Data Science on Mars

WEBVTT
00:00:00.510 --> 00:00:03.089
Welcome to the DataCafe.
00:00:00.510 --> 00:00:03.089
I'm Jeremy.
00:00:03.240 --> 00:00:08.009
And I'm Jason. And today we're going off world to talk about planetary science.
00:00:21.030 --> 00:00:27.449
Yeah, this is a fabulous Extra Terrestrial episode I've been really looking forward to.
00:00:24.089 --> 00:00:30.750
So tell us, Jason, what why are we talking about this? This is this sounds really, really exciting.
00:00:31.109 --> 00:01:04.290
Yeah, it's quite topical because NASA have landed a new rover on Mars. The rover Perseverance. It's a part of the new or at least latest Mars Exploration programme to land apparatus robotics on Mars, observing gathering data, and figuring out four main pillars of science, which I find really exciting. It's just such science fiction. Sometimes when you read it, you forget this is real.
00:01:00.359 --> 00:01:05.909
This is really happening. And people have probably seen it in the news.
00:01:06.180 --> 00:01:07.829
This is Perseverance.
00:01:06.180 --> 00:01:07.829
Right?
00:01:07.980 --> 00:01:11.730
Perseverance. It was launched in July last year.
00:01:11.760 --> 00:01:18.359
2020. Okay. Right. Yeah. Like launched during the pandemic as well. So much difficult disease at the best of time.
00:01:18.390 --> 00:01:21.750
Yes. So that's true.
00:01:18.390 --> 00:01:25.829
That's very social distancing, and all kinds of sanitation into the mix before we can get this thing off the ground yet.
00:01:25.890 --> 00:01:37.290
Yeah, you can see the model in the control room and the landing day so it touchdown the 18th of February this year, so it's just shy of two months ago.
00:01:37.590 --> 00:01:41.459
So what's Perseverance his core mission then? what's what's it there to do?
00:01:41.849 --> 00:02:14.490
Yeah, it's forming part of these four pillars. As I mentioned, the science goals of the aspiration programme, which are to determine if life ever arose on Mars number one, characterise the climate of Mars characterise the biggie. If they do, not so far, it's going to look at mineralogy as soil samples and ultimately get observational data and samples that could be looked at and studied for signs of fossilised microbial life, which will be amazing.
00:02:14.819 --> 00:02:27.810
I think that makes this sort of trying to extract 2% extra efficiency on a on a production line. Look, like small beans by comparison. I'd quite like to work on a data science project that discovered extraterrestrial life along the way. So yeah,
00:02:27.900 --> 00:02:30.270
Yeah. And not have somebody say, I think that's an outlier.
00:02:32.250 --> 00:02:33.120
How unfortunate.
00:02:34.650 --> 00:03:04.289
And, and the fourth pillar is to prepare for human exploration of Mars. So gathering all of this experience in landing on another planet, sends back all of the data of how it has encountered the atmosphere, how the landing system has worked, how successful it has been, and ultimately how safe it might be, and what we need to learn for the technological advancements to enable humans to ever journey to land on and explore Mars.
00:03:04.469 --> 00:03:09.000
Very Elon Musk then.
00:03:04.469 --> 00:03:09.000
Yeah, no. That's huge, huge.
00:03:09.449 --> 00:03:11.009
Now you've got investors interested?
00:03:11.460 --> 00:03:38.129
Yeah. I must have watched about four docu-dramas about landing on Mars and all of the challenges and extraordinary endeavours that that is going to take for a human landing. But yeah, we're really talking about a robotic landing here. This is a quite autonomous sort of effort. So why do you think this is a terrific exemplar of cutting edge data science, then?
00:03:38.759 --> 00:04:00.419
Yeah, the rover itself always reminds me of Short Circuit, the movie like it, as if we have put a robot representing us on another planet, roving around, picking up samples of data, sending back observations, it's taking selfies. It's an arm with a camera on it, and it has taken its own respective selfie and
00:04:00.569 --> 00:04:01.560
It's got a selfie stick.
00:04:01.930 --> 00:04:23.769
Basically, you can look at data live, not live but the most recent data of photographs, just photos of the surface of Mars, and it just looks like an old set from Star Trek. It's amazing. It's this barren, dusty wasteland that really does look otherworldly. It's phenomenal.
00:04:19.930 --> 00:04:36.610
And this is a robot up there that we've sent doing it and another really interesting bit is not just the Perseverance rover itself, but Ingenuity is the name of a helicopter the the first powered flight...
00:04:36.970 --> 00:04:40.420
Oh I heard about this.
00:04:36.970 --> 00:04:41.740
Yeah, yes. From from a rotor craft, I think they were calling it.
00:04:41.899 --> 00:05:03.860
Yeah, exactly. But in an extremely thin atmosphere. So right. What I find really interesting thinking about it is we make such a massive point in our own history about the evolution of flight. We are now I think tomorrow as we record this, it's been postponed to tomorrow going to have the first fight on another planet, which is mind blowing.
00:05:03.959 --> 00:05:11.550
So that's that's what a fully autonomous rotor drill rotor driven flight on it on another planet another world.
00:05:11.579 --> 00:05:39.779
Yeah, that's that's just amazing happening tomorrow and all of the sensing equipment, the data streams that are going to go into, you know making that flight a success and then making that mission a success that this is surely why I mean this is data science sort of plus plus Yeah, because the amount of data they're going to be getting, and then the decisions or the inferences I guess they're going to make from that data, it's gonna be really, really exciting.
00:05:39.839 --> 00:06:31.920
And you alluded to something there that acknowledges our previous episode, or Well, our first bite episode where we talked about the complimentary team, and what that means for data scientists and how important it is to have that diversity and wealth of knowledge and various backgrounds and everything. This is it at its Pinnacle, to bring together astronomers, physicists, engineers, computer scientists, geologists, astrobiologists, and your mathematicians, statisticians, everything comes into the mix to make this work to make it a success. And it has to be as successful as real female. Yeah, because it's it costs such a fortune, you know, in funding and in time and effort, and you only kind of really get one chance, you know, one mission only really has one chance to go right. And boy, if they got this one right,
00:06:32.110 --> 00:06:39.550
Yeah, there'd been a lot of failed attempts to get to Mars, certainly from from British space explorations efforts.
00:06:39.550 --> 00:06:44.920
It represents one of the most challenging planets to get to. Yeah.
00:06:45.250 --> 00:07:08.350
And these flights, they're not cheap. I mean, to get anything onto the planet of Mars, you're talking this one to$2 to $3 billion, or something, probably considerably more, once you start to think about supplementary flights and support flights and anything involving human beings, you'd probably multiply that by 10. So it gives you an interesting perspective on risk. I think when you're putting that kind of money on the line.
