Saturn's rings are dense. Sending one flagship probe through them makes no sense. It'll most probably be lost. But sending a swarm of small cheap probes does. Even if a lot of them won't survive, you're still getting lots of data. This is exactly what this NIAC project is proposing.
00:00 Intro 01:22 Observing Saturn's rings 05:21 A swarm of satellites 21:56 Current obsessions 23:36 Final thoughts
📰 GUIDE TO SPACE NEWSLETTER Read by 70,000 people every Friday. Written by Fraser. No ads. Subscribe for Free: https://universetoday.com/newsletter
⚖️ LICENSE Creative Commons Attribution 4.0 International (CC BY 4.0)
You are free to use my work for any purpose you like, just mention me as the source and link back to this video.
AI Generated·May contain errors
Video Summary
AI Generated
Dr. Mike Rubenstein discusses a NASA NIAC award-winning proposal to explore Saturn's rings using a swarm of 10,000 \"femtosatellites.\" By embracing \"robot minimalism\" and fault tolerance, the mission aims to gather unprecedented in situ data by deploying thousands of simple, low-cost units, accepting high loss rates in exchange for massive spatial data collection and system redundancy.
When NASA's Cassini spacecraft went to Saturn, it gave us close up images of Saturn's moons,
and it took pictures of the rings from a safe distance.
And the reality is that if you tried to actually fly through the rings, it would be a very
dangerous experience.
It would probably have destroyed Cassini if it actually tried to make that journey.
And yet, we want to learn more about those rings.
My guest today is Dr. Mike Rubenstein.
He is a professor of computer science and mechanical engineering at Northwestern University.
And he is the recipient of a NASA NIAC 2026 award for a mission called the actively steerable
femtosat constellation for in situ exploration of Saturn's rings, atmosphere and magnetosphere.
This is an incredible proposal for flying a fleet of 10,000 tiny satellites that would
aerobrake through the atmosphere of Saturn and then fly through the rings of Saturn.
And the rings are so dense that most of the spacecraft would be lost with every fly through
of the rings.
And yet, if you send 10,000, then you will bring together some really interesting data
and take some pictures of the rings from the inside.
So if this idea fascinates you, enjoy this conversation with Dr. Mike Rubenstein.
Mike, first, congratulations on winning a 2026 NASA NIAC grant.
Is this your first?
This is my first.
Yes.
Yeah.
And this idea is so cool.
So when I saw this one, this was like, I really hope Mike gets back to me.
Come on.
I want to be able to interview you on this one.
So so your idea is to observe Saturn's rings with a swarm of satellites.
Tell me about the idea.
So the idea is to instead of sending like a traditional, you know, flagship satellite
like Cassini or something like that, which is, you know, a really good science platform,
but it's also because it's a single satellite, it's they're kind of trying to avoid risk
of that satellite.
Right.
So you don't want to send it anywhere near.
There's a chance of like a particle hitting it or anything like that.
So we're looking at some advantages of swarms.
And one advantage of swarms is that they're fault tolerant and they don't really rely
on any individual working at all times.
So we asked, like, maybe we could send them in a dangerous spot.
What's the most dangerous spot that's scientifically interesting?
And I think the rings of Saturn are pretty interesting.
So the idea is that we accept the fact that if we go through the rings of Saturn or something
at orbital speeds, you're going to crash into particles and lose them, the satellites.
But if you have enough of them, it's OK.
Losing half the satellites as they go through the rings, you still have many satellites
after that.
Now, you have not begun the research yet, and I'm sure you are sort of still bringing
yourself up to speed on the environment around Saturn and the kind of density of the particles.
But have you got a sense of of just like how dangerous that environment is around Saturn?
We look mostly at the just the event of going through the rings of Saturn.
So looking at existing data about the particle distribution within different rings and then
given the size of our satellite, we know, like, what is the chance of a satellite being
lost for different parts of the ring?
So like the A ring, yeah, the A ring, which is kind of the outer part of the major rings,
you lose about half the satellites as you go through the ring.
The B ring.
Wow.
The B ring, you lose 92% of the satellites, Cassini gap, you lose like 30% of the satellites.
One pass through the B ring, you lose 90% of your satellites.
Yeah.
So B ring is very dense compared to A ring.
And even if you have very like our satellites are very thin and small.
So they have a very small surface area or frontal surface area going through the rings.
But even going through the B ring is too dense for them.
Yeah.
It's interesting.
I think people don't realize just how thin the rings are, that they're in some cases
10 meters thick, like the height of a house is the thickness of the rings that you would
just be above the rings.
Now you're in there.
Now you're out the other side.
But that density of part, I never thought that the density of particles was, was that
high.
