Public Lecture Series — The New Eyes in the Sky

Video Transcript
Download Video
Right-click and save to download

Detailed Description

Putting Drones to Work for Scientific Research
By: Jeff Sloan, Geographer, Project Leader — USGS National Unmanned Systems Project Office

  • Why is there so much interest in unmanned technology?
  • What are the rules to legally  y within the National Airspace?
  • How does this technology increase safety, lower costs, and lead to the collection of better scienti c data?
  • Will this technology become a commonly used tool for Scientists?


Date Taken:

Length: 01:18:58

Location Taken: Menlo Park, CA, US


[ Silence ]

[inaudible background conversations]

You should come
up here and say that.

- [inaudible]

Good evening. Welcome to the
U.S. Geological Survey

in another installment of our
monthly public lecture series.

My name is Leslie Gordon,
and it is always my pleasure

to introduce
the speakers.

I enjoy them as much as
all of you do, believe me.

I always have an announcement
because I always want you to

come back next month.
Next month, we’re having a talk called

Remembering Mount Pinatubo 25 Years 
Ago – a look back at one of the largest –

in fact, it was the second-largest
volcanic eruption of the 20th century.

Does anybody know what
the largest volcanic eruption

on Earth in the
20th century was?

[inaudible responses]

Krakatoa was 19th century.

No. Mount St. Helens
was pretty puny.

The largest eruption in the 20th century
was the Katmai Novarupta eruption

that eventually became
Katmai National Parks

and the Valley of Ten
Thousand Smokes in Alaska.

That was in 1912, I believe.

The second-largest was Mount
Pinatubo in the Philippines in 1991.

It was really gargantuan.

And John Ewert, who was there 25 years
ago, but is still working in our Cascades

Volcano Observatory, is going to be here
in Menlo Park visiting for this.

We’re going to show an
oldie-but-goodie episode

of Nova called In the
Path of a Killer Volcano.

It’s an excellent film even though it was
done in the – like, what, I guess 1990s.

The other reason for doing this is it’s not
only a quarter century since it happened,

it also happens to be the 30th
anniversary of a USGS program

we’re really proud of called the
Volcano Disaster Assistance Program.

We go all over the world with the
support of the U.S. State Department

and assist as needed or as requested
and cooperatively with foreign

governments and their local geologists –
respond to volcano disasters.

So the thing about
Mount Pinatubo is

that it was not as big a
disaster as it could have been.

Some of you may remember we –
the U.S. actually closed Clark Air Force

Base and Subic Bay Naval Base because
of the volcanic eruption and all the ash.

It was pretty amazing.
Okay. Next month.

Tonight it is my pleasure
to introduce Jeff Sloan.

Jeff is the project leader of
the U.S. Geological Survey

national unmanned aerial systems
project office in Colorado,

and we’re very glad that he took
the time to actually travel out here.

Jeff holds his undergraduate
degree in geography from the

University of Northern Iowa
and graduate work in GIS,

or geographic information systems, at
the University of Denver in Colorado.

He’s worked as a cartographer in the
U.S. federal government for 31 years.

And you’ll see – I love that
he starts his talk with a little bit

of history to
see where we are.

He began with the U.S. Department
of Defense working with the

defense mapping agency, then the
Department of Homeland Security,

then going to the Customs and Border
Protection office in Washington, D.C.

But the majority of his career has been
with the USGS in Denver, Colorado.

His background is primarily
in the area of cartography,

and remote sensing.

Specifically pertaining to domestic and
international mapping and monitoring.

He’s been involved with
UAS operations since 2008.

And in 2014, he actually
took over the project

as its leader in
Lakewood, Colorado.

So Jeff Sloan will be speaking
about New Eyes in the Sky –

Putting Drones to Work for
Scientific Research. [laughter]

Did I do – I did that.
Sorry. [laughter]

So please join me in
welcoming Jeff Sloan.

[ Applause and Laughter ]

I didn’t do that on purpose.

- Well, we have a qualified
pilot now, so I know where to go.