00:07:08.389 --> 00:07:22.100
Absolutely. Yeah, your your project managers is one that I didn't mention and how crucial they are, and managing that risk at some risk register to be looking at when you're building a mission like this.
00:07:19.399 --> 00:07:23.420
And again, just to echo what an amazing feat it is.
00:07:23.779 --> 00:07:33.500
We are lucky enough to talk to Professor Paul Byrne, who's planetary geologist who's really been involved with with this, this amazing story.
00:07:37.189 --> 00:07:45.769
Joining us today in the DataCafe is Professor Paul Byrne. He's Associate Professor of planetary science at North Carolina State University. Paul, thanks for joining us.
00:07:45.829 --> 00:07:48.860
Jason. Hi. I'm delighted to be here.
00:07:50.009 --> 00:07:55.199
So Paul, you work in planetary science, tell me about planetary science and the kind of work and research that you do?
00:07:55.480 --> 00:08:20.980
Sure. So I am a planetary geologist, I studied geology of Earth. And for the last 15 years or so, I've been studying the geology of other planets. And the whole goal of planetary geology, in particular, planetary science generally, is to understand how planets come to be and why they look the way they do. And the whole goal is to basically get more understanding of our own planet, and how that fits into like the broader context of planets generally. So that's what I do.
00:08:22.180 --> 00:08:26.620
Cool. And so how do you study that, for example, on Earth in the first instance,
00:08:26.740 --> 00:09:45.220
So there's, I think, arguably kind of three main strands we can take to study geology of Earth, we can do fieldwork, you can go and you can take samples or you can visit a field site, you can take measurements, you can take photographs. You can do remote sensing work. So for example, we use satellite data, or we fly geophysical measurements and aircrafts to look for ore bodies, if you're doing mining exploration, for example. And then we can do any kind of number of laboratory or computer modelling back home back back in the lab back in your office, where you might have numerical codes, medical solutions and stuff. Or you might have a geochemical lab, you might take rock samples and analyse them for their chemistry, there's a variety of ways you can do kind of back at home stuff, with the overall goal of understanding what it is you're seeing, and then inferring is in the case of geology, but in the case of a lot of different logic, science disciplines, you often have very incomplete data. And so the goal, for example, in geology is you know, what the surface looks like. And you might have some small inklings of what the subsurface looks like. And you got to make sense the whole thing in 4d, right? Not just the 3d shape of this thing, but also how it's changed through time is a good place to look for silver is a good place to look for oil gas, is just a good place to build a bridge. Is this going to suffer a massive earthquake? Is there any evidence of the past?
00:09:42.159 --> 00:09:51.909
So that's the goal of how we do geology on Earth. And we pretty much do the same approach from assigning the geology with the planets.
00:09:52.529 --> 00:09:55.259
So how do we study it on other planets then?
00:09:55.960 --> 00:09:59.409
With great difficulty and with great expense and...
00:09:59.409 --> 00:10:00.519
A bit of a leading question, right?
00:10:00.419 --> 00:10:54.320
Yeah, right. Yeah, it's difficult. And it's Yeah. So So basically, something changed in about the 60s in terms of planetary science, because up to that point, planets were essentially the almost exclusive domain of astronomers, with the exception of the moon, right, because we felt that there the whole time. But until around the 1600s, planets were in, we knew of them. We've no idea about them. We did not know how big they were, we had some ideas how far away they were, but not really, but we certainly know what the surfaces look like, because all we could see were lights in the sky by dots and telescopes. But in the 60s, human beings began to explore the solar system with robotic spacecraft dispatched out to the planets. And that turned, for example, Mars from a point of orange light in the sky, to a world. And we saw impact craters, and we saw volcanoes and we saw a giant canyons and we did the same for Mercury, the same for Venus, and then we did Jupiter and the giant planets.
00:10:51.260 --> 00:11:10.193
And suddenly this, this new discipline came to be, where we were able to actually study these worlds as worlds in their own right. And it's grown since then, right up to the Apollo project in the late 60s, early 70s, where human beings walked on the surface of another planetary body and retrieved samples and brought them home.
00:11:10.246 --> 00:12:13.745
And we've been for the last 50 odd years, been analysing these samples and labs and learning amazing new things as recent is area this year about lunar samples. So the way we do geology to the planets is basically the same except that it just, it's a lot harder, we fly spacecraft to other places, we photograph the surfaces from space, sometimes we land, sometimes we even rove so we're acquiring remotely acquired data, we now have at least for Mars, we have robotic rovers, that are able to basically do field work. One of the one of the geologists basic tools is what's called a hand lens. And you can imagine the physical kind of thing you hold up and you look at the minerals and identify rocks, there is a hand lens imager on the Curiosity rover. And it does the very same thing except instead of a human being looking at the grains and the rock it's beaming that image onto Earth. And then people at home are looking at that and figuring out what the rock is made up. And we have then analogue work with do fieldwork on Earth to look to similar locales, and draw inferences for ways rocks are formed. For example, where the Curiosity rover on Mars is right now it's in a place called Gale crater.
00:12:13.798 --> 00:13:23.309
And it's exploring up the flank of this thing called Mount Sharp is huge mountain in the middle of this crater. And we see layered rocks that form in water. And we know the form of water because of the certain chemical signatures that the rover is able to detect. And because we know that similar looking rock, literally just looks the same forms in water situations like in standing bodies of water in the lake on Earth. And so we're able to go and tell we don't have a direct measure of how old the rock is, we have ways of making inferences about it. We know that rock had... was absolutely laid down under water, which means there was a lake inside Gale Crater probably around three and a half billion years ago. And we're able to do that by simply comparing what we know of Earth to what we're seeing on Mars. So we're doing fieldwork on Mars, we're just using it as a robot, because we haven't put people on Mars, that's going to be that that's just such an expensive thing. It's gonna happen, but it'll probably be 20 years away. So we've had humans do fieldwork on the moon, we've had humans bring samples home from the moon. And we have lots of high resolution data of the moon from orbiters and landers that will help us make sense of what we're seeing, just like we do on Earth, just a lot more expensively.
00:13:23.789 --> 00:13:28.110
Yeah, that's really cool.
00:13:23.789 --> 00:13:47.850
And when you talk about some knowledge of the planets, and what you might think is happening as a world, is that based on some model that you have come up with before an admission is sent to one of these planets. Are you going out there with the hypothesis? Or is that, for example, a purely new discovery that you saw this evidence of a lake on Mars?
00:13:48.259 --> 00:13:50.629
So it depends, it's probably not one size fits all.
00:13:50.659 --> 00:13:56.419
In some cases, if we're going to visit someone for the first time, there are always going to be ideas as to what we will see.
00:13:56.990 --> 00:14:21.740
Often those ideas are wrong, which is good. Like that's how the scientific process works, right. And then those observations beget new hypotheses, which inevitably, then require new data from a subsequent mission, which might be decades away from being flown by these missions are extremely expensive, and they don't happen all that often. And so sometimes, there are missions that are flown with a very specific purpose, a specific thing they want to test.
