#37 Mathieu Johnsson | The challenges of building maritime drones, engineering culture, the future of war - Bang Two Three Podcast

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
So at that point, because it's experimental,

(00:02):
and there's very little knowledge,
you're gonna make, you're gonna crash,
and you have to crash because that's how you get better.
- Yo, is that your drone to your left?
- Yeah, that's actually an interesting story.
It's like an earlier version,
this aircraft was meant to fly.
And because it's also really printed,

(00:24):
we do modification at every single iteration of the aircraft.
And that does mean some times we get it wrong.
And so we did a modification, didn't realize we moved the wing,
not wanting to do it.
And by the time we assembled it,
we're like, this plane's never gonna fly.
So let's just turn it into a cool piece,
but this is actually like all manufactured
to be a fly with aircraft.
We just know if we're actually flying,

(00:44):
it would just fly upside down.
(laughing)
- That is so cool.
- So this is a demo aircraft.
- Is it possible, like if I have an aircraft
where the right wing is this high,
and the left wing is this high, can that fly?
Can you compensate for that some way?
- Yeah, I think I can not actually be a problem.
There are aircraft that actually are asymmetric.

(01:05):
So you have some plane that would have the left wing
would be forward of like the center of the aircraft
and the right wing would be behind.
You can actually, like these are like planes
that were experimental, but like it's all public,
you can find the images.
So there are things that look way more accurate
than just one being slightly lower than the other.
You have some literally where it's, yeah.

(01:27):
- Asymmetric.
- That is crazy.
What was your favorite subject in school?
- I think physics.
I mean, actually I think it was biology.
I think I was best in biology for whatever reason,
even though that's not what I did at all.
But I really liked physics.
I have a clear memory of like one day
where there was like, you know, in high school level
kind of physics where I was like,

(01:48):
oh, we're gonna model that piece of, I think snow
that is gonna slide from the top of the roof.
And then it reaches the end of the roof.
What is the speed that it reaches?
And then, you know, how far does it fall down from the wall?
It's fairly basic.
But I do remember thinking, wow, that's so cool.
They can predict this.
And I think from there, like, it's like,
physics is really cool.
- Exactly, right?
- And you say it.

(02:09):
You say it's fairly basic, but like there's a lot of things
that play with the snow, right?
So we're talking about like friction, right?
How big is it?
How much friction is there on the snow?
Also too, like if you wanted to get really technical,
how packed down is the snow?
Because here's the thing, like very loose snow
equates to tons of friction, right?
But like very packed down snow is essentially less friction

(02:30):
than actually rolling down a hill because it's slick.
So like there is lots of variables like that,
but it is, I totally agree.
It is really cool to say, hey, I'm dropping this bowling ball
from the roof and you can use math right now
to tell me how long it's gonna fall.
- Yeah, exactly.
- That is not real.
- That's true.

(02:50):
What was your favorite topic at school?
- Math.
I like, listen, Matt, you touching the touchy subject?
Listen, I was so bad in school.
I graduated with very, very low grades
because I was more interested in getting pretty girls
in trying to party.
But one subject that I did like regardless of anything

(03:15):
was math.
It was interesting because I was, you know,
I made Ds and Fs, I made very low grades,
but I was in an advanced math class,
like for gifted students.
It was so interesting because all of my other gifted students
in the math classes would make A's.
They would make straight A's in all the other classes,
but I would make a A in math,
but an F in everything else.

(03:36):
So I just really like math.
It just kind of clicked for me, like, say, okay,
there's a way to solve this.
I have some tools in my tool bag,
different ways to approach these formulas.
And it's up to me to choose the right one.
And that really like inspired me
and really like stimulated my brain.
- Yeah, I can relate to this.

(03:56):
I guess that's why I liked about physics.
It made it more tangible than math
because math can be like, it is very abstract.
So it's kind of, I mean, physics is just,
like the real word implementation of mathematics.
So I'm not out of it, I think.
- Do you know who is considered
like the birth person of physics?
- No, I've been in it.
- Yeah.
- I think they had to be on some drugs, right?

(04:18):
Because could you imagine it like the year 1500?
And you know, you're just sitting there
and you're like, wow, that apple fell from a tree.
I wonder if it falls the same speed every time
and I wonder if I can predict that.
- Yeah.
It's hard to really try to relate to those times when, you know,

(04:41):
I mean, everything we do today,
you always stand on the shoulders of whatever,
like the people before you build previous stuff
and we would be completing competence.
The two of us, if we were brought back to those times, right?
Yeah, I don't know what we would do.
But we, Oscar, you all said we'd be pretty useless, right?
So it's hard to imagine how you come up with,
yeah, those formulas to start with, yeah, maybe drugs, as you said.

(05:03):
- Or alcohol or whatever.
But it's more so like the intuition of,
like why are you curious about this?
I think like, especially back then,
there was a lot of things going on,
there was a lot of sicknesses.
We didn't know like a lot of stuff about health biology.
We didn't know a lot of stuff about a lot of stuff.
But some guy or gal was like,

(05:23):
I wonder if all of these objects are in relation to one another
in some way and I wonder if I could predict that.
Like that is, I don't know who that person is,
but I wish I did because like you said,
we're standing on their shoulders right now.
I wanna chat, so in 1986, you're probably familiar with this,
here in the United States,

(05:44):
we had the Challenger Space Shuttle that blew up
and killed everybody on board.
One of the engineers that worked on the Challenger Space Shuttle
on his way to watch the launch,
leaned over to his daughter and said,
the Challenger is gonna blow up and everybody is gonna die.
That's exactly what happened.
Now, with hindsight, we know that the engineering teams

(06:06):
were screaming from the rooftops saying,
do not put people on this Space Shuttle, don't do it,
don't do it, it's not ready for showtime.
But yet, management, whatever you wanna call it,
intervened and said, you know what?
We don't care, we're still gonna do it.
I am, I work with computers as my day job,
software engineer, programmer, whatever you wanna call it.

(06:29):
So I've been on some of these engineering teams
and I've seen like the dynamic where engineering
and management a lot of times are budding heads,
not like bad.
But for example, I've never been on an engineering team
where we released a product
and every single engineer was like, yep, that's it,

(06:49):
it's ready for showtime.
We always, we go to launch and we're like,
I cannot believe this, listen, we're gonna get hacked.
This doesn't work, we're gonna lose users,
the company is gonna go under, right?
'Cause we're engineers, right?
We get paid to perfect stuff, we find flaws
and we can't sleep until those flaws are fixed.
But management steps in and says, hey,

(07:10):
you know what, we need to make some money,
maybe the product isn't perfect,
but let's release it anyways, maybe it's good enough
for our users.
So you previously being an engineer
and now being a manager,
how do you deal with that dynamic?
- Well, first I'm still very much an engineer
because I still, I mean, I'm still the architect

(07:31):
of the product, so I'm still very much like hands-on
and my background is, yeah, aircraft design.
So that's, I would say, I'm kind of like having to
come for both sides, but how do I handle the dynamic?
Well, I think it's, what's interesting when you talk
about like NAS on the space shuttle,
it was very much experimental, I mean, by definition,
they weren't that many shuttles and that many flights.

(07:53):
So it's, and also, I mean, NAS I was not like made for
like commercial gain, right?
Is, so the dynamic of how do you get to,
it's not like there is a financial incentive
that managers are like, oh, we need to work
for the bottom line or whatever,
maybe there was like, oh, you know, whatever the budget
is constrained and we need to hit that milestone

(08:15):
otherwise the project is gonna shut down or whatever.
I'm not too sure what happened exactly.
I mean, from what I know on that specific flight
was like some issues with seals that were not designed
for the right, for the right flight condition or something,
I can't remember the exact detail, but,
so I don't know how this translates to like,
as a startup where, yeah, sure, you have to get to like,

(08:36):
some eventual economical benefit to what your solution is
at some point you might have to like, as a manager,
like, waste between like technical readiness
and releasing a product.
So yeah, it's an interesting question.
- Do you feel like a startup?
- Do you feel your product right now is like technically
complete, like is there or is there like a to do with this?

(08:59):
- No, no, no, no.
- Okay, yeah, there we go, no, no, no.
- Not at all.
I mean, it's, but that's not even our product,
that's just the industry.
Like, I mean, you know, like the safest mode of transportation
on the planet is to get onto an airliner.
It's like by far the safest.
And the reason it's that safe is because,

(09:20):
the any plane you get on board of today,
basically the same from like maybe 70 years ago.
Like the first plane that had the configuration
that planes have today, you know, like a fuselage
with a swept wing and like engines hanging below it.
But it almost, if you look at planes today,
you're just saying that's what a plane is like,
but that's not what a plane is like.
That's just that specific implementation of a plane

(09:41):
is like that.
And all the airliners have that exact same design,
which means the reason it's safe today is because
we have so much cumulative experience
of flying exactly that plane.
I mean, sure, there's been, you know, small improvements
and things, but when the improvement in planes today
in like airliners, I think it's like 0.3% per year,
like the improvement in fuel efficiency or something,
is very easily nothing's happening.

(10:01):
So, so that it's so safe because there are so many
of those experiments and cumulatively every time, you know,
there's a fuck up and something blows up,
then you fix it and it never happens again.
And then progressively you just become so safe
because you've covered every single possible thing
that could ever happen, right?
But when you're in a new industry like drones,
where, you know, we've designed this concept,
someone is gonna design another concept,

(10:22):
nobody has any experience in any of those concepts,
it's electric, there isn't much history
of building electric planes.
You basically restart from scratch.
So, none of the knowledge from before, I mean, physics,
you know, it's still the same.
Like the knowledge of the specific aircraft concept,
but it's not transferable, you need to redo everything.
So, so at that point, because it's experimental,
and there's very little knowledge,

(10:43):
you're gonna make, you're gonna crash and you,
you have to crash because that's how you get better
because you crash and then you're like, okay, well,
when let's fix this problem that we didn't know existed
and then it never breaks again.
And if you do it properly and then you get to the next one
and the next one and the next one,
until ultimately you fix like,
all of the one required to have like a plane
that is like super reliable.