I would have figured it would be more like the asteroid belt.
I mean, not quite as loose as the, in that case, you've got hundreds, thousands, sometimes
even millions of kilometers between objects that you, you would have to go out of your
way to hit one.
But you're saying that 90%, like what are the particles that you're, that you're going
to be crashing into?
Are they going to be like grains of sand?
Yeah.
We're assuming that, you know, at the speeds that we're traveling, like a grain of sand
would destroy the satellite.
So, you know, my understanding is that, you know, there's a, there's a distribution in
density and there's lots of very small particles probably down to like, you know, beyond the
size of the small in the state of a grain of sand.
And we just assume based on the day we saw, like if you hit anything, you lose the satellite.
Maybe that's not the case, but it was, we're just assuming that for safety, you know?
Yeah.
But, but it's funny because like, even this question, like you don't have the answers
to the questions, which is why we need a spacecraft.
And yet no spacecraft is willing to go anywhere near that sort of spinning blades of death.
So, so it's like very self referential.
All right.
So let's talk about your idea then, like what would these, uh, you're calling them femto
satellites.
What would they look like?
So they're kind of like, what is the simplest satellite you could build?
Can you make it as small as possible?
Uh, it's basically like a PCB with a motor on it, a solar cell and a battery and radio.
Uh, so what's her, what's a PCB printed circuit board, a little electronic board.
Usually they're green, like inside your cell phone or whatever.
There's a bunch of electric components built together on a circuit.
And so how small do you think you could get this down to?
We're estimating it's about 10 centimeters by five by two.
So it's, uh, you know, a little bit bigger than a business card.
Um, right.
I mean, that sort of fits into like a fraction of the, uh, like a CubeSat, like the CubeSats
are 10 by 10 by 10 centimeters.
So that is, I forget, they, they're like a fifth of a CubeSat U or something.
I forget.
There's a, there's a standard.
I think 10 by 10 by 10 is a one U.
I'm not sure.
It's one U.
Yeah.
And so you're a slice of a U. Yeah.
So I mean, that is, you can have, you know, a cubic meter, these things, uh, they weigh
like all together.
They weighed maybe 300 kilograms takes a cubic meter and you have 10,000 of these satellites
in a cubic meter.
Wow.
So you, you know, by making it very small and simple, you could build lots of them that
don't weigh too much.
Right.
Uh, right.
So one of the challenges of course, is when you're out at the orbit of, of Saturn, the
amount of energy you can get from solar power is like one, one hundredth, one twenty-fifth.
I forget the number, but it's, it's a fraction of what you get at earth.
Uh, is there enough power to get, like, is that, is the solar panel, the, the primary
thing you're having to deal with?
I mean, that's, that's part of what we're looking at in the study is trying to kind
of go from ballpark estimates to like understanding, like, you know, a little bit more with more
certainty.
Can you power something with a solar cell?
Can it survive the temperature?
Can it survive the radiation environment?
Um, those are all things that we think may be possible.
It's not like completely out of the realm of possibility and we just have to kind of
dive into the details and understand that a little bit more.
Right.
This is what the paper is.
This is what the research is going to be.
And so then one of the other things that you mentioned in your initial proposal was that
there would be some kind of propulsion system.
Have you thought about what that might look like?
So there, so there's no propulsion system.
Uh, they use aerobraking for, uh, maneuvering in, in orbit.
So the initial orbit starts, uh, kind of the lowest part of the orbit just touches the
outer atmosphere of Saturn and the highest part of the orbit goes a little bit past the
A ring.
And then when you go through the atmosphere of Saturn, you actually lower your, the point
where you cross the ring of Saturn.
So we can control the drag on these satellites through the aerobraking to control like where
they touch the ring of Saturn when they pass through them.
So satellites have, uh, one motor and they have two kind of passive directions that get
past the rings after through that atmosphere.
One is like a face on, so a very high drag configuration and one is edge on, so very
low drag configuration.
So you can control how much drag the satellite has during that, you know, aerobraking and
that allows you to control where it goes through the rings of Saturn.
Right.
Okay.
So take me through the whole story.
You are, uh, you've handed off your meter cube of femto satellites and it's been packed
atop a rocket and the rocket blasts off on the way to Saturn.
How, how does the rest of the mission unfold?
Uh, so I think it, nothing really happens on the satellites themselves until you release
them in this initial orbit.
And then, uh, they immediately start taking passes through the rings of Saturn.
Uh, every, every time they go through the atmosphere, they slow down to control where
they go and they make another pass through it.