00:14:21.919 --> 00:14:48.559
Sometimes they're flown simply just to see what the thing looks like, you know, but they'll always be to some extent, people cannot help but begin to speculate what you'll see or how something will work before it gets there. And then they can test their theories. And it's a great way of getting funding. If we have new data, we can test it this way. But But usually, usually there is if every time there is an element of brand new discovery even for a mission that is normally going somewhere to test an existing idea require very specific kind of data.
00:14:48.710 --> 00:14:51.259
There are new discoveries made.
00:14:48.710 --> 00:17:50.539
A good example to illustrate this is there's a mission called Grail missions. NASA flew back 10 years ago, and Grail builds on work from the Grace and Goce mission. Which are flown for Earth to characterise Earth's gravity field. And the long story short, one of the ways in which you can acquire the gravity field of an entire planet is you can fly two spacecraft in a row that has an extremely sensitive radio that can detect how far away they are from each other. And they are broadcasting back to base stations on Earth. And the idea is, if you have already existing which we did good topographic data for the moon, in this case, you can look at the effect of the accelerations on the spacecraft, the distances between them, which will shorten and lengthen depending on what they're flying over. And if you can subtract the effective topography, you can basically resolve you can develop a model for what the gravity of the subsurface is like. And the reason this is important, for example, for the moon is there's a there's a phenomenon geology called a mascot, which is a mass concentration. And it was discovered for the moon because in the early Apollo programme, when they were flying spacecraft over the moon before they landed, they were finding that the spacecraft are accelerating over otherwise by a very small amount, but accelerate nonetheless, over a very smooth otherwise features boring volcanic plain, so called Luna Mary, that dark splotches on the near side of the moon. And why would it be accelerating when there's nothing there's no big mountain there? Right, it turns out, there's an enormous chunk of mantel which is more dense than the regular crustal rock much closer to the surface. And the reason the mantle is there is because that basin has a huge impact basin. And turns out that within the first few few minutes of an impact forming which is 500 kilometres across part of the mantle, which is a bit underneath the crust, which has made it much more iron and magnesium which rock basically rebounds like a like a like a rebounding water droplet up and freezes in that position. So the mantle is closer to the surface, you don't see any evidence of this on the surface, but it's there. And it's gravity signature is felt. And that's what it was accelerating the spacecraft. And so long story short, we now have this unbelievably highly resolved like a degree in order 1200, or something very, very small spatial scale for the moon, of the gravity map of the moon. The mission was going to do this, this was the goal. But one thing we didn't know until the mission was finished that for example, the porosity of the upper surface of the moon is extremely high, which told us that the moon's crust is far more damaged than it sort of looks because of four and a half billion years of being hit by stuff. Yeah, so even though the goal was to acquire this gravity data, on which we're only really now, eight or seven years, since great lenders still beginning to just begin to pick through, we already made amazing discoveries about what was happening inside the moon that we cannot see from the surface. But that gravity mission Grail takes its inspiration, it basically follows the same approach the gravity missions for Earth if you don't wish, which we frequently fly, and update and tell us where there are subsurface concentrations of natural material, or even what the oceans are doing and how the ocean basins are changing in terms of fillers and what the ice caps are doing. So again, that's applying what we know of a geosystem on Earth to another planet.
00:17:51.769 --> 00:18:06.529
That's awesome. Because I've been itching to ask you about Mars and the Mars 2020 mission. And it's amazing how much is still being explored and understood about something as relatively inert as the moon from viewpoint here on Earth.
00:18:07.460 --> 00:18:08.210
It's fascinating.
00:18:08.240 --> 00:18:24.769
And I'll tell you just one thing, the real value of doing this kind of work is not just you know, national prestige, or getting kids into STEM projects, and developing new technologies and spin off patents and stuff.
00:18:20.809 --> 00:18:36.109
And all that's true. But one of the real values scientifically doing this kind of work is that Earth is an amazing world, but it's sort of a pretty bad world to learn geology on because Earth has plate tectonics. Okay.
00:18:36.170 --> 00:18:53.750
And plate tectonics doesn't seem the presence or really, I think in the past to have operated anywhere else. And it means that we recycle the crust the subduction process, which causes huge earthquakes, you know, your listeners probably have heard of plate tectonics and subduction and these massive earthquakes.
00:18:54.259 --> 00:19:01.519
And the oldest oceanic crust on Earth is around 200 million years. That's not very old for a planet 4500 million years old.
00:19:01.970 --> 00:20:21.200
And there are some parts of the crust of the continents that are really, really old, but they're usually slivers, and they're not in good condition. And they're subject to rain, and vegetation and people driving roads through or as on Mars and Mercury in the moon. you preserve that ancient rock record. So a good way to illustrate this is on the bit of a moon that we see those huge big splotches, those big splashes are vast lava planes that's why they look dark and they're inside causes circular things and it's caused by certain things are gigantic impact basins. Some of them are more than 1000 kilometres across. There are similarly sized impact basins on Mars and there are similarly sized impact basins on Mercury is not on Venus. That's a whole different conversation in a different podcast when Apart from this, but the point is what that tells us is Earth most absolutely helping similarly, so affected Earth monster must have had 1000 kilometre diameter impact basins, way back when but they're gone now because we have plate tectonics I mean atmosphere we've oceans So we are able to calibrate and understand the geology for the planets by having access to geology here in Earth. But we are also able to understand Earth ancient rock record, which is essentially lost loss by looking at those preserved records on other planets. So that's one of the real powers of planetary geology and history, science. Generally, we call it comparative planetology. It's not just understanding one word.
00:20:21.259 --> 00:20:36.349
It's how the whole thing fits together into a network. And one of the things we see is, there really, so far hasn't been a process, we've seen another world that we cannot begin to figure out this huge unknowns.
00:20:31.579 --> 00:20:58.759
But we kind of recognise the fundamentals and everything. And there's only a finite number of processes that affect a planetary surface. But the variety and the breadth in which those processes are manifest is far greater than anything on Earth. And what that tells us is, once we understand the basic tune, we can begin to understand the variations in that tune, and how it's manifest differently in different places because of different starting conditions.
00:20:59.240 --> 00:21:01.099
And that's one of the powers of what we do.
00:21:02.119 --> 00:21:12.589
That's amazing. So with regards to Mars, and the most recent mission that went up Perseverance as a rover that's now roaming on the planet Mars.
00:21:09.230 --> 00:21:14.779
Tell me a bit about your experience in seeing that mission and what it meant for you.