(11:03):
And nobody in the industry has gotten to that stage yet.
So, so if you look at like the most reliable drone,
probably on the planet is gonna be something from,
except probably Boeing, there's like a drone called
the Scanigol, which has been in service since like the year 2005,
it's flown like, probably millions of hours by now.
And an individual drone like this,

(11:24):
which is like a 25 kilo, three meter span drone,
so relatively big.
An individual drone lasts for 500 to 600 hours,
like the estimate and that's nothing.
Like, it's not really, you wouldn't get on a plane
that has most likely is gonna crash in the next 500 hours, right?
(laughing)
So, so when you say, are we, is the product finished?
If you mean finished as it's as reliable

(11:47):
and as safe as like a regular plane,
like where the industry is miles away from that.
And our specific iteration,
because we're more novel and pushing boundaries a bit more,
we're also on the slightly lower end of like the reliability.
So there's plenty of stuff that we need to figure out.
And so yeah, every day when we go flying,
there's always a question of, you know,
amongst the things that we're testing as part of this flight,

(12:10):
are we comfortable with the risk?
For us most of the time, because we are,
we do my time monitoring with our drones.
So we fly above water very easy to mitigate the risk
because you basically aren't gonna crash on anyone.
And the air traffic also is, like there's no air traffic
low level of above water.
So it's the safest environment that's one of the reason

(12:31):
we actually do mostly operation above water
is because it's much safer of an environment to do all these tests.
But yeah, so definitely it's not set, it's not sorted.
And we have to deal pretty frequently with like,
what's the risk of crashing?
And are we comfortable with it?
- So what metric do you use to measure success?
You were saying that Boeing's drone,

(12:52):
it's how many air hours you can do before it,
it could put before it goes out.
Is that what you're striving to do with your drone
is to increase the air hours or do you use a different metric?
- We actually do use very different metrics.
And I think that's probably one of the things
that is most interesting about what we do.
'Cause figuring out when you have a new industry
and you don't know yet what's gonna be most impactful

(13:13):
is, you know, and Mandelcraft, like, inevitably
when we have millions of them flying around,
it's hard to know in advance exactly
what's gonna be the most impactful outcome.
Figuring out, like taking a bet on what we think
that's gonna be and like figuring out the metric
that we be important for that use case is probably
what matters the most if you want to be
like really relevant in the future.
And it's a bit like I like the example of not for drones,

(13:35):
but an equivalent industry that had to go multiple times
figuring out what that metric was is the hard drive industry.
So, yeah, you know, you're in software,
but it's like, there might be some people in New York
and they don't even know before I says these in some way,
but it's like, you had like very large disks
and then the generation after it was like a slightly smaller one,
and every time the tradeoff was like,

(13:56):
"Oh, do I want the smaller one?"
"That will be mobile Apple?"
Like, I can't, I don't have to carry my laptop being careful
because those numbers I'm gonna damage the disk.
And, you know, you had, I don't know, you probably remember,
like you had those iterations, right?
And every time a company was a leader in one of,
in like the established size of hard drive,
they were like, "Oh, we're just gonna keep making it
as cheap as possible per unit storage."

(14:19):
Because that's what people care about.
They want just more storage, and that's true.
Up until, and you had those other companies that were,
"Oh, we're gonna do the smaller one."
That is way more expensive, but we're betting that one day,
people will not care that much about how cheap it is
per unit storage, they will care way more about, you know,
it's more compact, consume their energy, whatever.
And what happened is this actually happened multiple times,
and every time the company that was the leader in one domain

(14:42):
just disappeared before the next one that replaced it, ultimately.
And so it's kind of like, people really struggled
to see what is the metric that matters in your industry,
and every time these guys were getting it wrong again,
they were just focusing so much on like,
long term is dollar per gigabyte or whatever.
And every time, it's always true, but there is a time
when people stop caring and they care more about
like a small disk or faster or whatever.

(15:04):
And so for drones, I think there is an equivalent,
which is right now people all care about like endurance,
because the first drones were all in the defense space,
and the main use case for defense is,
how long can you, you know, you take off,
and how long can you stay above that enemy position
and keep me giving me intel?
We can stay 24 hours, that's fantastic,
you know, I don't have to think about it, it's just always there.

(15:26):
But if you're talking about like an industrial use case
or a use case where you want to, you know, survey an area,
it almost doesn't matter, like how long it takes,
you actually want something that goes very quickly, right?
It's like a bit for, when you get on the plane
to go somewhere, you're not buying flight hours on the plane,
you actually want the least amount of flight hours, right?
You just want to get to the location,
and the metric in that case is range or your speed
because that's what's gonna get you the most utility.

(15:49):
And people don't look at that for drones,
they just look at, again, endurance,
like how long can you stay in the air and give me information?
We don't look at all about endurance,
we just look about, actually the only thing for us
is dollar per square kilometer, which sounds obvious,
is like the metric for how much data
you're gonna be collecting at what cost.
And for that, it means you want a plane that flies fast,

(16:09):
so you do lots of kilometer per hour,
and a plane that collects data with a sensor
at very far is a side.
So you can imagine it's a kilometer that way,
and then kilometer per hour that way,
that's square kilometer per hour,
so you do hide that collection, right?
And then we use the cost of the flying of the drone,
and then you get your dollar per square kilometer.
And that's on metric, like the top level,
that's the only thing we care about, really.

(16:31):
- So if you could snap your fingers and 10X,
one aspect of your drone,
like in order to crush your KPA,
like the thing that you're solutioning for,
what would it be?
It sounds like it would either be more batteries
or better sensors, right?
- I know what the battery is,

(16:53):
there is like a weight issue and stuff like that,
but like, because I'm guessing that you do spend engineering hours
on like fixing stuff that is gonna have the biggest impact.
So like what is that component on the drone
that if improved would have the biggest impact?
- Actually, it's, it's, it's,
it's news of them actually,
while the one that will have the biggest impact

(17:14):
is actually happening next year,
and that is how solving the connectivity issue to drones.
'Cause right now when you fly a drone,
the only accepted mean is direct communication
to the drone from your ground equipment to like the aircraft.
The reason for it is one,
I mean, sure, you can have 4G in some locations,

(17:36):
but it's not that reliable enough,
so often regulators don't accept 4G
as like a primary mean of control because you might lose it.
It's hard to predict, you know, maybe some time
you lose signal on your phone is, it's okay,
but if it's a drone then you lose signal, that's pretty bad.
So 4G is not really used that,
it's not really accepted because of this,
so it's always direct line of sight.

(17:56):
And the issue with this is you have really strong constraints
on what frequencies you can use
and how powerful the signal is allowed to be,
which completely legitimate, you know,
like you don't want to like completely fill the,
and the electromagnetic spectrum
with everyone emitting stuff, right?
So it's, it's very constrained,
and regulation are different for countries.
So when you're someone like us who go and fly maritime monitoring

(18:16):
mission with drone in different countries,
you have to constantly change your thing,
you have to get new permits for like just,
and you're just talking wireless transmission
at that point is very painful.
So this would be the one thing that I wish was different,
and it actually is gonna be different
probably from this year or next year with Starlink,
'cause pretty soon is just,
I don't know if you've seen the Starlink,
they now are doing direct to cell.
So you don't even need like their antenna
and their stuff like literally on your cell phone,

(18:38):
like they did some demos where,
I think they were getting 17 megabits per second.
So what they said is this year,
they're gonna do text direct to phone.
So I guess you get a mobile phone subscription to Starlink.
And then from next year it's gonna be data.
And for us that's a massive breakthrough
because we can just use a 4G chip
that doesn't connect to the local network,
just connect directly to Starlink,
very reliable across the planet, you get it anywhere,

(18:58):
you don't need to adapt your direct communication system,
you don't need to travel with an antenna
that is connected to your system and talks to the drone,
the drones go and you don't need to worry
about wireless transmission.
That's gonna be a big breakthrough
for the whole industry.
It's gonna make everything so much simpler
because you wanna have to worry about transmission anymore.
It's just, you know, the thing flies, it gets data
and you don't have to worry about it.
It's gonna be great.

(19:19):
- That is beautiful.
So currently what happens if,
let's say you're flying your drone in a remote location
and you lose cell coverage, does it crash
or does it have some type of fallback?
- So we actually don't use cell, we use,
because as we fly mostly above water,
it's, you actually get really good 4G signal
to quite far out, but not everywhere

(19:40):
and it's hard to know in advance.
We don't use 4G, we use direct line of side communication.
So we have basically a transmitter receiver on the ground
from where the pilots are operating the drone from
and we have another one on the drone.
So it's like bi-directional communication.
The drone sends telemetry about, you know, its speed
or status of its battery, whatever gets to the ground
and we do the same data the other around.

(20:02):
We're sending commands, you know,
goes this direction, here's the new mission plan or whatever.
So we have one channel that's called telemetry leaf
for doing that and that is on one
specific frequency and power that tends to be like
very long range.
So this one is pretty easy to sort out.
And also it's very low bandwidth,
because you're just sending like strings of text like really.