And, um, you know, you lose like half the satellites every pass, but, uh, just based
on the initial estimates, you can have, I think, uh, 200 satellites actually make it
through the entire A ring, uh, going like seven or eight times through the A ring themselves.
Uh, a hundred satellites will make it through the A ring and then make it through the, um,
um, the Cassini division.
And then a few will make it through the B ring and stop.
So you'll have like all these satellites getting lost as they go, but you'll make 10,000 passes
successfully through the rings of Saturn, even though you're losing lots of them.
And I mean, I'm sort of imagining this arrow breaking maneuver where you've got this, would
they be a clump of satellites and then they're like a little bit stretched apart and then
they're all making their way through the atmosphere?
I mean, ideally they'd be as far apart from each other as possible.
So I don't, you know, colliding within the air breaking is probably something that would
break the satellites.
Uh, so hopefully when you release them in their initial release, you spread them out
a little bit.
So that way they don't, they don't collide with each other.
Um, and after that there's very little chance they could collide with each other through
the rest of the mission.
Right.
And then one of the great challenges is transmission.
So how would you get the information home once you're sort of sending these poor little
satellites through there?
Again, that's another, uh, question that we're trying to answer for the proposal or for the
phase one study.
The thought is that, uh, maybe the initial mother ship that dropped them off in orbit,
uh, will still be in orbit and can, you know, cause a large ship has a very, uh, high gain
antenna.
It can receive radio information from the satellites and then broadcast that back to
earth.
Um, once the satellites have transmitted to them.
And so what do you think is like the biggest source of risk at this point?
What, what sort of freaks you out?
Um, so, I mean, there's lots of like development, uh, questions, you know, solar power, radio
power, uh, and you know, looking at the details of those, making sure it's still feasible
afterwards.
Um, one thing that's fun about these satellites, which is kind of novel that enables this mission
is the, the, the way they actuate themselves.
There's not actually onboard propulsion.
They use a single motor to control their orientation in space as they're, as they're flying around.
And, uh, looking at that controller is something that I think is very interesting because it's
not intuitive how to get a single motor to point the satellite in three degrees of freedom.
Uh, and it's when it's trying to, you know, face of the rings or face the sun to recharge
or, um, reach these two drag states for the, for the, on the aerobraking.
And like, typically this is done with reaction wheels on spacecraft and they need minimum
of, I think two, ideally three to be able to control it in the three axes of, of, of
orientation.
How do you do it with one?
Yeah.
So the, I mean, the reason that we want to have one instead of three is you could have
three times as many satellites maybe because the, the, the wheels probably take up a lot
of the space and a lot of battery consumption on the, on the robot.
So making them as simple as possible, uh, is why we're kind of pushing it down to one
motor.
Um, and the way we actually control, uh, the, the three degrees of freedom is that we have
a motor and at the end of that motor is a, is a, a mass with a spring attached to it.
So like there's a spring between the motor and the mass, and if you spin the motor at
slow speeds, that's that spring doesn't do anything.
Uh, but if you spin up faster, that spring will stretch out and then you kind of get
your, um, moment of inertia changes as you're actuating the motor.
Uh, so you can like, you can spin up quickly and spin down slowly.
And then once you reach the motor in the original position, you are now be in a different attitude.
So it's kind of a non-holonomic controller, uh, for the satellite.
So, uh, it's a little bit, it's hard to control, but you can kind of go through these sequences
of motion where you control both the position and the speed of the motor, and you can end
up pointing it in any orientation you want.
And this is a thing that you've tested out and that's part of part of the, um, the study,
but we're pretty confident that we can get that to work.
Yeah.
I mean, that alone, I mean, there was, I don't know if you've been following the news, but
there was this mission to try and recover or help boost the orbit of the Swift satellite.
And there was a booster satellite that was launched just like last week and two of its
reaction wheels failed until it's down to one, right?
And they think they can recover with the propulsion system, use its ion drive to get it back on
on target.
But if you just had only one motor, I guess two is two is one and one is none.
So then of course there's that whole question of redundancy, but if you can just even operate
on one just purely by the, the spring mount that is around your reaction wheel motor,
that offers up some interesting ideas like, you know, now I'm just spitballing here, but
like if you go back to say the Kepler mission, that mission failed because of the loss of
its reaction wheels.
But if you had this sort of hybrid version where you have the motor reaction wheels,
but it's inside a housing that can kind of act like a, another axis if you need it.
Yeah.
I think if you, if you build and plan ahead for this type of control, you'd have to kind
of have the motor in a unique position and it may not be the same type of motor you'd
use for like these high speed flight wheels.
But it would be possible to use this type of mechanism on a bigger spacecraft.
One hard thing is that I think most of these spacecraft care not only about pointing at
a particular direction, but tracking something in a smooth motion.