That’s about all there is to it.
Let me know if you can’t hear me,

or if we’re good here.
Just wave if you can’t.

But thanks, Leslie. It’s an honor
to be asked to come out here.

Because what they –
our office in Denver –

we have offices across the nation,
and out here, they put together that

monthly lecture series and bring in
some of the world-renowned people.

The scientists that I have great
respect for within the USGS,

and it makes me proud,
but I’m not sure I’m worthy to be

in that line, but we at least have a
very interesting topic, if nothing else.

And as – whoops, I’m sorry –
as Leslie was flying this earlier,

all I have to do is –
there it goes.


So whether you like –
I’ll interchange the words, first of all.

It’s – we call – the names have
been debated quite at length,

and it has settled on a – like, an
unmanned aircraft system, or a UAS.

So I’ll use that term.

But also it’s referred to,
mainly in the press, as “drones.”

And, to me, it doesn’t matter.
I’ve heard the head of the FAA

calling it “drones,” so I think we’re
good to call them “drones” now.

And at least everybody knows
what we’re talking about.

But what it does do
is stirs emotions.

We get – either you
love them or you hate them.

I’m not sure there’s
an in-between.

What I’m trying to show tonight,
though, is how we can use this.

It’s just another tool
for our scientists.

And it may be a perspective
you never heard that we can –

that we’re able
to use these.

You may just see
them in the parks.

I thought it was ironic, however,
though – I was at a Starbucks over the –

waiting to come here,
and lo and behold,

right on the headline, Menlo Park nears
a drone ban in the parks. [laughter]

I go, now are you setting me up
here or what? [laughter]

But there is some
timing behind it.

There’s a big announcement –
or actually the rules –

I’ll talk about the rules to fly these
that are going into effect on Monday.

So good timing,
but let me get started here.

We’ve already flown our drone,
so we can keep going.

So what I want to talk about is why is
there so much interest in this technology.

I’ve been in the government in mapping
for 32 years – starting my 32nd year.

Nothing has come close to the interest
that there is in this technology.

And why is that? We’ll talk
a little bit about – I do want to

give you a history of – mainly mapping.
I come from a mapping background.

And I just want to set that framework
so you can see for yourselves

where these may just fit in as a
perfect tool for what we’re looking for.

And then, how do you
make this all work?

Flying the drone is one thing, with
maybe a camera that I’ll show here.

But how do you
make that all work

and put it into scientific data
that we can actually use?

And then, really, is it safer?
Is it more cost-effective?

And is that – is
it actually better data?

And the only way
I can show that is

what we’ve done
over the past five years.

So my ultimate goal is just to show
this is just another tool in the toolbox.

It’s nothing that’s going to revolutionize
all of mapping and all of scientific work.

So why is there so much interest in this
technology? Why are you all here?

You probably have your own decisions,
but here’s some ideas.

Is it a new way of life?

Is this just, like we said,
going to revolutionize everything?

Is Amazon going to be
delivering whatever,

maybe getting
confused at times?

Is Santa Claus going to be downsized,
and it’s all going to be drone delivery?

Was George Jetson a prophet that
could actually see this was coming,

and now we’re here,
where you actually have to get in

aerial mergers to be able to fly?
We don’t know.

But that’s a possibility.
What about privacy?

You’ve heard a lot about the privacy,
mainly people shooting them down

if they’re flying in their background,
which I’m all for, by the way,

but you didn’t hear me say that,
except this is recorded. [laughter]

But, you know, if they’re flying in
their backyard, maybe all is fair.

And do we have to
carry around drone swatters?

You know, is there just
going to be that many?

So maybe it is blown
a little out of proportion,

but still something
we definitely take serious.

Is it the cool factor? And to me,
this is probably the bigger part of it.

These come in a variety
of shapes and sizes – all the way –

I brought some with me. This is the
only ones I could pack on the airplane.

But these do compress down into
little packages, down to a little size.

We’ve seen them that big
all the way up to the size of a 737

where they’re being
flown without a pilot.

But anyway, definitely a cool factor.
I’ll show you some of that too.