(20:22):
The way, the other one that we need is video
to transmit the data from the halo down the aircraft.
So video images and this one, so suddenly one way
comes down from the drone.
This is much more constricted for bandwidths,
like we need a lot more.
That's also on a higher frequency
so it tends to be short range.

(20:42):
But it's also a direct line of sight.
So again, transmitter on the ground,
talks to the plane and the way back.
And this typically goes on top of a mast
because when you fly quite far,
you actually need to have quite a high level antenna
to be able to reach the drone.
So yeah, that's a lot of equipment.
That's why it's so much better for that starting.
- Yeah, no, so what about?

(21:03):
- Oh, go ahead.
- Sorry, I actually knew and answered your question.
You said like, what happens if one of these things fails?
So I just explained to you why we don't use 4G
that way we use right now.
If we lose connection,
I said we have one pipe basically for command
and control of the aircraft and another one for video.
The one for video can actually be used

(21:24):
for command and control as well.
So it's kind of a redundancy.
If we lose the main one for command and control,
then we can still use the one that is used for video
as command and control.
So we have some other redundancy there.
But if we lose both,
the aircraft knows that it's lost the communication
to the ground station.
So it's programmed to just come back at that point
and it has very specific location

(21:45):
that it's meant to come back to if it loses the signal.
So you plan these in advance.
And typically as soon as you start flying back
because you get closer, you get back the signal.
So it's pretty like you would just lose the signal,
flies back and maybe wait a minute
and then you have your signal back.
And all of this you design it so that no matter
where it happens, it remains safe
because you know exactly what procedures the drone will

(22:07):
go through and it's not gonna overfly a city or whatever.
So yeah, that's how we handle having issues with communication.
- That's absolutely brilliant.
I wish I was a software engineer over there.
That sounds so cool, man.
It's like, right, say, you're right.
Like this is such a great challenge.
It's like, okay, because I kind of worked on a project
that was somewhat similar.
Like we used to deploy sensors in like really remote places

(22:30):
and you know sometimes we would lose signal
and we had to have like these type of redundancies.
That's why I asked you.
But that's so cool.
It's like, okay, hey, we lost signal.
Okay, we need to turn, don't turn too sharp
or you know, there's aerodynamics and stuff like that.
You need to come back to this way point, all of that stuff.
That is awesome, man.
That is so cool.
Do you ever think like one day that your drones will be able

(22:52):
to fly fully autonomously?
Or does that not provide any benefit at all to your industry?
- Oh, I mean, 100% for sure.
It will, without doubt at some point be full autonomous
and it will be very different from what it is today.
So maybe I'll explain just briefly how it works today
because some people might have a misconception.

(23:13):
But so most drones can are now like fully automated
not autonomous.
So automated, which means you can just write the mission.
So you know, like for you like software engineer,
you could just write a text file that says,
you know, here are the coordinates and the time I want you
to be there and drone will just go and do it.
And with no supervision.
But it will also do that in such like as like blindly.

(23:36):
He will assume nothing is there.
And if there was something between like where it has to go
and where the drone is now, I'd just crash into it
like unless it's terrain.
Like I mean, but it's so, but it's mostly automated.
It does that automated job very well.
Like the drone can fly itself and handle changing weather,
changing wind, whatever, like no problem.
The, there are two difficulties.
One is, but all of that flight you still need human supervision

(23:59):
just from a regulatory standpoint.
And often that regulatory supervision needs to be very close
to the drone to make sure the drone doesn't crash into someone else
like another plane or something.
Often it's 500 meters.
So that means you need someone with, at all times,
within 500 meters of the drone,
which is like completely breaks the whole concept, right?
Because either you don't fly the drone far

(24:20):
or you need someone every 500 meters
which against completely different.
So that's the, that's the limitation number one today.
And the second one is actually let's,
let's imagine like the second one is the, the ground handling.
So like imagine when you land with an airplane on an airport,
you have tons of people that are there, like, you know,
hooking up the plane, like unloading the luggage,

(24:43):
refueling, whatever.
It's the same for the drone.
You need to like pull out the battery.
Maybe there's like some maintenance you need to do.
So you still need people on the ground to do that stuff as well
on top of the flying.
So those two constraints kind of right now,
they work well together because you need the pilot
to be there from a regulatory standpoint.
And you need that pilot to be there anyway
because some work needs to be done on the drone.
But inevitably in the future, probably pretty soon,

(25:07):
you're going to have better regulation.
So you don't need the 500 meter person
at all time away from the drone
because there would be, you know, an atrophic management
that makes sure the drones don't crash into each other
than other airplane users are safe as well.
That's like that's going to happen for sure.
But then at that point, the big problem
will be how do you automate the ground handling
and everything because you don't want to have fully autonomous drones
but then every time when it lands,

(25:28):
you need someone to swap the batteries, right?
So you will need an infrastructure of kind of like robotic airports
or something.
And but these two things will happen for sure.
And then when they do, drones will be a lot more like,
I mean, if today they're like,
like today you go somewhere with your drone, you fly it,

(25:49):
you pack it back and you go back to your location.
That's typically how it is.
It's like very bespoke if you want to collect data with a drone.
In the future, it's going to be a lot more like satellites
where if you want, you know, today like we could log in on a website
if you wanted to buy satellite imagery from somewhere,
you're not going to buy your own satellite and then go to the location
and then launch your satellite and collect the data, right?
That's what it used to be a long time ago.

(26:10):
A long time ago if you wanted satellite images,
you had to buy your own satellite, put it in space,
wait a few years and then you could get your data.
Drones are like that today.
In the future, again, it's going to be like satellite today
where there's going to be just an infrastructure of drones.
So some form of robotic airports like hangers,
by which drones take off and land all the time.
You don't really see them. You don't really think about them
and the only way you interact with these drones

(26:31):
will be through web interfaces where you could just log in and say,
"Okay, I want data in five minutes, I want data over there every 20 minutes,
I want an update please."
And then you just get that data, I think that's how it will be.
So at that point is just complete automation
and the only interfaces through the, so that web interface.
Why would we need like a network of drones kind of like hovering above us

(26:52):
if we already have a network of satellites doing that exact same thing?
Well, that's a good question and I think you'll need both.
So the issue with satellites, again, I think with movies,
we get a misconception of how powerful satellites are.
So the first thing with satellites, most constellation

(27:14):
are maybe two or three satellites.
You don't have that much, you don't have any control of where the satellite is at any one point.
You can't steer it or slow it down or just it's going to go around
at a very predictable rate.
And if you want data somewhere on the planet, you can't say,
"I want it at that time tomorrow," you're just going to get told,
"You can get it tomorrow at that time, all the day after at that time."

(27:36):
So let's say you live in wherever it's just going to be,
"I can only get data maybe every two days, maybe every day at 3pm."
If you want any data, any other times, not going to work.
So first, that's kind of limited.
That data is good enough if you want to do maybe a map of an overall region,
but if you're trying to track whether there's a wildfire,
but you're not going to get one update this morning,

(27:58):
you're not getting another one until tomorrow.
That's kind of a problem.
The second thing is you're dependent on the weather.
So if there's cloud coverage, for example,
and you're talking optical sensors, then you're not getting
it in coverage.
There are other sensors that see through clouds,
but then they have limitation on like,
so like synthetic aperture radar, for example,
you can see through clouds,
but it's very different resolution, it's much lower.

(28:18):
But yeah, so that's like those two problems.
And then ultimately, yes, you do have a resolution problem,
because you're not going to get as much data as good data as if you're
flying in the atmosphere with a small drone.
So for example, we satellite, you're not going to get to
reading, like I don't know, on the side of a vessel,
what is the registration of that vessel?
Or you're not going to be able to see,
are the people on the vessel currently,

(28:40):
what type of fishing are they doing?
Is it just a vessel or is it a vessel that is fishing right now?
And these things are important when you're trying to enforce,
for example, in my time protected area.
You want to know these things.
And satellite, I'm not going to be able to give you that.
And that's true even if you're talking nanosatlights or CubeSats,
because these ones, sure, you can do a lot larger volume.
So you solve some of the problems of like, revisits.

(29:02):
So you know, you don't have potentially to wait until tomorrow,
same time, you could just probably get it in 20 minutes.
But then because these are smaller,
you get smaller sensors that don't have the good optical quality,
to be able to see like crazy high magnification levels.
So probably that does mean that even if you look at trends
into the future, slow-alone costs, drones are still going to be much better

(29:22):
if you want to do like very high-revisit, right?
So you know, get the data like every five minutes or real time.
If you want to do high-resolution, if you want to do independent weather,
if you want like, critical response to things,
like a fire or whatever, like drones are going to be much better to this.
But and also satellite is very expensive, like crazy expensive,
if you want to buy data, which again, like we,

(29:44):
you don't think about it this way when you see movies and things,
but it's, yeah, it's like 20 to 30 dollar per square meter,
which doesn't sound like much, but if you want to do a large area,
but also if you want it, like let's say if you want it every five minutes
on updates, then get crazy expensive.
We constantly talk to people that tell us,
like, "Can't use satellites just to expensive."
Yeah.
Yeah.
So it sounds like, doesn't it sound like your vision of the future

(30:06):
that you just described, which is like pretty cool.
I've never heard anybody with this idea,
where we just have like a network of drones flying,
and you can tune in, like to a website, a live stream of like a drone,
to see like, "Hey, I want to see what Sweden is looking like today.
I want to see what China is looking like."
Isn't that vision of the future contradicting your design of the drone?

