How a filmmaker became a military drone innovator: Meet Vermeer's Brian Streem

Hi, I'm Jen Judson, land warfare reporter for Defense News, and I am here with Brian

Stream of Vermeer, CEO of Vermeer. Brian, talk to us a little bit about your background and how you fell into the defense industry

Yeah, sure. Thank you for having me, Jen. So my background's a little strange

I went to NYU film school. I studied cinematography, movie producing, and for a while I used to, you know, I produced

independent movies and TV commercials and things like that. And I kind of got into this via strange

method. In 2013, I saw a little video of a little drone flying on YouTube. And I kind of thought

damn, that'd be really cool to put a camera on. This was before DJI Phantoms had cameras

And I ended up buying a very big, expensive Swedish drone with the idea of, you know

kind of renting it to my Hollywood film friends and filming movies with it. So I ended up buying

a big expensive Swedish drone. The drone crashed maybe the first week of having it. And I, you

know, fortunately I got a very large check from the insurance company. And then I bought a strange

Estonian drone. And then from 20, and then I was one of the first legal drone operators in America

I had a 3-3-3 exemption. I think I was number five or six in the country

And then for a couple years, from 2014 to 2018, I probably had the biggest drone cinematography service company in the world

We did Spielberg movies. We did Marvel movies. We did movies that won Best Picture at the Academy Awards

And then it was a very brutal, terrible business. We were making decent money

but like, you know, logistics problems, regulation problems. It was a service business

We were using technology, but it was hard to scale and duplicate

So I started to build software to essentially help drone cinematographers get more precise drone footage

And I essentially one day I fired my entire drone service company

except for the one or two software engineers I had. and I started to build this software product

And then in 2019, I was sad and desperate in Buffalo, New York

That sounds like a movie. It sounds like a common scenario. And I met a guy who worked for Techstars, which is like a technology accelerator

And Techstars was hired by the U.S. Air Force essentially to find innovative technology companies

and he told me, I showed him what I was doing. He told me that the Air Force would be very

interested in what I was doing. And I kind of was like, what are you joking? Like, doesn't the Air

Force have this already? Like, if you're coming to me for help, like we're in a lot of trouble

And he said, you know, you would be surprised. And he told me that the application for his program

closed in 20 minutes. So I got up from the table, filled out this application. And then a month

later, I moved to Boston from New York, and I was part of this U.S. Air Force-funded Techstars

program where they essentially took me and nine other companies and it was designed to inject us

into the bloodstream of the DOD. And that was kind of my foray of how I got started

All right. I mean, that's very interesting. We often hear about how hard it is for, you know

small, non-traditional companies to get in, um, in this way. So it's nice that you were approached

as opposed to having to, you know, knock on many, many doors. Um, yeah. Well, I mean, I did have to

knock on many doors. I mean, I, when I, when I first got here, when I first got into Boston at

this program, uh, I realized that, okay, like, you know, the DOD has a trillion dollars. How do

how do I get them to give me some of it? And I have a very simple rule of sales

which is you can't sell stuff to people if they don't know you exist

So what I did was I realized every DOD email ends in dot mil So I went huh that interesting And I built a tool that scraped every email address off of Google

and I sent 50,000 cold emails. I think we all need that. Sorry

I think we all need that. Yeah, and then for the next nine months

I had about 2,000 cold phone calls with people at the DoD. I mean, it was brutal

Doing that amount, you know, that's a Monday through Friday, 9 a.m. until 6 p.m. every day

that's all I was doing. I was just kind of introducing myself and I was asking people

what their biggest problem was. And everyone told me their biggest problem was GPS denied

navigation. And this was in 2020. So this drone cinematography tool I was building

I kind of thought to myself, huh, I think if I tweak it a little bit and I talk about it a

little bit differently, I actually think I have a way to solve this GPS problem

And my first year, I probably won like six or seven million dollars of like non-dilutive capital from the Air Force

And then, you know, two years later, a big land war broke out in Europe where the first thing to disappear in Ukraine was GPS

So how did you get your capability into Ukraine? I know that that's your next step here

So talk about how you managed to get into the country with your capability and get in front of the right people there

What do you do? You take a train. I mean, like, you know, I had, I think part of, like, my thing is, like

because I'm an outsider since I don't know, like, the typical path

I'm naive, and my naivete is somehow, like, my superpower, because I'm just like, well, I don't really know