And is it even more
data than we already have?

Just another way of
looking at all this.

But what I want to try to show is
it actually is a really useful tool,

especially for
our scientific work.

So just to give you a little
scope of what’s out here already,

a little over a year ago, the FAA
set up a registration –

a voluntary registration for
those that already had their drones

or that were getting them
for Christmas, in particular.

And they asked that – if they would
go online and voluntarily register these.

They get a N number just like
an airplane has an N number.

And you were supposed to
put that onto the tail somewhere.

But in the first 30 days,
300,000 of them were

voluntarily registered
into this database.

I believe – I’ve checked about
two to three days – weeks ago,

and it was up
over half a million.

So with what I’ll talk about what’s
coming Monday, right here, actually –

the new rules for these
small UAS are coming out.

And it was announced back in June,
but it’s going to take effect

on August 29th – Monday.

So what’s that all mean?

This is the new rule,
mainly for commercial entities –

those entrepreneurial –
especially here in Silicon Valley,

where a lot of these little
companies are popping up.

I saw a statistic that
Menlo Park probably had

the most registered commercial
drones in the nation.

So it is something you
probably have to be aware of.

But it paves the way for various rules
where they can fly a little more free

without having to go through a
waiver process that I’ll talk about.

But anyway,
that’s all coming Monday,

so if you don’t like drones, you probably
really won’t like them on Monday.

It’ll take a little while
to phase this in, but …

So let’s talk a little bit about the history
and where we see these fitting in.

And the best way to
describe it is to go through it.

What we’ve done in the national
mapping part of the USGS – back in –

I can say “the good old days” because
I’m getting to that age, and I’ve been

wanting to do that for a long time – but
our survey crews were out in the field.

Actually, full-time they would
be out there, and they used

the technology of the times using
alidades and transits and levels.

And they were actually able,
with a plane table, to create

a topographic map that you see
on the right there, right in the field.

You could actually draft –
with knowing distances

and establishing the benchmarks,
which a lot of you are familiar with.

However, the young crowd
doesn’t really know what these are.

They’d be cemented in, and some
of them – a lot of them still exist.

A lot of them have been disturbed.
That’s where this one came from.

So the field crews would
go out and identify those.

But that’s the framework
for which you start.

You always have to have
a benchmark or a starting point

from which you could
turn your angles.

And that’s how you
made the maps back then.

Along came aerial photography –
the analog world,

where an airplane
would fly along.

And as long as it was taking pictures
with a little – with a little overlap,

we could then start to use either a
hand instrument like a stereoscope –

just like you’ve seen in
many different forms, or these –

like up in the upper right-hand corner,
the analog stereo plotters.

So however the airplane was
collecting it, you just mimic

that orientation, and it would project it
down onto a 3D collection device.

And a lot of us did that.
I did it when I started.

And then it would go on to
a map finisher, and we’d use

little tools like scribers.
This is where I started.

I did this for three years, where you’re
actually carving off plates and making

a photographic negative, which went
into the color separation process.

And that’s how
the maps were made.

Along came the satellites at that
time too, and we have a rich history

with our Landsat satellite – several of
them – that started over 50 years ago.

So still just different
ways to see this.

And the technology just fed into
better ways of doing it –

not necessarily more accurate –
well, a lot of times more accurate.

And as the – as the techniques
improved, you continued to use that.

Along came
the digital era.

And including the global
positioning satellite, which is

probably the most revolutionary
thing in the mapping world.

But all the sudden, now we have a
way that we can use devices like

a GPS receiver or a survey grade
where we set up and get very accurate –

down to centimeter and
millimeter levels of accuracy

for those – for those
starting points.

Even the iPhones have the
capability now down to

a meter or so that – and you know
what all that has brought about

for capabilities and just apps and
everything else that goes with it.

But the digital world brought in –
not necessarily better ways,

but maybe a better way
to manage the data.

And in the middle
there are different layers.

And the digital layers became our map
overlays. It’s all geographic data.

And the aerial photography – the
cameras were big, and they were fancy.