(30:27):
Because you said that one of the things you're not solving for is airtime.
You're like, "Hey, listen, you know, airtime is cool,
but that's kind of not our thing."
So are you concerned about not being part of your future?
That's a really good point.
That's a good way to put it.
So there's, I guess, these two elements to this.
One is a bit like, you know, when I said that there are different technology,

(30:49):
for, so I sang for a hard drive as the dollar per kilobytes,
or something was getting lower.
At some point people stopped caring, and they started caring about like a smaller drive,
which was more expensive, but at that point,
it was so cheap anyway, the dollar per kilobytes,
so that they were moving to the next tech stack, and it happened multiple times.

(31:10):
I think for drones it might end up being the same thing,
where right now, sure, people will, for some use cases, people that,
where they want high endurance, they might not use our drone.
But by the time all of that stuff is implemented, batteries will be better.
And at that point, a drone that may be today,
flies for five hours in the future, while five for 10 hours,
and then people won't care about flying 10 hours,
ours will fly five hours, but it will be like three times faster than the other one,

(31:33):
so we collect way more data and you'll get very, very faster.
So there is that one aspect.
The other one that I think is way more important, actually, is the issue in our plane design is,
I mean, in some way it's very easy to get more endurance.
It's simple, you just make a bigger plane.
It's not like you're being like a better physics, so whatever,

(31:53):
there's literally bigger aircraft that has bigger fuel tanks or better
more batteries, and then you get more performance.
The downside of this is because it's bigger, it's way more expensive.
And so there is a real question always in our space stuff,
is like you want to tailor what you're building to the use case you're serving.
What that means is, for example, if you fly short-range flights in the US or in Europe,

(32:16):
you're going to have the smaller aircraft that is tailored to do between two small cities,
and also it's just designed to do those missions, not like something much bigger.
The bigger aircraft that are designed for the longer range,
you don't use them on those small flights.
It's kind of like you wouldn't go on a 747 to fly between,
I don't know, I'm not going to pick a city pair in the US because my US driver is not great,

(32:37):
but like I don't know, you wouldn't be a 747 to go between New York and Washington DC or something,
that'd be stupid. I don't even know if there are flights between the two pairs.
And so for us, we've done those drones that are smaller, higher speed,
longer, smaller endurance because we think for a lot of use cases,
you don't care about spending 12 hours in the air.
You might occasionally want to do a very long-range flight,

(33:00):
but it's important. A bit like a 747, but for most of our flights, you should be doing the smaller ones.
So we're building the smaller aircraft that is used for short-range flights if we're doing in
comparison to air travel. We're not beginning the bigger one because most of the time,
you just want for the short mission, just use the smaller ones better.
So it's kind of like those two things.
Okay, this is great.

(33:22):
So not only are you a great technical person, but you're a good CEO.
So it sounds like you're saying, hey, look, in the future, there's going to be drones,
that's up there for 12, 14, 48 hours, depending on how good batteries and technology gets.
Say, look, I'm not trying to compete with those guys is what you're saying.
Say, look, I think that there's going to be a great use case for a different type of drone.

(33:43):
And I'm going to build that drone and listen, Boeing, the big guys, they can fight over there.
But you're a small scrappy startup.
You're saying, look, we're going to corner this area of the market and just focus on that.
Yeah, 100%. Actually, maybe a good example is, I don't know if you've heard in the UK, but between,

(34:03):
oh, by the way, yeah, so we're based in the UK, but I'm French, I've been in the UK for 12 years.
But yeah, so between the UK and France, there's quite a lot of illegal vessels crossing.
It's like 125 people a day on average.
And I decided that it gets probably not the biggest problem, but it is like a super political thing.
And with the government that is in place at the minute, it's like part of their

(34:26):
electorate care about like illegal immigration. And so they're spending tons of money on solving this.
And the way they're doing it is they're spending actually a lot of money on drones to monitor the area.
So you have to picture this area between France and the UK. It's like 30 km wide. It's really small.
And they're flying a drone that flies for maybe 12 to 18 hours. So it's like 150 kilos,

(34:47):
two engine, internal combustion. So like, you know, this is 10 kilos, like,
really different the drone they're using. And the headline figure for that contract is 300
million a year to fly that one drone. That drone flies for 12 hours, but what it does is it does like,
it basically does like a loop that we could do with our drone probably in like an hour.

(35:07):
It does that one loop in 12 hours, and it just occasionally does like circles when it sees like a
vessel or whatever. And you could imagine like that's a completely stupid use of a drone of that
kept capacity. You'd be much better off with just like, you know, five or six, much more than one,
you just send one to do the full loop in one hour. And then if you find something just dispatch
another one to spend like an hour looking at whatever vessel you find. And, and, and yeah, the reason

(35:30):
people do this in the UK is because they buy on the headline number of like, you know, you can
if I put 12 hours and, and maybe it's from like a big company. So it's like a reputable, but the
consistently like people look at those problems and they don't break it down as a scientist would
to think, you know, what is actually needed and what's the best solution to just look at headline number

(35:51):
that historically have been the one that's been important for like surveillance drones.
So yeah, it's going to be interesting to see how us trying to focus on other smaller niche markets where
basically what we do is we go after the market that can't afford those 300 million a year that
just talks about that they can spend. And then we go there, but they can't afford that. And we say,
look, look, we have something that is much cheaper and they're, they like the fact that it's much cheaper,

(36:16):
they can't afford planes, they can't afford like those very expensive drones, but it can't afford
us. But ultimately, they're not going to end up with a better solution that the UK spending like
100 times more. And the view is at some point, they'll realize that and then we can go pick up those
those businesses as well. Yeah. Well, how do you tackle that? Because it sounds like your sales pitch
is highly technical, which is how I think as well. It's, it's like, hey, let's, let's break it down for

(36:39):
first principles, right? It's like, hey, we have a problem. Is it a problem? What's the best way to
tackle this problem? Leslie out five solutions and let's like make a decision based on pros and
cons, technical trade-offs, cost, all of these things. It sounds like that is your sales pitch. But
a lot of times is, it's really hard to get non-technical people or governments or stuff like that

(37:01):
to buy into that. So how, like, what, how do you sales pitch like a politician or something like that
from a technical standpoint, saying that I have the better solution?
Yeah, that's a good question. I take them at dinner or get them a glass of wine maybe, tell them
they're pretty. Is that, is that how you do it? I think that's how the big companies do it.

(37:23):
For sure, that's, I mean, yeah, in defense, like it's all like in relationship driven and that's
probably, yeah, you get people to dinner or you already know them anyway. Yeah. I think for us,
it's, again, like, whatever answer is definitely working progress because that's not easy to
to figure out exactly what's the best way to do it. I think it's like, do you have to play the,

(37:52):
do you have to play the game of the big companies, right? That's the question, right? Do you have to
play the, hey, I need to schmooze politicians. I need to donate to their political campaigns. I need
to know the right people, right? Is this because like it's work for them, right? They're getting the
$300 million contract. Do you feel like that you have to play in that ball field or are you going to

(38:14):
try to approach it differently? I think at our stage, we're not there yet. So, I mean, there's probably,
for sure, we need to get to know the right people, but we're not at the stage on budget where we'd
going to do any donation or anything. And also, I think what we do is compelling enough that we
probably don't need to do that yet. But like, so the question, how do we, how do we start considering

(38:34):
it's like highly technical? I think people are not going to buy the highly technical thing. Like, it's,
it's interesting to, from me, to explain what's different because when you talk to people that
are interesting in the technical field, like it's exciting when you say, oh, this is a very different
way to approach a problem that's really cool. I hadn't thought of it, but yeah, they don't care about
that in general. What works is when we've started doing demos now of like the actual fully integrated

(38:57):
solution where we, like, it's very, very striking. How it's different to how we're approaching things.
So, for example, the haven't even mentioned this, but like the traditional way you collect data
is a drone. It's not actually, drones actually don't really collect data in somewhere like in
surveillance. Traditionally, it's just you have a video, a camera that's, you know, like a gimbal

(39:18):
camera. So you can rotate, you can tilt, whatever, and then it translates video back, you can zoom in
and out, and that's remotely controlled by, controlled by an operator. And so the data you get out of
the drone is just that video imagery that you can store or whatever, but they say, yeah, that's
typically hard works. When we operated, we realized that actually, that's a super and effective way

(39:38):
to monitor an area when you want to find boats because if you look, when we looked at track,
afterwards of operators that were looking around with the camera, it was just a little bit there,
a little bit there, and it was like not very systematic, so you were actually missing most of the area.
And people are not like, why would you hire someone to just go left and right, this is really ineffective.

(39:59):
So instead, what we've built is a completely different sensor where you have cameras that are
pointing just sideways, so you don't look forward or backwards is literally like five high resolution
cameras with fixed focal length optics as well. Looks so much better. And exactly like a satellite,
we just have that slit of imagery and as we fly high speed forward, you just map in real time.