Like, maybe we should, this thing we're building for war, maybe we should go test it out in the war

But, yes, you could fly to Warsaw. I was at the time, I was in Riga, Latvia, doing an integration with a company

And I decided to fly to Warsaw, got on a train, and I took a train from Warsaw into Kiev

Didn't know anyone. And I was just, I thought, okay, this is easy

They don't have GPS. My thing provides GPS. I'll be there for two weeks. I'll make a zillion dollars and I'll go home

I've lived there now for two years. I have a team of five or six full-time folks there

and yeah when I first got there didn't know who to contact so what did I do

I went on LinkedIn and I searched for hashtag Ukrainian defense and other things

and I started to find and I would just reach out cold to people

if you don't know what else to do reach out cold to people and see what sticks

can you talk about the type I know that there's some things that

you can't discuss this capability is being used in operations in Ukraine. But talk about some of the lessons that you are learning and how that is

you know, causing you to adapt your technology and think about your technology and what's kind

of next for you when it comes to this capability that you're developing. Yeah, I mean, I think

I think kind of like broad strokes, it's very hard to kind of, I think my big lesson from Ukraine

like just, you know, from a high level is like fighting a kinetic, like actually fighting a war

and also trying to innovate at the same time are just two brutally challenging things to do

You kind of need to have the innovation already there and the new tools

If you're working with stupid, simple drones and then trying to build AI computer vision on top of it

you're falling behind. It's a very bad situation. It's just too much stuff to do

There's not enough hours in the day and stuff like that. in terms of kind of just like, you know

you know, so we build a vision navigation system that helps drones navigate

where there's GPS jamming and spoofing. And the reality is it's like, you know

it requires doing this a lot. Like you need to like, you need to like

you know, fire drones a lot. You need to then collect them. You need to like look at data

You need to like test them in like real jam. Like in America, like, you know

oh, are you going to be at White Sands for the experiments

Oh, well it was in June, you missed it, but there's always next year

There's always next year. It like no we need to do this five times a week to figure it out Building products like this actually work like yeah they take 10 years to develop sometimes the way we do things but if you do it in a place

where you're able to like fire drones constantly at like you know high speeds low altitudes high

altitudes you could just consolidate compress that 10 years into six months if you just like

do it every day. So it just kind of like the, the doing it a lot just yields all sorts of benefits

What's the feedback from users in Ukraine, um, when it comes to using what you have, I mean

how is this serving them? Is it, you know, helping with targeting and what are some of the things

that have, have been useful when it comes to... I mean, they are, uh, they're, you know

very, very pleased with the results. Uh, they wish, they wish it were cheaper. Um, but you know

I mean, uh, there's kind of, I'm kind of like, do you want it to be cheap and not work or do you

want it to be a little bit more expensive and work? Describe what this capability is. Can you

talk about what does it look like? Yeah, sure. And you know, without divulging your secret sauce

you know, what is the capability? Yes. Yeah, sure. So my, my company, we essentially build

So the problem is that in a conflict zone, Ukraine, Russia, or the Chinese have sophisticated abilities

They essentially have satellite dishes they can put on the ground that essentially pump the environment with fake GPS signals

And all of our missiles, all of our drones, all of our ground vehicles, all of our UGVs, USVs, they all use GPS

But all these systems are essentially designed to essentially spoof it, meaning manipulate the reality of GPS

So you think you're at this altitude, but really you're about to crash into the ground or they just turn off GPS entirely

And they're doing it. We're doing it. So like at the front, there's no GPS

A lot of systems require GPS to initialize. So like we could send some of America's fanciest systems we could send over there

But those fancy systems usually require GPS just to turn on. So you can't even turn them on

So essentially my company, we build like a little box. The box has two to four simultaneously running daytime or nighttime cameras

Those cameras go into a little NVIDIA computer with our software on it. Attached to the computer is an IMU

and attached to the computer is a little hard drive that has a previously acquired 3D terrain satellite map database

or celestial database. In this case, we're doing mostly terrain navigation. So essentially, as the drone flies through the environment

the cameras see the terrain, and in real time there's a deep neural network that matches what the camera sees

to the previously loaded 3D terrain map database. It's like facial recognition

When you buy your iPhone, you take a couple photos of your face and five years later, you're older, you're fatter, you're wearing sunglasses

you're growing a beard, it still knows it's you. So even if there's a lot of change in the terrain

it's still able to kind of figure out your position. And you can't jam it, you can't spoof it, it doesn't emit or receive any signal

it works daytime, nighttime. We have a box that goes on top that essentially looks up at the position of celestial objects for overwater navigation as well