They were multispectral and all sorts of
different ways of collecting new data.

And then the
analog compilers.

The first time I saw a aerial photo
on a computer screen, I thought,

wow, that was one of the
coolest things I’ve ever seen.

That was mid-’90s.
It’s not that long ago.

And then the lasers are coming along.
That’s probably the biggest push

right now in the USGS is
to collect what’s called Lidar.

It’s a laser signal – an active signal
sent out with a very accurate return.

So that’s being
compiled for the nation.

But all that – all that feeds
into geographic data.

There’s no difference. It’s – whether it’s
digital or analog, but it all goes into,

for us in particular,
the 7-1/2-minute USGS quads,

which there’s 56,000 of them
that cover our nation.

And that was a major
accomplishment when we did that.

But whether that’s digital or
in paper, it doesn’t matter.

It’s still geographic data.

And I think you can start to figure out
where the drones feed into this –

mainly in the aerial
photography area.

The Department of Interior
has a humongous job.

Everything shown in blue there
is what they are – they’re mandated

to federally manage.
And that’s one out of every five acres.

And that includes not only the –
well, the USGS doesn’t manage land.

We’re the scientific branch of that.
But Bureau of Land Management,

Bureau of Indian Affairs, Office of
Surface Mining, Bureau of Reclamation,

Parks Service,
and U.S. Fish & Wildlife Service.

So that’s a large responsibility that,
really, the only way to even

do that feasibly is through ways
to sense the data from a remote object,

usually such as a satellite, and then
you just work your way down.

We have aerial photography
that fits in at various heights.

Well, along comes the drones,
and a lot of these can fly –

some of those bigger ones can
go clear up to 65,000 feet,

and they can fly up there, and you
work your way down to 20,000 feet.

Where we really see our role,
and what I brought along here,

are the really
small ones.

And that’s a gap in the data
that we’ve seen –

right around
1,000 feet and below.

So that really filled a gap,
and we noticed that right away.

Give a little history of
how long we’ve been in it.

We actually started
looking at this in early 2000.

A lot of work was done out of NASA
Ames with the Forest Service,

and they did – they modified Predator –
one of the bigger planes to fly wildfires.

And that’s a – was one of
the ways we got our start.

But really, in 2004,
the conferences started.

And I remember some of
the earlier ones, I’d go,

and they –
it was all military.

All the technology was
coming out of the military.

And we’re leveraging
that now because –

we’re as far along as we are
because of what was done there.

But they literally say,
what’s the USGS doing here?

You know, they had
no concept that this could be

used for mapping
or scientific data.

It was usually video, and it was
usually at an oblique angle,

and the whole idea being
looking for the bad guy

and feeding that data feed back
to the people that needed it.

So we saw this.

It wasn’t –
it wasn’t going to go away.

And we saw an incredible
resource there that we could utilize.

So in 2008, we actually set up the office,
which was not a small ordeal.

But we set that up in Denver.

About a year later,
the Redstone Arsenal –

the Army down in Alabama –
surplused the Ravens –

the fixed-wing – what I’ll
call a fixed-wing. It’s a little airplane.

And they gave us about 17 systems
in various conditions.

They had been
through a couple wars –

Kevlar-coated, hand-launched,

But they told us they were
going to destroy them.

They were going to dig a trench,
they were throwing them in,

and they were going to
drive a tank over them.

And I don’t know if that’s true or not,
but I like the story, so I keep telling it.

[laughter] But we came along
and rescued them,

and we put them to work
for about five years.

We made every modification
we could to make those usable.

All up until even last year.

And then, shortly after that, we got
some more, and I’ll tell you

about those in a minute. But we were 
trained on those a year later.

And then we’ve
been actively flying.

We got into the national airspace
in 2011, and it hasn’t stopped.

It’s been – I turn down work on a
regular basis because we’re overloaded.

So how do you
make all this work?

You know, it’s – collecting the data
is one thing, but how do you

make it work into a geographic-type
data layer that we can use?

First of all, though,
you have to talk about the rules.