(40:21):
So what we produce is a real time map of what's in the what's in the need. And that's very different
again to just like a drone that flies and just sends you video of like, what's happening over here
in the app afterwards, but here. And now that we're starting to have that full system operational,
we can go somewhere and give that image real time. And when you when you have someone from like,
say emergency services or like the military or whatever, and they have that interface that they can see

(40:45):
the map being built in real time, it's kind of like, wow, okay, yeah, that that now I understand that's
nobody cares about the technical details of how we do this. That's the compelling message. I think
that's what how we we win people over, rather than like the technical pitch. That okay, that is
beautiful. And that is really, really cool if you can. So as it's flying, it's like it's beaming the

(41:07):
data real time, real time. And then you have like some website or some UI where it's like, oh, wow,
this is a map being built. It's real time. You know, I can check it out. That is a really compelling
feature that I think would knock the socks off of most regular people. Exactly. Yeah. And the
thing that gets exciting when you're talking to like people in the software space is that right now,

(41:29):
you can imagine like, it's much easier to build a startup just purely software on AI is like the name
of the game at the minute. So you have tons of people that are building software to do like AI
analytics of drone data. But as I've said, like most drones collect like video feed off a little bit
here, a little bit there and it's like, and they don't fly that off and they're in reliable.
So you're kind of like starting to build, it's like if you were like YouTube and haven't been

(41:51):
invented and you were like building some AI tool and you need a training data set and there was
just no video anywhere. Like, why would you do that at that stage? And it feels for drones, it's a
little bit like this. So what I've just described now like that data collection tool, like I think
is pretty obvious that it's going to be much lower. It's it's built to collect way more data.
And especially what we start with, which is above water, it's very easy to do computer vision on

(42:15):
top and like pick up the boats automatically because it's not really difficult to spot a boat
above water. So what I've just described on top of being exciting for just like, you know,
it's cool that the map built it to build itself as we fly is that it's a perfect thing to do some AI
feature on top afterwards. So you know, detect the boats, label them, whatever, like, and then

(42:37):
and then it's like, you have a real time view of exactly what's happening and that's that's the next
thing yet next ever afterwards. Okay, I got it. I got it. Listen, I'll give you this one for free.
You don't have to hire me. Yo, this is a great idea because what you see a lot of especially
tech, tech companies specifically software companies. Now all of a sudden their data sets are like one of

(42:59):
the most valuable things they have. Like it is super valuable, more valuable than employees, all of this.
And because you're collecting so much more data, you can compile a data set that you can say,
hey, look, yeah, we're a drone company. But yo, we have this data set that nobody else has. And
it said a resolution that nobody else has. And for example, you can fly over a certain segment of

(43:21):
the Atlantic Ocean just hypothetically. And you can do this every day, 100 days in a row. And then
like you said, you could take that data and now you can train some very simple AI. This will be very
easy to do just to detect anomalies. It could be a ship. It could be a damn whale, right? Anything
like this, right? But you have the data set and you could say, hey, listen, uh, government of the UK

(43:43):
or whatever. Um, I know you're interested in this 30 mile patch of water, but I can bet my life
right now that nobody on earth has more data on that patch of water than we do. And we have AI that
will detect it. And so you don't need real humans. It's way cheaper. And our AI hit rate is like 98%.
And so yeah, it becomes like a no brainer if you could compile this data set. Is that something

(44:06):
that you're interested in? Yeah, yeah, completely. And actually it's, um, that's, I mean, in some
way, it's also like when you pitch to investor, I mean, that's obviously that very sensitive to
those pitches because it's like, yeah, just it matches the existing patterns they have about what's
going to be successful is like, you know, data sets and like software. What's it's harder to pitch,
you know, building drones and like, it scares people's off often when you have hardware. So,

(44:29):
so definitely that's like, that's a way to think about what we do, as well like, so almost like we
were, we're doing that hard bit of building the drones and the infrastructure and getting
the permit and all that stuff. And then that's like the defensible elements that then gives us
exclusive access to the data that nobody else can collect. And then that's like a compelling thing.

(44:50):
The other thing that is interesting as well, and again, I'm going to make that comparison to
satellite is that the way drones work today is again, like, let's say you're a, you're a component,
like, um, accompany that manages, um, container vessels that are living like, you know, the San
Laurent, like, North, North, uh, Canada and like North America, uh, you know, the, uh, yeah. So you

(45:15):
imagine those vessels and you need a very frequent imagery in that area because you need to know where
the whales are. So you don't hit it and hit the whales and you need to comply with the required,
like, uh, environmental protection. And also you can go faster because if you don't know where the
whales are, you're going to go slow. So currently you're flying, uh, planes all the time and it's
very expensive. And maybe right now you've started using drones. And in that case, the way we would

(45:37):
have used drones is you would have called the drone company, you would have said, okay, can you go
in that area? And then it would say, okay, we're going to set up an operation. Uh, for the next year,
we're going to be there every day. And they will be flying drones just for your operation. So the
data they collect will just go to you as a company that operates those vessels. And then you're going
to solve the problem. You're going to know where the whales are. Fantastic. That looks a lot like

(45:58):
satellites at the very beginning where again, you want the data. You just had to put your own
satellite in orbit and you would use the data you collected just for yourself. But satellites
today don't work like that at all. You have the people that operate the satellite, collect the data.
And then they sell that same data to tons and tons of users. Yeah. And that's how it's much better,
like it's a much better business, but also you can make it much cheaper because more people are

(46:20):
depreciating the hardware drones are going to be this way as well, which gets back to where you're
saying about the data and how it's interesting to think about that data is that or let's say you're the
vessel operator and you call that company that said, okay, we'll come and fly our drones in the area
and we'll charge you for the full year of all our expenditure because you're the only one using the
data we're producing. Our view in the long term is we'll just have our system installed there. So

(46:45):
we'll have robotic hangers where the drones are flying from and we'll just say, yeah, okay,
we're already flying there. We're selling the data to the coast guards. We're already selling the
data to the people that are managing the fisheries in the area. Yeah, we can sell you the data. We
already have it and it's going to be like 10 times cheaper that whatever that other company quoted
you and you'll have it faster and whatever. So yeah, it's completely different way to think about

(47:06):
drones rather than, yeah, someone you just hire and it will come with the aqua. It's just going to be
that data that everyone shares and can use the same data set really. I wonder if you're drone, it seems
like your drone would be perfect for some of these large container ships and other type of ships
that go in these hostile waters where there's pirates, right? Because it's so like, couldn't you just have,

(47:32):
like you said, a robotic hanger, right? And I'm on the boat and then your drone launches and it's
like it can just scan the boat for hostile threats such as pirates and stuff like that and do
something about it. Have you ever thought about that use case or am I watching too much TV?
No, no, yeah, we do have renders for that use case. So, yeah, I think specifically is

(47:55):
like insurance companies that would love to have like better understanding of how risky any situation is
so you can better price how you do your insurance. So there's definitely a use case there.
It's just a lot of what we've talked about is like long term implementation of drones like into the
future and when you talk about use cases, there's also to consider between where we are today

(48:17):
and that long term vision, what are the steps you're going to take to get there and some use case
are going to be close to today and then some use case are going to be the very later stage one, right?
So, for example, the where I said long term, it's going to be that infrastructure of drones,
everyone shares the data when that's the case, that's when the data will be the lowest. So the people
that have the least money will be able to get the full benefit of the technology at that stage.

(48:41):
Today, you probably want to go after people that have a bit more money because you're not going to be
yet able to have a full installation everywhere and everyone could get the data, right? Similarly,
when you talk about operating from a vessel and that's kind of a high risk environment because the
vessel is moving constantly, there's probably a lot of risk if we were to crash, you know, do we
damage the vessel and things? Also, you would need people onboard the vessel that we need to maintain

(49:04):
the aircraft in that vessel that maybe would be in dangerous waters by definition based on like
the use case you've talked about. So it's 100% something that will, I think for sure, at some point
inevitably have a system like the one I've described, but it might not be the first one. For example,
maybe an earlier one is going to be military vessels. So, you know, like the one of the biggest

(49:27):
changes you had in navy is when they started having helicopters on vessels because it massively,
massively amplifies the capability of a vessel if you can have a helicopter on it that can
extend the range of how much you can see around it. But most vessels can't have helicopters because
they're too small and in some way, like, you know, especially these days, you don't want really big
vessels that are susceptible to like one's tracks and things. You want smaller vessels. So there is a

(49:51):
lot of demand for like smaller military vessels where you could have drones on them. So that will
have, I think, happened before you have commercial application that the one you've talked about,
like anti-piracy, I think before that you might have smaller military vessels that don't have
helicopters on right now that will have drones that can give you that surveillance capability around
it because they have a bit more money and they're probably, I exposed to more risks so they're happy

(50:14):
to like start implementing that technology first. But yeah, so you already had 100%.
Yeah, yeah, like, there's so many, there's so many things just popping into my head on like how
this can be used in like really cool ways. And one of them is like, I would just like to fly it,
it seems like fun. But would you ever consider putting small weapons or explosives on the drone?

(50:39):
Make it all sense. I mean, we've definitely been, yeah, we've definitely been asked because I think
our, the design we have is like perfectly suited to it. So I didn't say but this drone is,
so it's fully through the printed, it is printed in SLA, so it's like a plastic. And it's a plastic
that is very not thermal conduct, so many conductive. So what that means is it's the heat that is

(51:02):
generated by all the components just stays inside. So the aircraft has no very little thermal signature.
The aircraft is fully again, plastic outside of the batteries inside. So in terms of radar
signature is also pretty small. It's like much smaller than, and like, a carbon fiber aircraft or
something. And it's pretty small. You can see from the size and it flies 100 miles per hour.
So all of these things mean that it's actually very difficult to like see where is thermal sensors.