00:21:15.259 --> 00:21:39.500
I mean, it was incredible the technological achievement to do what they did. Well, let me just back up, right, so so Perseverance is the second it's a twin to a rover called Curiosity, which we landed on the surface of Mars in August 2012. And Curiosity, Perseverance of the fourth and fifth rovers, respectively, that NASA has landed on Mars. So it started with with rover called mission called Pathfinder and a rover called Sojourner in 1997.
00:21:36.410 --> 00:22:03.289
And Sojourner is the size of like a large microwave oven. And it was, it was essentially a tech demo. It was a proof of concept. Can we do this because we've never we being human beings had never roved on Mars before, we had automatic rovers robotic rovers on the moon, the Russians have done that. But no nation no human ever roved anything on Mars. And the way that Mars represents an event.
00:22:04.460 --> 00:22:49.369
It's a tricky place to land stuff on. If you want to land on Earth, you can come in really fast like think this is how SpaceX doesn't sell the Apollo missions, do it to the Soyuz capsule to come in really fast that heat shield, and then at some altitude, you blow a panel and out comes a chute, the drogue chute and the main chute and you parachute down. If you're the Russians, you do what's called lithobraking, or you slam into the Kazakh steps a few miles an hour. If you're the Americans who landed in the ocean, the point is, you're coming down on chutes, because it's thick atmosphere to do that. Yeah, the extreme of this is Venus. If you want to land on Venus, which the Russians did in the 70s, and 80s, you bring a parachute. So you come in your heat shield coming in, and your hypervelocity speed is coming from faster than orbital speed.
00:22:49.369 --> 00:23:06.289
Usually, you come in, you have a shield, drop all that stuff, and then you have parachute. And then at around 60 kilometres and Venus, you drop the parachute, you cut it off, and you freefall because the atmosphere of Venus is at the surface is 90 times atmospheric pressure on Earth.
00:23:02.930 --> 00:23:58.160
So you don't need a chute, you have a drag plates, you can kind of start the thing spinning too much. And you basically fall off the bat, it's like dropping a probe off the side of a ship onto the ocean floor. So Venus, okay, separately, the temperature pressure will kill you and the sulphuric acid. But it's like dynamically, it's not that hard to get them to surface. Earth got parachutes for, let's say, mercury, which we've never landed on, or the moon, which we have no atmosphere. So you can't use chutes, but you can use retro rockets. And we've done this many times we've been humans, we've had human beings actually fire themselves off the surface in the Apollo programme. We've had a robotic Russian American Chinese landers. And as long as you have the fuel, you can land and you could do a soft landing with a retro rocket and you just land it down. The problem with Mars. You don't have enough of an atmosphere for big chutes, but you have enough of an atmosphere that you can't use rockets all the way down. So it's challenging.
00:23:58.640 --> 00:24:02.180
Okay, yeah, I read that Mars was, like 1% density of
00:24:02.029 --> 00:24:03.859
...at the surface, right.
00:24:02.029 --> 00:24:51.109
But it turns out that if you're Earth's atmosphere coming in at interplanetary speed, you're still that's how that answers Do you. So the way you land on Mars is with difficulty. You come in as normal high speed heat shield, then you have a parachute. But to safely land a lander, you can then drop the chute at a few kilometres and come in on a rocket and that's for example, what the Insight lander has done, which is currently operating on Mars as a seismometer on board. The Phoenix lander in the late 2000s landed in the Northern Hemisphere near the North Pole, it came down on rockets. But if you want to put a rover down, the rover has to somehow come in some sort of like rover base station, which is what Sojourner does, and so it came in on airbags. And the way it came in is it came out a parachute and then some few kilometres up, dropped and then just ballistically fell gravity is a third of what it is on Earth, you're still gonna accelerate.
00:24:48.740 --> 00:25:53.839
And then basically the airbox inflated and this thing looks like a knot is looking like a bunch of Maltese just stuck together and it bounces onto the ground and then it rolls and stuff right and then the airbags deflate and And lo this thing unfolds and outcomes. I mean, that's an amazing technological achievement. The rovers that follow Sojourner, Spirit and Opportunity use the same technology they used airbags although it turns out that if you Google that the images of the airbags for spirit opportunity, these things were huge, maybe like 10 metres tall or something, the whole network of airbags. But what NASA determined was that that's probably so the Spirit and Opportunity rovers are perhaps maybe about two metres by two metres long metre high with their solar panels. They're hefty, but they're the size of I don't know, like deer. So what NASA worked out is that that basically represents the maximum size thing you can learn airbags, okay, and so when it came to putting something like Curiosity, which is the size of a mini cooper on Mars, NASA had to come up with a new way of doing airbags wouldn't cut it.
00:25:50.569 --> 00:26:39.470
And you don't want to come in on rockets all the way down, because that's going to potentially seriously damage the rover and in terms of how you would safely have the rover on board and not getting debris on and stuff. So what they settled on was this idea called sky crane. So you come in and your heat shield and your interplanetary speed. You blow your beauty shield after a while that you drop your back shell after a certain time, basically, politically free falling, and then you come down on rocket. So now you have the rover underneath this thing called the descent stage, which is nicknamed sky crane and it comes down basically like a jet rocket pack. And then when it's hovering in about 10 metres over the ground, it starts to drop, lower down the rover on cables, which is now deploying its wheels. It continues to descend slowly. And then when you get wait on wheels on all six wheels and a bunch of other criteria.
00:26:40.220 --> 00:26:58.940
Then those cables are snapped by guillotines, and the desensitise rockets away. And that all happened and this all has to happen in about 767 minutes from the moment you get atmospheric interface to getting to the surface of Mars. And this all had to be completely automated.
00:26:54.950 --> 00:27:05.000
And they did this in 2012. And they did it again, in 2021.
00:26:58.940 --> 00:27:31.430
Perseverance. The biggest difference between what happened 2012 and 2021 is that normally, when you are landing on the surface of any planetary body, you have what's called a landing ellipse that you've got to try and land in. Landing ellips basically is an area o uncertainty, you are going t land somewhere in that ellipse And the size of the ellipse, and the aspect ratio of the llipse represents a whole ile of things, the features f this uncertainty model.