And this control that we have doesn't, it can't track, you can only point.
So think of it like if you're driving a car and you want to go from pointing in one way
to another, maybe you have to do lots of weird back and forth until you point to the right
direction.
This does something similar to that.
So if it's pointing one direction, it needs to go to a different direction, it might do
a bunch of strange wiggles and then eventually restart.
Right, right, right.
So you're not going to be tracking say some target perfectly as you're moving.
You're just, you kind of end up where you need to be through a lot of, yeah, a very
messy set of movements.
That makes sense to me.
But still, it's kind of interesting how there's this idea that I'm really excited by where
there are various metamaterials, various kinds of existing stuff that naturally wants to
solve a problem.
That instead of you trying to come up with some kind of engineering solution for it,
which is a potential point of failure, you use the right kind of material that can, or
the right kind of configuration that provides you that solution just as part of the process.
And so it's like one less point of maintenance and failure and so on.
If the underlying just structure of what your thing is made of, you know, these various
bimodal metals or things like that, right?
I think it was a trade-off between the traditional systems, which are, I guess, easier for engineers
to understand and easily control, and then these more exotic systems that are maybe more
efficient in mass or something like that, but are not easily controllable or not very
intuitive to understand how to do that.
So did you start with, what could we use swarms to study in the solar system?
Or did you start with, how could we study the rings of Saturn?
We start from the question of swarm.
So that's my research background.
I build swarm robots.
So trying to understand how to build lots of simple things and control them to do behaviors
that we care about as a group.
And we've been looking at one-motor vehicles for a few years now.
And the question is, can we have a one-motor vehicle in space?
And then by going to one motor, you can make them very small and then make lots of them
for the same type of payload and mass.
And the question is, what's a mission that can take advantage of that capability?
And that's kind of where we landed with this going to the rings of Saturn idea.
And so were there second ideas?
What were some other places in the solar system that you thought might be interesting to send
this kind of a swarm to?
So not only is it fault-tolerant, but it's also distributing sensors across a wide spatial
area.
So you can actually, we have some algorithms for getting them to spread out equally in
an orbit.
So you start them all in one clump, and then you come back a few weeks later and they're
equally spaced along the entire orbit.
And that way you can have lots of sensors along lots of areas.
So maybe if you care about looking at very quickly changing magnetic fields on a planet,
like maybe Uranus is important to understand the magnetic fields, then maybe this is a
mission that would be helpful for that.
Because if it's changing very quickly, it's hard to measure that from a single point that's
moving around.
I mean, I'm thinking about something that, like, for example, say the plumes on Enceladus,
where you could send them in a stream through the plumes.
Yeah, you're going to have some, some are going to get hit by ice or whatever, but you're
going to then be able to measure the dynamic quality of the plumes.
Are they changing in strength and so on as they're, as material's blasting out?
I mean, having them go through would be helpful, but like, how do you get them to move in that
way is a hard problem to do.
And that's why we take advantage of this aerobraking around Saturn, which should probably be harder
to do around a moon.
Yeah, it's trying to aim to do your aerobrake and get through the plumes at Enceladus, which
is however many millions of kilometers away from from Saturn.
The other idea, you know, a lot of people are thinking about these, these tiny spacecraft,
even smaller versions, chipsets, for interstellar probes, you know, this was the idea of the
original Project Starshot and and, you know, there are they like, how do you make these
things small enough and accelerate them to relativistic speeds?
You can cross between here and another star system in a small amount of time.
But the other issue is communication that you won't be able to send the flagship mission.
And so, in fact, these tiny spacecraft can work together to to mimic a larger radio dish
that could be sending a transmission home.
Yeah, I mean, I feel like that's a order of magnitude harder problem than the one I'm
looking at.
So I'm, you know, I'm looking at using mostly easily available components, like there's
nothing novel about any individual components that would be going on the spacecraft is kind
of just the control behind them and maybe, you know, the motors as part of the spacecraft.
What you're talking about, I assume, is building the entire thing from scratch out of nothing
you could buy currently.
Probably.
But but I think that, you know, people have jumped straight to, well, let's build an
interstellar probe.
And I think what you're addressing is let's develop swarms to explore the solar system.
And then you could do things like, well, instead of them being carried on a giant carrier
spacecraft, let's attach a solar sail to them and accelerate them with lasers.
And maybe now you're sending 10,000 tiny satellites to 10,000 different asteroids.
That is, you know, you're not getting that swarm, but even swarms exploring individual
asteroids is is a very interesting process as well.
But but just this idea of us getting really comfortable exploring the solar system with
small satellites, as opposed to trying to make them, you know, large flagships that
that disposable nature.