(51:28):
It's really hard to see on radar. It's very hard to like see visually. Also pretty silent because
it's electric. So it fits like lots of boxes if you wanted to use for an offensive capability.
And also because through the printing and the size, they're really inexpensive compared to
what people, I mean, drones in the military that size would cost like 200,000 pounds and

(51:48):
they're much, much cheaper for us to manufacture because of the choices we've made. So,
so yeah, there's definitely an application. We've been asked about it. In general, that's not really
something we're interested in. Because especially nowadays with all, like, more conflicts happening,
and it's way more polarizing, like, not everyone agrees on what's right than, you know,

(52:09):
with different conflicts that are happening in the minutes. So it's just, it's not something
we're interested in. Like, it's way more exciting to vision. I've talked about it. Like, how can you
collect that data set at a large scale? How could you empower, like, you know, civilian and military
organization to know exactly what's happening at all time? And maybe you could have a much more,
and bias view about like, you know, what are the events of this country doing stuff around here,

(52:29):
or stuff around here on, and I think that's some more interesting things for us to build. Yeah, and,
and I totally agree. I know it's really hard being a CEO. Like, it is kind of a moral judgment. Like,
can you sleep at night knowing that something you built may be killing civilians and stuff like that?
But in the market are, what if you have a competitor that comes out and they do have weapons on it,

(52:50):
and they eat your lunch? Are you okay going bankrupt to that?
I don't think that would happen because it's, I think, I mean, there are plenty of companies in the
drone space that are going after that at the minute. And we don't really care because I think
long-term that's not going to be the big market. I think people are going after this now is what,

(53:11):
because one of this, like, bigger budgets, so, you know, it's easier to make money.
It's also like a big hype, so you have actually a lot more like a private capital going into
funding stuff that is for defense. So it's very easy to go into this right now. But it's also very
short-sighted because this is going to be quite cyclical. So yeah, it might be great for the next five years.
So, but it's probably going to go off afterwards. You're also very dependent on politics and

(53:33):
whoever decides to buy your stuff. And also, even longer term, like, the trend in general has been
that R&D is very updated in defense. It's like the best stuff is not in defense. It's in commercial.
If you look at satellites, sure, they start in defense, but like most of the expenditure now is
from commercial, so probably the biggest market in the long term are not going to be defense. So

(53:54):
today it sounds exciting. So we're trying to do some R&D that is related to defense, which is like
mostly like surveillance, providing the insights and everything. But I don't think there is a
there is a compelling outside of there is a compelling use case to go heavily into defense and
like offensive aspects outside of like short-term profits and going down like a rabbit hole that

(54:16):
really doesn't need too much in the future. So I think even like, yeah, from a purely
business side of things outside of like the ethical element, I don't think it's the right thing to do
anyways. Yeah, I agree. And it sounds like you make a compelling case that, hey, there's a market
that doesn't require weapons. And it's a big market and you can sustain your company times 10 with it.

(54:38):
And you can also help people without blowing them up. Speaking of blowing up people.
What do you think? So like, you're right, there is lots of conflicts going on around the
world and stuff like that. Something as, you know, a technologist or whatever you want to call it,
a nerd, myself, I really don't quite understand why we still have boots on the ground fighting wars.

(55:00):
The technology is out there. The technology has been out there for years to essentially like
have like a drone warfare, right? Why would you ever like a drone will beat a boots on the ground anytime.
Me versus a drone, I stand no chance, right? Same thing like even against boats and stuff like this. I
don't quite understand why we still have people with guns and hands whenever we can build powerful,

(55:26):
cheap drones with weapons on them to fight these wars. What's your opinion on that?
Well, I think there are big issues with drones still today. One of, so if you look at Ukraine, for example,
like where you constantly have both sides thinking, how can I defeat drones? Because drones have taken
a massive, yeah, they have a massive role in the world now. So everyone's trying to figure out

(55:52):
how do you disable them? And there are plenty of very effective ways and there's plenty of research
now trying to counter that. So for example, drones all like the main source of data they rely on to
know where they are and figure out where they need to do to get somewhere. It's all based on GPS
and it's pretty easy to jam GPS. So if it's pretty easy to jam, you can actually jam GPS in such a way

(56:16):
that it's not just the drone doesn't know where it is anymore. You actually make the drone believe
that it's somewhere else and you can actually backwards fly the drone this way. So these, so that means
it's not just straightforward like, oh, drone works in your backyard when you're flying like a quadcopter,
it's just going to work in an military airspace. It's not like there is this. There is also,

(56:38):
I talked about that line of side communication stuff. We're using like a radio transmission between
the drone and the controller. Easy to jam as well, that stuff. Also, if you were to go to
your crane and fly a quadcopter next to the front line, people would detect you're there because you're
emitting signal. And so you'd get you'd get shot at. So it's not straightforward. And there's

(57:01):
plenty of reason why it hasn't turned out like just to you know, drones versus drones. Yeah, so that's
probably that was probably one of the reasons. But I guess as well like drones right now, I can
mean, maybe it's different when you're talking about Gaza for example, where there's no really
opposition. Like there aren't people that have on the ground like much, you know, anti, you call that

(57:25):
anti electronic warfare like so things to like jam GPS and things. I would assume this is probably
quite limited. You're still constrained by like, you know, people are in buildings. So you have to
go into buildings. You can't really go to the drone, I guess. That would be the perspective. Yeah.
Yeah, that's actually a really good point. I didn't know about the GPS jamming technology and some

(57:47):
of the other stuff. So that kind of sucks because if if we go in a war, I am not trying to go in a
in a trench with a gun. I would much rather it'd be like video games. Do you think one of your drones
has been like mistaken for a UFO? Are you familiar with some of these weird like alien UFOs that are
like often seen by pilots over like large bodies of water? They say that they go into the water and

(58:12):
out of the water. Like, have you seen any of these with your drones? And if you have seen them,
like, what would you do? Oh, interesting. So I haven't heard of like the one you're talking about,
or like something coming in and out of the water. I mean, the stage are which we're at. It's not yet
we're not like scanning the whole planet every five minutes and knowing everything,

(58:35):
what's happening everywhere. Like even though I think long term, whether it's also like other
manufacturers like drones will eventually get to that. I'm pretty sure. So we don't we don't we're
still deploying in a super limited fashion. So it's not like with detected stuff like not at all.
So but like could a drone be be mistaken for a UFO? I mean, probably. I mean, ours are pretty

(59:02):
small and pretty fast. So you might but but they do they do look they still look like planes. So
yeah, I don't know. Yeah, I mean, I'm for sure you could that we could design something that looks
like a UFO behaves weirdly. And I'm pretty sure we could do that. Yeah. And you should put like
like an alien face on it, paint it just to fuck with people. That would be awesome. I'm sure this

(59:26):
is why I'm not the CEO and you are. Have you been keeping up with like some of the issues and the
quality issues that Boeing is experiencing? I mean, yeah, I've been seeing the news in the
background. I haven't been looking into crazy amount of detail. Yeah, I mean, my understanding is
like so I used to work in the Airbus. So you know, like the the main antagonist, I guess.

(59:52):
Do you fly Boeing? Oh, yeah, I mean, yeah, you don't really have a choice anyway. Like,
you wouldn't go like discard some airlines based on like the planes are flying. And also, I mean,
when you're talking Boeing, like all the issues recently has been on the specifically on the 737,
right? And mostly like the most recent ones. So I think the main you don't in Europe, we don't have

(01:00:15):
that many 737s because those smaller aircraft, the equivalent from Airbus, the A320 is super popular
in Europe. So you outside of 102 airlines, you don't really get to on the plane. So I get, I don't
I don't have to ask myself the question that often. And on the bigger planes from Boeing, they've been
flying for a very long time and the, I don't think there's been many issues. So, but yeah, I guess the,

(01:00:36):
what's from what I understood and it's an interesting thing is it's not, it's like an overall
company cultural issue. Yeah. And my understanding is that Boeing used to be a very much like engineering
driven culture a long time ago. And the whole defense because of the Boeing is partially a defense

(01:00:56):
company like that, that holding aerospace industry, the same in Europe, they after World War 2,
when you had tons and tons of company over, like, you know, the following decades, all of these
company ended up merging with each other and becoming those bigger groups. So, and Boeing in particular,
I can't remember one exactly, but at some point they merged with McDonald Douglas. And that company

(01:01:17):
apparently at the time had a way more like financially driven culture when Boeing was very much
engineering driven. And the way they did that merger, even though it just stayed remained Boeing,
like you don't, I don't even know if you know about McDonald Douglas, but the, the executive for McDonald
Douglas took over a lot of their, like the executive position in Boeing and took over that kind of,

(01:01:40):
like, financially driven culture into Boeing. And I mean, you know, you can imagine in a large company
like this, for cultural changes to take to, to be effective, it takes like decades. But I think what
we're doing, what we're seeing now is like, there is a lot of that merger and how, like, it just
changed, going from an engineering driven culture to a, you know, bottom line trying to optimize

(01:02:01):
and make more money from, from like existing planes. Yeah, I think it's, what, what I said at the
beginning, right? It's who's winning, right? The engineers of the managers. If with your current
understanding of the Boeing situation, how would you go in and like, what would be the first thing you
would do to try to like fix that culture? I think you'd probably need to spend a few years just

(01:02:23):
going around and trying to understand what's happening because it's like, yes, I mean, I don't know
how big Boeing is like, Abbas is 150,000 people. And I mean, that's ridiculous. Like, how do you,
how do you implement a culture at that scale? Like, yeah, I don't know. So, I guess you'd,
yeah, you'd need to figure out what's happening. Like, so I don't think there's like a simple solution.