00:27:28.039 --> 00:29:50.430
ut how exactly fast are yo going because you don't have a speedometer on a spacecr ft, we have to use things lik radar tracking for grad class s. And the way the Russians did t so like basically, when y u land on Venus, to this day so... okay, America's never la ded on Venus, this is a kind of a sore spot for a lot of us. B t the way the Russians did it as, it was ballistic. Like onc you set this thing going there's very, very litt e attitude, there's some attitud control, but there's very lit le propulsive control in ter s of actually steering or drivi g this thing. Yeah. And what t at meant was, the Russians deci ed roughly where they wanted to and and had a very large err r ellipse, landing ellipse, and then ba To this day, although if you go to Wikipedia, for example, it'll tell you where the Venera landers are, it gives us I think it quotes it's like one decimal place in terms of lat long decimal degrees. We don't actually know where they are, we have a rough idea, because because of the accuracy of the, the radio tracking, but not to the point where we could say it's exactly, you know, 40 metres from here, you know, we've never seen these things from orbit because you never had the resolution to see them. If we could work out where the Venera landers were, because we have photographs from four of those landers, of their landing sites in terms of the terrain and not big landmarks, as you can see that far and the way the cameras worked. But if we could understand from the orbital scale, and this is a perennial issue we have everywhere, if you can, and one of the reasons why the rovers are so powerful at Mars is we are now able in the case of Mars to go from the grain scale to the planetary scale. Completely uninterrupted to continuum now from the hand lens imager I talked about, to images of aircraft from the rover to high resolution images of the rover to low resolution, huge regional scale maps. But if we had we had it for the moon to in a few places, and again, from Mars in only a few places. But if we had that for Venus, for example, we'd be able to say okay, now we know what radar data because you have these radar to see thick atmosphere Venus, because it's opaque to visible light, we will be able to say, Okay, now we know where this thing is. And we know what the land looks like, on radar at this scale, because that's what we've photographed at this scale. And we know exactly how to relate the two and then we can use, we can look for see exactly similar dielectric, or backscatter properties and the other radar that we have for the rest of the planet and start to work out what the surface might look like. We can't do that yet.
00:29:51.029 --> 00:29:51.450
So...
00:29:52.140 --> 00:30:17.970
So when you're talking there about how they land on the planet in the first place, part of what I read about the Mars 2020 mission was that it was a whole new demonstration of what they referred to as the Terrain Relative Navigation - TRN. Have you seen that used before? or what does it mean to use now for future missions, because you're saying this is going to revolutionise how we land on any given planetary body? Is that true?
00:30:18.319 --> 00:33:59.059
Yeah, it is. or it has the potential to be. So no, it's never been used before. This is brand new. And there's a very specific reason why it was developed. So when we land on planetary bodies, normally, we have what's called a landing ellipse, which is basically an area that shows the uncertainties where this thing is gonna land. And the aspect ratio of that ellipse, the general size of it, that all reflects a bunch of different factors, like the exact moment of atmospheric entry. Depending on the time of day the atmosphere will have different heights if you live different scale heights, and so you might encounter drag sooner or later or all kinds of the factors right. Now, if you look at the size of the landing ellipse of Mars missions over the past few decades, from say the Viking landers in the 70s right through to today, those ellipses were originally hundreds of kilometres long and they've been shrinking down. Now, Curiosity, Perseverance's twin the land in 2012 had up to that point, the smallest landing ellipse ever for a Mars lander. And the reason was because it had the ability to, to a limited extent steer itself during atmospheric entry. It had a rough idea where it was where is open at that point, historically, it would only ever be a ballistic entry, which means it's in the hands of the gods and luck and fate. This thing had the ability to a limited extent to basically control the entry corridor. And that meant that engineers could could then work with a relatively small around the left because the area of uncertainty as to where it would land was proportionally smaller. That meant that Curiosity was able to land in what had been up to that point, a completely unrealistic and forbidding place. It is inside a crater called Gale crater. And on the on the edge of the crater, you have these big walls tower of a couple of kilometers. And in the middle of Gale Crater, you have this mountain called Mount Sharp, which is where we have these layered rocks. And that is one of the most interesting places you could go and look at the history of climate on Mars, but landed between a big cliff of the crater wall and a huge mountain. No engineer will normally say that's okay. But there's a compelling science reason to land there. That's what led engineers to develop the ability for Curiosity to steer itself a little bit on the way in to shrink the landing ellipse enough to give mission managers confidence that it would safely land in the flat area between the cliff and the mountain. For Perseverance, it has a very specific purpose, which is basically it is setting the stage for a much larger, longer and more ambitious project called Mars operator for it to do this. Ultimately, mission managers in the science team and the science community generally identified a particular place in Jezero Crater, where we think we see evidence of a remnant ancient delta, basically a fossilised remains of a river delta, where we think a river was carrying sediment into a standing body, this crater existed already, it was a standing body of water, it was a lake. And this stuff just came in and overtime built up now, those kinds of settings on Earth are rife with bacterial life. And if we are ever going to find evidence of potentially ancient fossilised microbial life on Mars, one of the foundational questions we have is, are we alone? And if so, why? And if not, why not? One of the most exciting places you can go was to look at these kinds of rocks and look for that evidence of fossilised life is in something like the layered rocks in Jezero Crater. The problem is that the actual area you want to get to near near the delta is an extremely difficult place to land. And if you're relying solely on ballistics, and even simply on that little entry interface thing you could do, where you can kind of see yourself during your entry sequence. Your landing ellipse encompasses parts of the crater wall parts of the much ruggeder hummocky land beyond the delta and parts of the Delta itself.
00:33:59.420 --> 00:35:12.590
There's just no way you're going to risk a two and a half billion dollar rover with something as dangerous as that you're just not. That requires, driven by science, always fundamentally driven by science requirements that required folks at the Jet Propulsion Laboratory, building Perseverance, to develop a new way of helping the rover identify a place and land itself that will ensure a safe landing on the surface of Mars. Now, because of the relative position of Earth and Mars, it's something of order about seven light seconds and Perseverance made its made its landing. It depends on where Mars and Earth are relative to the Sun and each other. There was no prospect for having what we call ground in the loop, you cannot have the controller in mission control with a joystick because of the light delay. By the time they send a signal thing is already on the ground. It takes about seven minutes to get a signal from Mars. It takes about seven minutes to reach from the top of the atmosphere to the ground. So by the time we started to get the first signals back in during landing day for Perseverance on February 18th, but the timing of those first photons back saying, hey, I've entered the atmosphere, the rover was physically on the ground now, the rover was always going to be on the ground. The question is whether the rover was going to be in a little crater, bits of debris everywhere or physically in tact, right. And people have at that point, it's, it's, it's up to the computer.
00:35:13.130 --> 00:35:14.780
Yeah. But happily It was the latter.
00:35:14.000 --> 00:36:43.460
It was to be clear it landed, it landed. And that's where the Terrain Relative Navigation system comes in. So basically, long story short, because of the quality data we have from multiple orbiters from decades at this point in orbit of Mars, we were able to make a really high resolution topographic map of the area of interest in Jezero Crater. And then that map was uploaded to the computer's brain and the computer in the spacecraft and the rover, which is doing all the controlling. It wasn't picking an arbitrary site on its own. It had this map in its head. But what was amazing with the Terrain, Relative Navigation was it's called relative. It's because it has this pre loaded map in his head saying, here are the safe bits. Here are the slightly riskier bits here, the bits you must avoid at all costs. And then what it's doing is during that last sentence coming down on the sky crane and this jetpack, it is acquiring data with LIDAR, and it is processing those data in real time and comparing them to the pre loaded map. And then it is making its best decision as a function of where it happens to find itself to direct itself with the vector thrust on the descent stage to go and move the rover to an area it designates as safe, which means the mission engineers knew roughly who would be but they did not know exactly who it would be until it made the decision. And it was the computer making the decision for itself as to where to safely put down and if you go on, you can Google the image of the Terrain Relative Navigation map where you can see where the seat touched down, it landed in this little sliver of perfectly safe surrounded on either side by no go. I mean, the computer did it perfectly.