How do you get into
the national airspace?

And actually, before I do that,
I want to boil it down into

what I’d put into
three categories.

This is really what you
need to make it all work.

You need the drone –
the platform.

You need a sensor – a camera of some
sort that can – that this drone can lift.

And then you have to
have the software.

And each of these categories are
developing at lightning speed.

It’s hard to
keep up with it.

All the emphasis is on the drone,
but these little cameras –

even point-and-shoot-type cameras
you get down at Walmart or Best Buy

are of such good quality, it matches –
what I’m able to do with this right now,

I could – it would take a mapping –

mapping-type camera just
three to five years ago.

So that’s what’s
making this work.

And then the data processing –
the algorithms that we’re –

that are being developed change.
Literally, monthly we’ll have to

change our software because
it gets better and better.

So let me talk about the
first part of that – the platforms.

I’ve already introduced you to the Raven,
which I brought along with me.

We had 17 of those systems in
various conditions. Got those working.

When I say a “system,”
that’s three aircraft.

So we had three airplanes –
three times 17.

We were given quite a
little gift there for free.

And then, shortly thereafter
came the T-Hawk, which was

a Honeywell product.

And that was given to us also
by the Redstone Arsenal.

Those literally had not been used.
They came to us brand-new.

We got 22 of those systems –
a system being two aircraft.

So we had 44 of those.

And then even the Super Bat,
which is a larger fixed-wing,

that was surplused to us as well from
the National Institute of Standards.

And we kind of joked we became
the dumping ground for what

these people didn’t want, but the
price was definitely right for us.

And it got us into the ballgame.
Our whole – our whole mission

at the time was to figure out how these
work with the FAA – how that works

with our own aviation services
within the Department of Interior.

And it was painful
because they were –

they didn’t know how to handle
these any more than we did.

So we learned together,
and eight years later,

I think we’re getting to
where we want to be, actually.

So last year – early last year –
the way the Department of Interior

wanted to handle this
is fleet aircraft.

These are classified as aircraft,
so we have to treat them that way.

We have to be class II
medicals at the time.

This is kind of going away just
recently, and Monday even more.

But we had to have class II medicals.
We had to be – we had to stay current.

We had to
fly every 30 days.

We had to be proficient.
We had to have check rides every year.

So treat it just like
a manned aircraft.

But we – they decided, and we decided,
within the Department of Interior,

this is available to everyone,
not just the USGS –

categories based on weight –
and that’s what some of the rules

are all based upon
is weight.

So actually, while I’m thinking of it,
the rule that’s coming out on Monday

is for 55 pounds
and below.

So that’s not necessarily
a little small aircraft.

That can be that big one that
you see down on the bottom.

But the way we
handled it was by weight.

So four different categories.
The micro level – just the little 3DR

that’s about this big, which you see –
you can get them down at Best Buy.

That’s becoming a
very useful tool for us.

It’s one that we really see our
scientists using and being able to

throw in the back of the pickup
and go, or in his backpack.

We’re working on five cameras
that will go on that – interchange.

Then we had a fixed-wing category –
a little bit bigger than our Raven.

The one we settled on
was called a Falcon unmanned.

Actually, they won
both categories.

This went out for a solicitation
and a bid process.

And both the Falcon and
the Falcon Hover we could

interchange payloads –
the cameras.

So we could put them
either on a fixed-wing or

on a hover, of which
we see a need for both.

And then we could fly that off the
same type of navigational system.

Most of these will fly either
off of a laptop, a tablet,

or you hand it off to a
PlayStation-type joystick-type flying.

And you can interchange.
You can hand off.

And then we did see a need for the
heavy lifter – the heavy category.

These are able to carry –
it’s a pretty healthy system.

It’ll lift up to
10 pounds pretty easily.

So we had so many geologists
and hydrologists come in with

all these ideas of sensors
we could put on these.

And we had to have a way to lift that.
So that is why we went after that.

That, by the way, is about
the size of a snowmobile.

A little – it’s a pretty
healthy system.