(01:02:47):
Yeah. Yeah, no idea. Yeah, I don't either. And especially I would imagine, right? You just come in and
you say, hey, listen, all middle managers, you're fired engineers. You need to think for yourself.
We want to hear from the engineers. We don't want to hear from managers and board meetings.
But it sounds like Boeing may have painted themselves in a corner by, you know, being so profit-driven

(01:03:12):
profit-driven. And now you say, how do you go to your investors? How do you deal with the stock price
falling and stuff like that? Whenever you say, hey, look, we're not going to worry about the bottom line
so much. We're going to fix our engineering culture for the next five years. That sounds like a
recipe for like almost bankruptcy or something like that. So it really should be like a warning

(01:03:35):
to people, hey, listen, you know, profit-driven is great. But whenever you're in an industry like
aerospace or aircrafts or anything like this, you have to listen to your engineers to a certain extent.
It's not even like profit-driven because if you look right now, like the financial results from
Abbas are much better than Boeing. So it's not like you haven't achieved your desire to financial

(01:03:56):
lot of them anyway. Yeah. So I mean, you can't, I don't, a big reason. So yeah, I used to work at Abbas
and I didn't like it because it's just not what I'm good at. And I like, you know, with having like
deep technical discussion about what the future looks like and like doing quick in iteration and
like being comfortable with crashing and things and like figuring what is the next race through

(01:04:18):
or whatever. Like that's very different to the culture you want to have in an abbas or a Boeing
where they're doing like small iterative improvements, but they need to do that in such a way
where there's no room for failure. And so you need a very different culture. And a lot of it
is being based on processes and following those processes that have been set up and like understanding

(01:04:40):
why they work well or whatever. And so I assume they must have all of that still working in
Boeing. I mean, you know, even despite all the failures, it's still like, you know, it's not like
every aircraft's crashing. Like there's still plenty of things that are right. So I don't think you
need to like, you shouldn't take drastic decision in like removing people and stuff. I think it's
probably a long-term process. And especially as like what might be boring of like having good

(01:05:05):
structure and everything is what makes it safe in the first place. So, so yeah, I think the it's hard
to know is not knowing the details of in those specific departments where they've had the issue
on like specifically the 737 MAX program like what there are things that were different between
that program and another aircraft like yeah, I don't know, but you need to know the details I think.
I would love to be a fly on the wall, right? Just to be in there to see like, hey, what's going on?

(01:05:29):
What are these board meetings like? You know, what are we talking about? What's our objective? What's
our KPIs all of this type of stuff? Because man, if I was running Boeing, I would shit myself if I
cut a single dollar from our quality assurance budget, right? Yeah, but I don't, I thought that it's
as simple as that as that might be like, you know, you create incentive for managers to be having a

(01:05:54):
specific way, which mean later down the line, they've changed something and then like, you know,
three layers down that you've had that impact. But to be fair, I don't know if that the top level
they end up cutting those budgets. Yeah, I mean, what's interesting though is that if you look at
Abbas in particular, so if you were to buy from Abbas, like the smaller airplane, like the A320,

(01:06:18):
which is the equivalent of the 737 from Boeing, I don't know how much exactly the number now,
but there's a massive, like, there's a massive lead time. If you were to buy an aircraft today,
maybe you get it in five years. I mean, don't quote me on the number, but it's like, it's really,
it's pretty high. The reason for it is there is way more demand and supply in like,
planes at the minute, especially on the smaller ones, which is where like Boeing is fucking up. So

(01:06:42):
as a society, like the whole planet, we can't afford like, buying to fuck up too much because we,
there are, there's only really two companies that are supplying the market. There are quite a lot
of others that are doing it as well, but like the resmal, and virtually, so you need, everyone needs
Boeing. Even, even by Abbas, I think if Boeing was completely fucking up, they would be like,
how can we scale up? Like we can't scale up. Like, yeah, the main thing Abbas is struggling with is

(01:07:05):
like, how can you increase production for these aircraft and demand for travel is continuously going
up. So, so it's an interesting situation, but it's not even like everyone is like, oh, this is so good
for us. You know, we're going to make a good business because Abbas Boeing is not doing well.
Everyone needs Boeing to sort it out. Why do you think there isn't, where's the third competitor?
Why is it only two? Because it's very difficult, I think, like the complexity, I think we get used to

(01:07:30):
it. I feel like, oh, yeah, we're just getting on airplane on its works. It's like, I mean, and we're
like excited with like something new happening in AI where, you know, two people can do it and
like you use some existing libraries and like, if you think about like building a plane company
from scratch is like, it's mind-boggling. We just, we, again, we forget about it because you just
so used to it, but the level of complexity and safety you need to have in place for these is just ridiculous.

(01:07:53):
So, so it's just really hard. And so I mean, the other companies are doing it. There is one in
Brazil called Embraer. There was one in China called KOMAC. So China has been investing massively
in this for a very, very long time. And eventually they'll get there, but it's just, yeah, it takes
takes forever. So, so that's why it's not easy. Super difficult.

(01:08:14):
Damn, it must be difficult because we've had like Boeing and Airbus for quite some time and nobody
could break through. So credit to them. Whenever you talk about your own drones, why did you choose
to go battery powered because batteries are incredibly heavy? Yeah, it's true. I mean, it's,
so yeah, it's not just the way it's like the specific energy. So it's like, we get less energy for

(01:08:35):
the same weight because ultimately we can put more weight on the plane. Where did you go batteries?
One is like fundamentally not wanting to use internal combustion, just like sustainability aspect.
But actually the main reason is we're thinking about everything we do is for that longer term.
The way we believe longer term drones will be implemented. And if we say longer term,

(01:08:58):
it is a robotic infrastructure where people occasionally go to those hangars to do a bit of maintenance
and everything, but otherwise it just works by itself because you want to take people in the out of
the loop completely. That's how you reduce the cost. In that scenario, you don't want internal
combustion engine on those drones that need to be like frequently maintained because if there's
like internal combustion engine, it's pretty bad reliability. Also, what do you have like a

(01:09:21):
automatic like spot refueling on the drone or something? It's like it doesn't feel like the future.
The future is like batteries that get swapped or maybe the drone lands and it just gets recharged.
That's just that's yeah, that's just I think I can't imagine the future being internal combustion with
like robotic nozzles coming and start refueling the drone. That sounds ridiculous.
How do you balance long term future with short term sustainability because a lot of companies,

(01:09:46):
small companies go bankrupt while thinking about the future 10 years from now?
Well, it's like balancing, you know, within the roadmap that you need for that future implementation,
what are the things that make sense to do today technically and what are the things that are
going to be beneficial for customers today? And then you start with these ones and you try to

(01:10:08):
minimize the amount of like side quests you need to go on to that maybe are necessary, but I like,
you know, long term that side quest is not really beneficial. It's not like the main story,
but you still need it. You're happy to take it because short term you need it to sell to customers.
So that sounds pretty simple, but yeah, I mean for us, it's like, you know, specific features we
focus on at first. So it was, for example, just doing the aircraft, not doing the robotic hanger

(01:10:32):
and everything because today it makes no sense because you need a pilot on the ground from a regular
history standard. If you need a pilot and you have a robotic hanger, you're just going to have
the robotic hanger, but you still have the pilot next to it. So you won't get the benefit of
having your fully automated system because you'll still need someone next to it. So that's how you
trying to pick up what makes sense early on to develop. Yeah, with the side missions, that's
such a great way to put it because I used to work for a very small company too. It was a startup

(01:10:56):
and our CEO would go out and they would meet with maybe a potential client. And our CEO would sell
the client, say, hey, listen, do you want this? You want this? And he would sell the client
on a feature that we didn't even build yet. It was just, he'd be like, yeah, man, hey, I just sold this.
I made $35,000. We have to have this built by next week or the check is going to bounce, right? Is

(01:11:20):
that the type of stuff you have to deal with? I mean, you constantly have to sell what you're trying to
build as a startup because you have to paint that vision ultimately and you have to balance that
between, yeah, you have to balance selling that and what you have available today, like it's for sure.
How, like, I don't know what's with, I mean, ultimately, especially for us, we're trying to sell

(01:11:49):
big government contracts. So it's not like, you know, I don't know what you, the product you were doing
before, but let's say if you were selling to, you know, like, be, to consumer or maybe to like, small
companies that are probably like, there are billions of people, there's probably millions of small
companies. So if you, if you were to sell something that you can't deliver to someone and you
fuck it up, then you're like, what's okay? There's like a billion other potential customers or a

(01:12:12):
million other potential, who cares for us? It's very different because the pool of, you know,
governmental organization that need to do my time monitoring for like, fisheries protection or whatever.
It's like a pretty small pool. So if we start fucking it up, it's, yeah, we're done basically. So,
so we have to be more careful about this, but that the, because the industry is in material, most of

(01:12:33):
these organizations will have seen other people selling them stuff before that didn't work. So by
just having, you know, good communication where you say, look, guys, this is what we're building.
Ultimately, this is what we want to sell to you. And sometimes we show renders and people are like,
wow, we saw this was science fiction. So, for example, you know, the robotic hanger I've talked about,
we're talking about putting that on the vessel, but a robotic vessel. So it's kind of like a robotic aircraft

(01:12:54):
carrier at that point. So you imagine a small robotic, like a 25 ton, so small-ish vessel hydrogen
powered because out in the Atlantic and then you have the robotic hanger on it and it just flies
the drone. So I say, and sometimes we show these vision to those people and like we, we're clear
that this is not what we have now, but we're going to have that in, you know, a year and a half,
and you can use that narrative to get them involved in the product we have today. So I think it's

(01:13:17):
possible to navigate between the two, but okay. Yeah, that was a key distinction. I really didn't,
and you're so right, and you nailed it, right? Whenever you do have a very small client pool,
you can't always be flying by the seat of your pants. You know, you have to make sure because,
you know, it's a, it's a small group of people and I'm sure they talk, right? And I'm like, hey,

(01:13:38):
this guy Matt has company sucks, you know, his drone crashed and next thing, you know, your reputation
is ruined. Whenever you talk about the decision to go with batteries, we know that batteries do not
like cold weather. Does that limit your ability and your range in cold weather? No, yeah, there's definitely
something. Interestingly, most of our operations right now have been like in Africa and so in