00:36:43.730 --> 00:36:47.210
Wow. And those no go areas would have been in its ellipse.
00:36:47.449 --> 00:36:51.559
They Absolutely.
00:36:47.449 --> 00:37:04.820
Absolutely. Yes, there were much of its ellipse, it could not land in safely either because the landing itself would damage the rover. Or worse, the rover might land safely but be unable to drive out of that locale and just become base station.
00:37:02.000 --> 00:37:28.250
Because it wasn't just about landing safely It was about them being able to be mobile and move thereafter. Terrain Relative Navigation system executed perfectly. Now, the reason this matters for future missions is because this now opens the door to other robotic uncrewed non human in the loop or ground in the loop landing systems. Let's say for example, we want to land on the surface of Europa, which is an idea that's been kicked around there isn't money for it yet. But But NASA wants to land on Europa
00:37:28.699 --> 00:37:29.719
One of Jupiter's moons.
00:37:29.750 --> 00:37:55.820
One of Jupiter's moons, Europa is a fascinating place it is it's got an outer crust instead of rock, it's ice, there is a strong evidence, we suspect that there's a liquid water ocean below And then below that, again, it's rock. The kinds of environments where for example, we know life may have started on Earth and say hydrothermal vents may be present on Europa not we don't know for sure, maybe.
00:37:51.559 --> 00:38:38.570
However, we don't know what the surface of Europa looks like right now, at the scale of a lander. This is an issue we often have. And it was an issue that NASA had in the 70s with the Viking landers, we did not have the resolution to know where we'll be safe at the scale of a lander, because the images we would have from orbit were much coarser than that thanks to for example, certain instruments, we now have on the moon and Mars in orbit, we now have images and other data, it was a surface roughness of the surface at the scale of the lander, but we don't have it for most places. Now. NASA has a mission called Europa Clipper under development and Clipper is due to launch mid this decade, possibly on a Falcon Nine Falcon Heavy, and it will get towards the end of this decades to Europa, it won't orbit Europe, it'll actually orbit Jupiter.
00:38:35.449 --> 00:40:02.480
But it will do many flybys of this. Jupiter has an unbelievably strong magnetic field and very, very high radiation environment. And it just wrecks instruments and electronics. So they've had to be great creative and how they devise this flyby. Hence, it's called Clipper. But But the goal of one of the goals one of the many goals of Europa Clipper is to resolve the surface of Europa in decent small areas, at scales relevant to a lander. Once we have those data, there's no reason we can't then process them in the same way we did for Mars, feed them into the landers computer, and then using a very similar approach with this sort of rocket jetpack basically look, ultimately, to drop something off the surface of Europa, again, with humans not in the loop allowing the computer to make the decision for the safe place to go is depending on what you fed into in the first place. So I'm sure there there are scenarios that I haven't even thought of you know Mercury, more landers on Mars landers on the moon. Once you have the ability of telling a computer, here's what you what the landing site looks like, here's where we want you to go, here's what we don't want you to go. Now you have the capability of acquiring and processing data for comparison, in real time during the descent, when there is no second chance. Yeah, we've already validated that technology, which means we will see it fly again. And that is going to open up at the very least it shrinks landing ellipses. Yeah, which means that we can now be more selective on where we want to go instead of saying we have to target this boring, vast plane of lava.
00:39:58.460 --> 00:41:04.400
Because we know at 100 kilometer length scale, it's safe. Now we could say let's go near that really interesting, weird outcrop we don't know what the mineralogy for space is, let's go land there. That's what things like Terrain Relative Navigation, the rover itself has some enhanced autonomy in where it's able to drive now It never drives on its own. And again, because the light light, there's no human with a joystick, or mouse, it's all command lines that are uploaded in the command sequence and the thing moves and it takes it's kind of take all these images it beams the images home, which is how the each day the rover operations teams can see you know, where it's moved is where they think it should be with the right direction. But there are definitely ways the rover has now of looking at us landscape using its navigation and its hazard cameras to basically make informed decisions to the team as to where it thinks is safe. That's a level of autonomy we haven't had before, we're gonna continue seeing that kind of autonomy developed and enhanced for subsequent missions. All of it in service ultimately, of more precise, more accurate landing and, and surface operations on planetary bodies such as Mars.
00:41:01.909 --> 00:41:57.650
And this gets the whole idea of Mars sample return, because what's going to happen is, and one of the goals of Perseverance is to collect little vials of soil and rock, it has a drill to look and take cores. And ultimately, through a very complicated, convoluted and expensive process, return those samples to work. So if you can imagine the amount of work it's going to take in the coordination it's going to take to get the samples, deposit them in particular places. The goal is later to have what's called a fetch rover land near Perseverance, which again, will require unprecedented accuracy, because you don't want to land 300 kilometres away, because the rover will never last 300 kilometres, right, it's probably won't, right. So you want to land nearby, fetch rover will come up and collect these little caches, and basically, put them on board. And ultimately, it will collect all the cache, all little samples, like that Perseverance has been kind of pooping out around the place.
00:41:58.369 --> 00:42:49.519
And these little vials, it's very complicated sampling system. Fetch rover is then going to bring it back to the thing it landed on, which now has a rocket of its own called the Mars ascent vehicle. It's going to basically this is all autonomous again, yeah, I mean, you can programme I put it, there's no human doing this in real time. It takes these little samples, these vials this kind of module or capsule, puts it in to the rocket, which then closes up the rocket then fires and goes into Mars orbit. Then a third element of this entire mission, which is basically I think it's called the Earth Return orbiter, which will be the largest orbiter ever flown to Mars, because it's not just going to Mars like every other orbit, it's coming home. Yeah, the Earth Return orbiter will arrive into Mars orbit, spend while there, checking its systems out. And it's then going to rendezvous with this capsule, which is now an orbit of Mars.
00:42:45.409 --> 00:42:56.420
Grab the capsule. Imagine how difficult is to say get the spacial space station I know the Soyuz capsule dragon, right?
00:42:51.920 --> 00:43:16.730
This is an autonomous robot and in different planets. Finally a capsule besides probably about a football, it grabs this thing, ingests it, then turns around fires his thruster to leave Mars, which never had a spacecraft have to leave Mars orbit, before it'll fly back to Earth. And then it will let the capsule go the capsule in a reentry system. It'll come in.