And the applicability of that is obviously for not just drones, but UGVs and USBs as well

You've had a lot of experience in Ukraine. Is that helping you make some headway, say, with the U.S. military in terms of getting that back to them

and seeing if they're more up for taking on your technology now that you've spent a lot more time

fleshing it out in an actual operational environment like this where it's really challenging

Yeah, absolutely. I mean, it's strange, but living over there for two years with my team

I mean, we probably just have more expertise than anyone in the world

barring maybe a few other people, on kind of integrating vision navigation system

into aerial platforms and actually deploying them in a jammed and spoofed environment

So yes I working on several programs at the DoD where they have told me well we looked around and we tried to figure out who the hell knows anything about this stuff And your name keeps on popping up And I also have you know probably 40 plus customers A lot of those customers are uh either other nation states or you know

U.S. drone manufacturers. Um, so, but, you know, being, uh, being in Ukraine has given me a lot of

you know, it's excelled my, you know, my, you know, my technology's capability, uh, shown that

it's effective and also just like, you know, doing this stuff is not easy. I mean, we now

we know how to walk the walk and, you know, that's important to customers that are building

large drone platforms with, you know, billion dollar programs behind them. What are the opportunities that you're eyeing when you see, you know, we're at an army conference

So what are some of the opportunities that you're eyeing that the army may be pursuing right now

where this would fit? I mean, there are probably dozens of large

drone programs right now that the Army is pursuing that are loitering munitions

missiles, one-way attack drones, ISR drones. Everything needs GPS

Every robot needs GPS and every robot is going to encounter in future conflict

a lack of GPS. So there's more than enough opportunities out there

I'm going to, you have a technology, but what are some of the things that, you know

you see as challenges potentially for scalability? I mean, if this is something that could go on so

many things, how are you thinking about, you know, the future if this really takes off

how you build it at scale? Well, I mean, you know, building, you know, scale is a funny

I mean, what does scale mean? Scale means like, you know, putting one joule of energy into something

and getting 20 joules in return or whatever. So, I mean, like, our device is pretty simple to manufacture

We can, like, assemble one box in, like, eight minutes. It's a little computer

It's all NDA-compliant, off-the-shelf computer, camera, and IMU. It's like three or four components

And then you screw those components into a little box. And that's it

Are you doing the assembly in Ukraine as you go? Right now, we're assembling them in Europe, and we're assembling them in New York

Okay, excellent. Vermeer, that is a Dutch painter. Talk about why you named your company after a Dutch painter

Well, you know, it maybe made a little more sense when we were kind of trying to sell into this artistic market

But, you know, Vermeer was a 17th century Dutch artist. and the thing about him was he was very, very good at capturing light

and painting light in a very photorealistic way, unlike some of his contemporaries

So we kind of think that artists, they maybe just kind of possess some sort of magic in their eyes

in their brain, in their muscles, but some people have theorized that maybe he didn't possess this magic

Maybe he was more of kind of like a mechanical tinkerer. so people have kind of thought that maybe he used a version of something called a camera obscura

which is so essentially imagine i wanted to like paint you but i knew nothing about painting so

imagine i had like a canvas below me blank canvas i have no experience painting i would essentially

take like a little round mirror i would kind of like put it on a stand at like a 45 degree angle

and then i would kind of you know dab some red you know blushy paint and i would start to paint

the bottom of the canvas. I rock my body back and forth between the edge of the mirror and the

canvas so that when the edge of the edge of the mirror disappeared into the canvas, I, it meant

I made an exact replica of your face. So people have been able to duplicate this method that they

think maybe Vermeer used. So maybe Vermeer wasn't like a artistic genius. Maybe he was like an

optical engineer genius. So what are we doing? We're kind of like rendering optical light into

a unique thing and doing cool stuff with it. All right. Well, thank you so much for joining us today

I hope that the rest of AUSA is fruitful for you. And I know that you're demonstrating some stuff in a parking lot nearby

So that's correct. Yeah. Thanks for having me