(01:14:01):
South America, in Central America and Caribbean. So it's all been like actually really warm weather.
So we haven't even done the mass of how it affects us, but for sure would affect us.
There's plenty of other drone manufacturer that I've solved this by having, you know, some
heating of the batteries. We do have quite a lot of like excess heat on the plane when we're flying

(01:14:26):
like the motors generally, you know, based on the motor, maybe like 90% efficient, still means
like 10% of the energy you're like you're using is gets out in a heat. So there must be a,
there's probably a good way to like heat up the battery, but for sure it'd be in the defectors.
I was just thinking another thing on the, what we're talking about earlier, the,
like, you know, the, how do you manage the, like being ambitious and like selling the next thing when

(01:14:49):
you, but at the same time, you know, you need to be able to then ever that next thing and it's hard
to know in advance in advance how reliably you're going to be able to deliver what you think you're
going to be able to build. I think it's also important to be kind of ambitious and like really pushing
yourself and, and, and telling to a customer, we're going to have this because then if you, if you do
that, you have a legitimate deadline for in, in, in the company that motivates yourself to try to

(01:15:14):
hit something because if you don't set that, that kind of like, it's like you set an external constraint.
So like a customer, if sold something to a customer, so no, it's not just internally you're saying,
oh, we should have that feature by then because we want it and that's our objective. Now you've said,
oh, we sold that to this guy. So we need to have it by then makes it way more compelling to like,

(01:15:35):
work crazy hard and like be super motivated. So it's, I think you work as a company, you're more
effective and it's, it's just some way more fun, like you're more motivated to work towards something
when you've committed to someone externally. So that's also an important factor to consider. I mean,
I don't know how it was when you talked about your, your, your company when you're, your CEO might
come back and say, well, I'm just 35k for this feature. We need next week. And if it's next week,

(01:15:57):
maybe it's not for beneficial, but, well, it's motivational sometimes to have like a crazy deadline or
it was always, uh, fuck that. It was another, yeah, yeah, I'm getting paid anyway, fuck it. Uh, no,
and you probably know this whenever you work at a startup is, it takes like a different type of
person. Um, if you want to punch the clock, go home to your wife, watch TV, maybe a startup's not

(01:16:19):
your jam. It was totally my jam. Oh my gosh. I loved it. Like I, my favorite days working at this
startup was like the days where I would stay up till 12 p.m. or 12 a.m. in the morning, just coding
because it's like, I have to get this done. I have to get this done. It was extremely
motivated. And so like the CEO would come in and you know, we had like, we had like 10 people in

(01:16:41):
the company, right? And he would gather everyone around and he would say, Hey, listen, I just talked to,
um, like, uh, whatever it was a sensor company. And we just sold this feature. We need to have this new
mapping feature UI and they're going to have these types of sensors overlaid on it. And he would
just give us details is and maybe we would do one it for an hour. And then all the technical people

(01:17:04):
would come be like, how can we make this happen? And it was like, it was very, it was like almost,
it's kind of like whenever you talk to people in the military and you say, Hey, what was the best time
of your life? It was like the time I was at war. It's like, that makes no sense. You're about to die. But
the time that I would work like 14, 15 hours a day, where my favorite times, I love that shit. Oh,

(01:17:27):
man, challenge me because I want to see like I want the target and I want to try my best to hit it.
And I will do it at all costs. And if I fail, I may cry. But if we succeeded, which we did most of the
time, um, we would always celebrate afterwards. And it was it, it, it, working at a startup for me,
more felt like just hanging out with buddies in a garage than it did work for a company. And I love that.

(01:17:50):
Yeah. No, I think it's like it gives you that sense of purpose, like very clear because if someone
buys a feature you haven't built yet, it means like, you know, they actually want it. So it's like,
what I'm doing is you is generally useful. And also when you're in a startup, like you said, you
know, 10 people will also release more at the minute. It's, you have a direct connection to that
deliverable. It's not like when you're a bigger company like you're in a bus, you're one person in 150,000,

(01:18:15):
you know, your, it's hard to see how what you're doing is contributing to anyone. Um, so those two
things are like, you know, it's going to be helpful to someone who has paid for it. And you have you,
yourself have a very tangible impact. I think those two things, yeah, makes it to be exciting. Yeah,
super exciting. And also the coach at a startup, you just kind of do whatever the fuck you want. Like,
I, like, it's in a lot of big companies. You know, you kind of have like these handcuffs where

(01:18:38):
like these managers, hey, you can't touch this. Don't do that. At a startup, it's like, we need this
and I don't really care how you get it done. We just need it done, right? Obviously you don't want to
be on, I think, right? You don't want to like, right, uh, software with no tests or you don't want to
like do things like cut corners, but you could do anything. Listen, if you want to go work on the

(01:18:58):
rooftop and that's your jam, go do it. If you want to like use some like different library,
whatever, you don't need management approval. It just feels like really, really natural to me.
Did you enjoy working at a big company? No, not really. I think it's, um,

(01:19:18):
I mean, I think it was super valuable for two reasons. One was, you know, when you,
it's useful to have like a stamp of approval. Yeah. That it, like an external validation of your skills
when you're building a company, especially early on. Um, so having work at Abbas, which is still like in
Europe, you know, that's the biggest like aircraft manufacturer, kind of difficult to get a job

(01:19:41):
there. So it was, it was positive to have worked there from that stamp of approval, even though I
don't think I learned that much. Um, the other thing I was useful is the one, I actually went to
Abbas for one very specific job. I just didn't want to be specifically in Abbas. I wanted to be in
one department, which is called future projects. So that's, uh, I don't know if you know the Skunk
Works in like, you know, Locking Martins, like the terms Skunk Works. And so it's kind of like the

(01:20:04):
equivalent, it's like the secretive department, who that does, you know, the cool aircraft concept
that Abbas might do 20 or 30 years from now or might not, but you know, they have to constantly
figure out what could be done because if Boeing does something, you know, we need to be able to,
like quickly respond to it. So you look at the coolest kind of concepts. Um, and it's also typically
the most passionate people in the company because they're like into that stuff. So I always wanted to

(01:20:26):
get that job, but it's very hard to get because it's like maybe 20 people, you know, out of the 150,000
and a lot of people want that job because it's, it's a cool one. Um, so, so yeah, what I learned was one,
like having the, what was valuable is one, the stamp of approval and two, because I wanted that job,
I had to try to get it. I applied 10 times. And I kept getting rejected. So it like really taught

(01:20:48):
me like, you know, resilience and, uh, and just keep going at it. And like, you know, I didn't
care if I was being rejected, you know, you tend to, you just assume, well, I'm not good enough,
right? If you get rejected. So it's like, how do you fight against this and be like, no, I'm just not
that I'm, I'm just going to work harder next time and do it again on you. Do that like 10 times in a
row. Uh, I think those skills are very transformable to building a business because constantly, you know,

(01:21:08):
it doesn't work like the only way it works in a business is you constantly hitting your head against
like problems and you keep going anyway. So, um, so those, yeah, even though I didn't like working in
a big company because it's not for me. That's just not my personality. Um, yeah, I definitely,
it was super useful. I think it was a complete, a central to what I'm doing now. Yeah, big companies
are cool. There are some things you can learn and see kind of like how they operate and why they

(01:21:32):
operate that way. For your company, Marble, what do you think your company looks like five years
from now, assuming that you are successful? I mean, I told you like the, the vision of where I think
drones are going to be. So I think we're going to be on, I can, one of the big companies that implement,
like one of the companies that makes that happen. So, yeah, again, that vision is having networks of

(01:21:56):
robotic hangers deployed mostly on coast. If you're talking five years from now, later on,
we'll just be everywhere. So along most coast on the planet, you'll have a robotic hanger, maybe
every 150 kilometers. Each of them will have, you know, three aircraft operated from each of them will
just have like, you know, stalling or an equivalent straight from it. So they will be collecting data
in real time across the whole world. On all these coasts, they'll be aggregated in one location.

(01:22:19):
Anyone can get access to that. Just you pay for the data that you consume. You can also task to
drone compared to a satellite. You could actually say, you know, in five minutes, I want this data
over here and you could actually just get that concrete that was satellite. So we will be the
company supplying that, we'll have built a hardware and will be selling that data as a service.
They'll probably be a lot of additional layer of intelligence on top. We talked, we touched a

(01:22:41):
little bit on like artificial intelligence, computer vision. So yeah, five years from now and we'll
probably have some of these not just on the coast, but also on vessels across the, you know, a bit
further out when you want data, not just like, you know, on the coast, but like a half-way across the
Atlantic. I think I am going to buy lots of your data and I'm going to try to find the Lochness

(01:23:03):
Monster and I'm going to be very famous after I do that. Matt, thank you so much for coming out on
the pilot. How can people find out more about you and your company? What the best, I mean, is just
our website. So it's, you know, marble.aero. And then yeah, just on LinkedIn, that's probably where
most active. That's, yeah, and if you want to get in touch, it's contact@marble.aero.

(01:23:26):
Awesome. Once again, thank you so much for coming on, Matt. You're super smart. I can tell every single
opinion you have you spent many hours thinking about it and always appreciate that, man. Thanks.
Thanks. It was great to have a chat with you.
[Music]

All Episodes

#37 Mathieu Johnsson | The challenges of building maritime drones, engineering culture, the future of war

Episode Transcript

Popular Podcasts

Dateline NBC

Stuff You Should Know

The Nikki Glaser Podcast

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}#37 Mathieu Johnsson | The challenges of building maritime drones, engineering culture, the future of war