00:43:13.730 --> 00:43:26.960
We get the samples off the surface. I mean, it's the amount of technological the challenges that haven't even been encountered yet. Yeah. All in service of getting samples back from Mars to Earth.
00:43:28.429 --> 00:43:50.389
It's absolutely fascinating. it's mind blowing, it really is, is a phenomenal endeavour. And I purely cannot fathom the amount of like nerves that must be involved when you see such a feat of engineering and technological advancement in the exploration of another planetary body. It's just
00:43:50.000 --> 00:43:50.840
It's stunning. It's stunning. I mean, we know we amazing.
00:43:54.260 --> 00:44:27.769
weren't around for the Apollo missions we will be around I hope for Artemis or whatever replaces Artemis when we see humans back on the moon. But until then, this is this is as close as we can get to see the the amazing technological achievement to put something on the surface of Mars with the accuracy and precision that Perseverance to see the images your listeners can go if you Google Perseverance raw images and Curiosity raw images. NASA uploads raw data before it really any processing. So basic processing every day, you can go and then upload onto every day, depending on what the rovers do.
00:44:28.039 --> 00:44:43.699
But you can see these images that come down unfiltered, you can go and modify and play with them. You can see what it's seeing every day. And it's starting you can you know, what was it like today? It was cloudy on Mars today. You see that picture? Yeah, to live in the er where we have that is just a
00:44:44.409 --> 00:44:54.610
It brings it home. It's it makes it so real. Yeah. Paul, that has been really, really amazing to hear. Thanks very much for joining us at the DataCafe today.
00:44:54.659 --> 00:44:56.909
I absolutely enjoyed it.
00:44:54.659 --> 00:44:56.909
Thank you, Jason.
00:45:01.420 --> 00:45:22.090
Wow. So I think that's that's a really impressive challenge, then to get this set of data off of another planet, I think that brings it home dramatically how one investment in technology in engineering.
00:45:17.019 --> 00:45:36.909
And then you've got this data, you know, either through the experiments that that you've been running or are running, and then the experiments that you will run in the future on all of these samples, potentially. So what an extraordinary story what an amazing application for data science to be involved with.
00:45:37.300 --> 00:45:47.530
Yeah, even when you mentioned the ongoing work on lunar samples. How much more can there be to do with a rock from the moon? Just
00:45:47.900 --> 00:45:55.579
When were we last on the moon? In the 70s, maybe I think just before I was, yeah, born.
00:45:51.889 --> 00:45:56.269
So that's, and yet it's still going strong?
00:45:56.300 --> 00:46:14.719
Exactly. And I think it's because of the advances in scientific techniques in computational power in modelling the cutting edge things that we employ, as data scientists in a lot of our areas are being applied to the same materials that have come back from other worlds, which is really impressive to me.
00:46:15.050 --> 00:46:25.730
And in fact, Paul talks about this, this this field, I'd never heard of comparative planetology, which I thought, but but when he explained it, I thought, That's fantastic.
00:46:25.730 --> 00:46:45.619
That's exactly what data scientists do. They try and infer statistical patterns from one set of data, and then try and apply it to another scenario where they think there is this commonality. And that's exactly what they're doing between not just within an industrial application area, but between planets. Wow!
00:46:45.650 --> 00:46:56.869
Yeah. I mean, I'm, I'm laughing, because I'm thinking, are they setting up an A B test between two planets? Because that can't be happening, right?
00:46:53.420 --> 00:47:07.460
Fundamentally, yeah, this is what we're doing are comparing planets and the discoveries from the observations of plants, because again, all of these observations are only samples.
00:47:03.139 --> 00:47:22.190
And they're very limited in, you know, the extent to which we can get an overall picture of what it is we're observing, which is typical of every science, everything you do is as simple whether it's something out of really miniscule level or a global level. And here we are doing it with planets.
00:47:22.000 --> 00:47:56.920
Yeah. And when you when you have enough, though, when you have enough data that you can, with confidence, say confidence, I have seen that, you know, set of striations on a rock before on planet Earth. And I can tell you with a level of confidence with a certain probability, I guess, that there has been water formation on that rock were involved with the creation and the moulding of that rock. And I'm transferring that knowledge from one one environment to another, nonetheless, able to make that kind of connection. I think that's that's just spectacular.
00:47:56.980 --> 00:48:05.260
Yeah. And I was, again, mind blown when Paul said Earth was a bad place to yes, these observations. I mean, no,
00:48:05.260 --> 00:48:20.980
Yes. Not not a great place to do geology. We're really struck home. And so the idea that this was it 4.2 billion years of basically missing data on Earth, because we have these tectonic plates, which just suck everything down and destroy it and recycle it.
00:48:21.010 --> 00:48:22.539
Yeah. That was amazing.
00:48:22.750 --> 00:48:34.750
Again, you know, it's such a classic data scientist problem, how do you how do you cope with missing data? Well, you find another comparable setting? And you say, Oh, well, look, I've got pictures of rocks,
00:48:34.960 --> 00:48:36.130
like a time series.
00:48:36.570 --> 00:48:43.289
Right. And I can use that I can use that, to discover the pattern that I need, and then patch the missing data that I haven't got.
00:48:43.289 --> 00:48:49.679
Exactly yeah, Earth's time series is botched or incomplete. Yes. Points in its history.
00:48:49.739 --> 00:49:02.219
You know, there was so much that came out of this from from the data science perspective, the fact that we want to, you know, show with this mission, so we're doing hypothesis driven approach.
00:48:58.679 --> 00:49:11.969
We're pivoting when we discover that our assumptions, which often are wrong, right out of the box, but you know, allows us to go Oh, wow, that's completely not true. So what is true? What can we learn from this, this data that we're collecting.
00:49:12.570 --> 00:50:02.820
The demonstrations of technology as well off the back of that, that he mentioned, everything is led by the science, but enabled very much so by the technology and we see it in the rover technology that you read about Steve and made the wheels more sturdy, they've strengthened its arm for the camera, they've changed how it senses its environment and they've put together this caching You know, when we were talking about geocaching, you know the idea of going to certain coordinates and picking up a sample. It's a fun hiking activity that people do here on Earth. Paul talked about a Mars return mission where they'll send up another robotic spacecraft and go and retrieve the samples that the current Perseverance is putting on surface for for future science.
00:50:02.900 --> 00:50:24.710
Yeah. So I was trying to connect how they were getting how long it was going to take to get these rocks back because it's required this sort of complex synchronisation of two, or even three, missions to bring these samples back to Earth. And it's, it's at least seven months, isn't it to bring to do a one way mission back from Mars. So so you know, this is this is a...
00:50:24.710 --> 00:50:25.969
Depending where the planet is. Yeah.
00:50:26.000 --> 00:50:32.599
Right. Of course it does. Yes, yeah. So I asked quite a trek, just just to get just to get some rocks to be able to do the experiments with.
