhere. And I've seen some videos where they would create this holograms in a room, big holograms. How big are these projectors? Just to have a reference...

PG: So it depends on the technology used. Um, so a lot of things in the media that are called holograms are not really true holograms, that has clever ways of overlaying a flat image and have a lot of value. Right? I forget now some...

SL: How about the concert with Tupac at Coachella or Michael Jackson. Were those real holograms or they were 2D images?

PG: They're 2D images. The magic of good art is that you don't notice the difference.

SL: Well, if you're too far away, you're not gonna notice, right? Unless there is a drone moving around and you can see the parallax, because the trick with holograms is that you can see the parallax effect.

PG: Yes. I'm trying to think of it as a very old phenomenon as Pepper's Ghost, which is effectively the same thing as is used for... So the idea was that if you wanted to show a ghost in an old school theater, you'd put a big sheet of glass at the back of the theater, which you couldn't see the glass, and then underneath the glass, you'd have your blanket and your ghost sheet moving around. And then you'd get the reflection of that off the glass without actually the audience be able to see the glass. And that's known as Pepper's Ghost. There's an awful lot of variation. So when people say holograms, what they typically mean is some image that appears like it's floating in space, which is a legitimate understanding. Microsoft, HoloLens has no holograms in it. It's just overlaying your glasses with an image. My interest in holograms is taking, doing an actual 3D image that's independent of context. If you go to a cinema, 3D cinema, and you get up to go to the toilet halfway through, it will appear like all the actors in the cinema are following you around the room, because your point of reference is important. Which isn't true for 3D holograms.

SL: So do you think we will reach a point in which these things, these projectors, could be portable things you could have like, place them in your room and then play video games...

PG: Room-size systems long way away? Headset systems already on the way...

SL: So you could also have that in my headset? 

PG: Yeah. One of Cambridge's biggest company, biggest new startups is VividQ. They make the software for headsets, Facebook reality labs, there's as a couple... I wouldn't want to bet too much on this, but I'd say within five to 10 years, you'll be able to buy VR headsets and do this. Definitely within the next couple of years, you'll be able to get headsets include some of these technologies. Headsets have an advantage because a hologram shows you everything. It doesn't make any selection on what your view is or which bits are in focus. It takes a lot more information. It scales very poorly. So the amount of projectors you need to fill a room would be a lot, which is the big advantage of a headset set because you know exactly where someone's eyeballs are. You can show just the hologram for that.

SL: So how would it work. So you have a headset that projects a hologram where?

PG: In front of your vision, it would be like using a VR headset. Just it will be holographic as opposed to a classical projector.

SL: But then you said the problem would be resolution.

PG: So resolution will suffer, because, again, because holograms have to show a lot more information than a normal image. They can't make as many compromises. So if you have a 4K holographic projector compared to a 4K, 2D projector, the 2D projector will look better. But they're more accurate. So there are many scenarios where accuracy is much more important than visual quality. Let's say you're trying to train to be a pilot or something...

SL: But if you are watching a 3D movie or something...

PG: I don't think you'll be watching movies on it in the short term. But in a few years may be, as we scale up to 16K displays, and then whatever's next 64 I don't know what's next in the pipeline. That's when things will get interesting, because then the...

SL: I'm not sure about this scaling of 16K. Because to be honest with you, there are many videos about that, but if you are a bit far away from that TV, that's 4K TV, you're not going to be able to notice the difference between a 1080p and the 4K screen.

PG: Absolutely. Whereas the hologram difference...

SL: Because that's in front of you

PG: So that's my point is that I think we're gonna run out of the need for more pixels on display, but they'll make holograms better. And there's no theoretical upper limit, I guess, to a hologram... obviously, there's the amount of computer power.

SL: Okay, so what's your next project at the University of Cambridge?

PG: So I'm working on this 3D one. I've just got the money in to buy a LIDAR drone system. I want to test one of these 3D projectors in a real system... I've got a big lithography project on the go with Hannah Joyce and Tim Wilkinson. And I've got two other side projects. Enough to keep me busy.

SL: Okay, now another thing. I came across some projects that you worked on a archery software or something like that. Is that a possible?

PG: Archery software? I did write an applet about catapults once upon a time.

SL: What for, reenactments?

PG: No, right. My brothers in law and I built own own catapult. We went to Wix and bought a load of wood. I took some of my dumbbells to use for weights and we launched beanbags across the room. It was pretty good fun actually. We got a pretty good like wood launch this thing for... we launched big beanbag across a large church room, that was pretty impressive. It's good memories some time ago now. 

SL: Did you run into troubles for that?

PG: And no, we got away of it. I didn't break any windows. So...

SL: Okay, Peter, thank you very much for your time. It has been a great pleasure. I have learned a lot from this.

PG: Always happy to talk about engineering.