I mean, I think that's a cool idea.
I think, you know, having a swarm like this lets you distribute them spatially.
So now you can set a satellite to every asteroid that you know about.
And, you know, it makes it low cost and fault tolerant.
You know, if you're sending it to every asteroid and you miss one of them, it's probably not
the big deal.
Yeah, yeah.
The idea that I liked about this to solve the communication gap is that you you send
them on on trajectories that will then bring them back to the Earth.
So you say you pick a you pick an asteroid, you send the small set towards the satellite.
It does a flyby, you know, reaches the very height of its perihelion.
And then, oh, sorry, it's appealing and then comes back down to Earth and does a flyby
and then transmits its data when it's close to the Earth and then, you know, can be disposed
of or or goes off to another target.
So but now I'm just now I'm just rabbit holing.
Um, well, Mike, what are you obsessed with right now?
Uh, I mean, a lot of stuff I've been doing has been looking at robot minimalism.
So I'm really interested in, like, how do you get robots to be as simple as possible
and still do the things that we care about in robotic systems?
So instead of having a four robot, a four motor flying robot, can you with one motor?
Can you have like a satellite that has one motor?
Can you have any other type of robot that's very simple, but yet capable of doing the
things you care about?
And that makes it easier to lots of them.
I mean, really, for me, like when do they wash my dishes?
You have a dishwasher, which is a robot that everyone has.
Yeah, I know.
But it's not picking up the dishes from the cabinet and it's not pulling them out of the
sink and it's not putting them away when it's done.
It's good, but it's not a problem.
Yeah, I know.
I know.
Yeah, I think I think that is the that is the line for most people, which is that if
you can have a robot that even a flying robot with knives for hands, I don't care.
Have it come through and do my dishes and clean up after me.
And I think we'd be willing to accept any kind of risk.
You know, I sort of imagine Terminators stepping on the skulls of humans in the future.
Fine.
As long as they're cleaning your dishes.
Yeah, for 10 years we had our dishes done.
Yeah.
And then and then the robots took over.
It was worth it, though.
Yeah.
Well, my congratulations on the Knight Grant.
Good luck with with the missions they're working on.
And I hope to see the man pictures from Saturn's rings inside the rings.
That'll be crazy.
Thank you.
I hope you enjoyed this conversation with Dr. Mike Rubenstein.
As always, I did another longer version of this exact interview where we talked about
another project that Rubenstein and his team are working on, which is to make an invisible drone.
So it's called the Phantom Twist, and it is a drone that the entire object spins so quickly
that kind of like how a propeller on an airplane makes it so you can't see the propeller.
Now you can't see the whole drone.
Now it's still very loud, but it is much harder to see.
And actually, the goal is not to make an invisible drone,
but to make a safer drone for people to be able to be nearby.
So if you want more information on drones and just kind of like drone society in general,
I think you'll really enjoy this longer version of the interview.
As always, it's over on Patreon, but it's completely free, open access.
You won't see any ads.
I'll put a link down in the show notes so you can go and check it out.
Now I'm going to give you some final thoughts.
But first, I'd like to thank our patrons.
Thanks to Abe Kingston, Andrea Pedretti, Brian Bowdoin, Caridin, Chuck Hawkins,
Commander Baylock, DarkFinga, David Gilton, and David Matz.
And for all the reading and math for toddlers, Evan.pro, James Clark, Jeremy Madden, Jim Burke,
Marcel Smits, Michael Purcell, NASAOceanGuy, NordSpace, OneStepAnimals.org.
Please follow my nephew at vbrick6994, Ren Kaidu, Richard Williams, Sean Sargent,
Stephen Findlay-Munley, Team49, Telescopes Canada, Vlad Sheplin, Wolfgang Klotz,
Zelda, oh my God, we have proof of aliens who support us at the Master of the Universe level,
and all our patrons.
All your support means the universe to us.
Well, this is great.
I'm really enjoying my run through the NASA NIAC awards so far.
This is the third one that I recorded.
I've got at least another five in the queue so far.
I think there's 18 that were released this season.
And as always, I'm going to try and get every single one of them to talk to me.
If not, well, that'll be too bad, but I'll try.
And I hope you're enjoying them as well as much as I do.
And of course, we're going to have all of the interviews in a collection on Patreon
or on a playlist on YouTube.
And so if you want to go back through not just the ones that we've done this year,
but the ones we've done in previous years, they are all just bonkers ideas,
totally out of left field, out of the box thinking, technological readiness level zero.
But these are the kinds of ideas that may be able to eventually carry humanity to the stars.
All right, we'll see you next time.