00:50:32.889 --> 00:51:07.329
Yeah. And that alignment of missions is really impressive, I think, in my readings that this Mars Exploration programme is planned up until 2026. And what really impressed me at that point was it will have been going for 30 years, covering a number of missions, I mean, you can read up on them online. And beyond that, I don't know what's proposed, what I presume that a lot of the learnings of the back of a mission like this will then enable those proposals and the funding and the additional missions.
00:51:08.289 --> 00:51:40.449
It drives a huge amount of excitement, it drives a huge amount of just high quality scientific publicity. But but most of all, it drives by necessity, development in the underlying technologies that, you know, are required to just just perform this extraordinary feat, that I mean, I was really taken by just the landing technology that was needed and the development along the way that Paul was talking about to be able to do that.
00:51:40.510 --> 00:51:59.949
That whole landing ellipse idea. I mean, it just speaks to me so much as a scientist, just generally, that you have to have transparency on your certainty or uncertainty as the case always is, and an error bar to quantify that somehow.
00:52:00.489 --> 00:52:20.769
And this is what our landing ellipse is when he's talking about these missions, you know, saying, we know it's going to land somewhere here. But how big is that ellipse? And how small can we get it with all of these advancements in technology? And yeah, I think you're alluding to the Terrain, Relative Navigation system right?
00:52:20.829 --> 00:52:26.949
Yes, that blew my mind, that was just such a phenomenal piece of technology.
00:52:27.050 --> 00:53:31.579
It did this wonderful thing for me of saying, and, of course, there's no Sat Nav on Mars, they are very specific satellites around Earth, you can't just put a whole infrastructure of Sat Nav on Mars, you know, it's so it just, there's no silly moments where I realise what I'm taking for granted on my phone in my pocket, yeah, and the amazing technology that we have, and then we send that off to another planet, and it's on its own. And we're sending up batches of commands. And what we need to do is build in a level of autonomy, that gives us confidence that our ellipse is manageable, that our lander is going to be successful. And that TRN, it speaks to me for all the advances in computer vision in AI. And we talked about self driving cars, you know, ideas of bringing in sensors that allow your machine to learn, hence the name machine learning and make decisions.
00:53:31.960 --> 00:54:17.710
It was a really decent example, a really powerful example of a reinforcement learning agent taken to the sort of the next level if you like, I mean, this is a mission, a spacecraft that has got to, it's been given an initial best guess, best scan of the terrain of the layout, the geography of the planet that it's going to, but it's recognised that that's only okay at a certain resolution. And when the mission is descending through the atmosphere, it can then have a better view on where it's going than any prior map or prior prior understanding. So it's got to update its knowledge, doesn't it?
00:54:17.840 --> 00:54:49.159
And imagine how little stake it has. It's not alive, like when they talk about landing on the moon and looking at the window to make sure that you avoid the area that's, you know, problematic to land in. I mean, I'm an astronaut. My health is on the line. This is a hazard to me, you know, to land somewhere wrong. This is a machine we need to invest the capability in the machine to now make a decision. That means all of our hopes and aspirations on the back of it are not dashed.
00:54:49.570 --> 00:55:07.420
Absolutely. Famously, Neil Armstrong had to adjust the landing position of the first lunar mission. Because he could see that that where they had originally thought it was going to land wasn't going to work.
00:55:03.610 --> 00:55:24.730
And there's this fantastic tape you can listen to where you know that you can hear them playing the amount of fuel that I've got left as a as a running metric so that absolutely everybody in Mission Control and on on Apollo can hear it. Say, you know, this is this is how much time you've got to play with to make sure this mission ends well, Neil.
00:55:24.760 --> 00:55:26.289
My hearts go and just talking about it.
00:55:27.019 --> 00:55:36.079
I know, and yet, so but this thing's doing it without human intervention, it cannot have human intervention. Because of the light distance from from Mars to Earth.
00:55:36.260 --> 00:55:55.070
We almost forget sometimes just how far away these things are. It is a vast expanse of empty space, where the signal? I think Paul, Paul mentioned that was seven minutes. I mean, yes, that's a huge wait to be sending batches of code up and not knowing has the outcome worked in the way that you thought?
00:55:55.570 --> 00:56:02.289
I think he said by the time you think it's entered the atmosphere, it's already landed.
00:55:57.940 --> 00:56:09.099
I love that. And then it's going down. It's applying this fantastic reward function, not just as oh, yeah. Am I going to be successful in landing here?
00:56:09.099 --> 00:56:13.480
And now I've got this refined model of the terrain beneath me.
00:56:13.480 --> 00:56:24.460
But also, am I going to be able to navigate a rover out of there, which is like that's quite a complex reward function, you're having to apply to what already a pretty high risk and real time reactive system.
00:56:25.300 --> 00:56:27.280
Yeah, yeah, very much.
00:56:25.300 --> 00:56:27.280
So.
00:56:27.610 --> 00:56:39.460
I love that I thought that was that was such a priceless piece of science and data science that they must have been through. So the amount of testing they must have had to have done on that must be phenomenal.
00:56:39.929 --> 00:57:01.679
Yeah, there's some lovely interviews with the engineers and technicians that were involved in the build and the proof of concepts here on Earth initially, as part of the whole proposal and demonstration, the tech that was needed to achieve it. And, and well, yeah, they achieved the same hearing Paul say, it did it perfectly.
00:56:56.849 --> 00:57:02.070
There's no better proof than that.
00:57:02.639 --> 00:57:14.460
You used a word in the middle of the interview, Jason, which I think perfectly sort of summarises this, which is, you know, what a terrific endeavour.
00:57:10.650 --> 00:57:20.699
That, for me sort of summed it up. And I think maybe, in the future, when we're all driving our self driving cars and sort of
00:57:20.730 --> 00:57:25.320
Not driving as the case may be. All right. Sitting back,
00:57:26.250 --> 00:57:27.150
Hence the name.
00:57:27.690 --> 00:57:29.039
Listening to a podcast,
00:57:29.610 --> 00:57:46.590
Listening to you listening to the 800 and 72nd episode of DataCafe. I think we'll be we'll be relying on the sort of technology that's come out of the Perseverance mission and what an extraordinary effort that is an endeavour that is to have got there.
00:57:46.630 --> 00:58:07.210
And it touches upon such beautiful messages for the importance of space exploration that just gets me so excited by everything that we're doing, right at the point of sitting down and writing some code, knowing that somewhere some piece of code that somebody wrote, is driving a rover on another planet.
00:58:09.309 --> 00:58:26.500
Thanks for joining us today at the DataCafe. You can like and review this on iTunes or your preferred podcast provider. Or if you'd like to get in touch, you can email Jason at datacafe.uk or Jeremy at datacafe.uk or on Twitter at datacafepodcast. We'd love to hear your suggestions for future episodes.