The second part, of course, is the
sensors. [musical tones in background]

And like I said earlier – oh, that’s a
pretty song. [music stops]

We started with a
little GoPro camera,

which you’re probably very familiar
with, about, you know, $300, $400.

But that’s a 3-1/2-ounce camera.
So we could easily get it into the Raven.

We just took out the camera that’s
in there and stuffed this in there.

And all the sudden, we were –
if you’re not familiar with these,

these still are probably some of
the best video capture devices.

They’re high-resolution –
4K, if that means anything to you.

But you can get very good resolution
from that, and it’s lightweight.

We then – but mainly we
wanted more still-frame imagery

for the type of
work we do.

So these little point-and-shoots
work real well, all the way up to

a digital SLR camera
like many of you have.

All these cameras are
under $1,000, probably,

that are extremely useful for the
quality that we’re getting from them.

And so we talked about video –
Sony Action Cam or a GoPro.

The FLIR camera – this is
a little inch-by-inch sensor.

Like I said, this technology –
the drone technology is pushing this

technology to shrink everything and
to make it available and affordable.

So these little thermal cameras –
even though they’re still about

$4,000 to $5,000,
they give us a temperature.

So we can fly at night and start to
pick up wildlife and whatever else

we can use thermal for –
maybe water temperature differences

and things like that.
But about the same size as that GoPro.

And then probably some of the later
stuff – this multispectral-type sensor.

And multispectral may not
mean a whole lot to some of you,

but really all it is is starting to
push into the near infrared.

That’s what the eye can’t see.

And we’ve used that for years and years
for vegetation and geologic work.

But now, not more than a
year or two years ago,

the only capability for multispectral
was a $100,000 camera.

This is now a point-and-shoot-type
size that can go onto a drone.

But there’s five separate cameras in
there, all oriented at different bands.

And even some of –
two of those into the

infrared region where
the eye can’t see that.

And I’ll talk a little bit –
and it’s easier to show you

what that means
than to tell you.

And then we’re already – of course
our geologists are thinking all the time.

Sometimes thinking
way too much. [laughter]

And we’re looking into some of
the sensors that we can –

that are available right now
that we can put on these drones –

geomagnetometers, hyperspectral
sensors – if five bands isn’t enough,

they’re up to 28 that we’re working
with the University of Florida,

and we’ve already
proven that to work.

These medium-size format cameras –
the one on the far right – that’s about

the size of a normal digital SLR
camera, but it’s 100 megapixels.

So every time they say “megapixel,”
that’s a million pixels, which to us

means we can go higher and collect a
bigger area at that same resolution.

But it also means you need a whole lot
more storage on your computer.

Because the amount of data coming
off there is overwhelming.

Radar being developed – we have a
Austrian company that’s building that.

And then a telemetry one which
we’re going to try next month.

So we’re actually flying a
drone with NASA Ames.

They’re coming out with us to
Missouri, flying the Mississippi River.

They have tagged
some Asian carp,

which is a big problem
in the Mississippi.

And we’re going to
fly a drone over that

as a test to see if that
will pick up the radio signal.

A lot of ideas coming out.
Part three – this is putting it all together.

Now remember,
when we fly these airplanes,

we have different ways
of triggering the cameras.

They’ll either be on a timer as it’s flying
along – snap, snap, snap our pictures.

Or we’re triggering it
off the GPS off the airplane.

Some of these cameras have GPS
in them, and that will tag that image.

But if you notice, there isn’t
a whole lot different.

If I can just trigger that at a
certain speed – or, we could fly it

at a certain speed and trigger it, all the
sudden, I’m getting my stereo overlap.

This is no different than
the way we’ve always done it.

The software now is maybe
handling it just a little bit different,

but we’ll collect the video.

What these do, when we start
to run it through our software,

it takes the overlap, and it will
use triangulation – geometry.

Comes down through the photos,
and where it hits the ground,

it creates a point – an X, Y, and Z
value – a latitude, longitude, elevation.

But it also paints on a color.
So if you look on your computer,

you’ll get red, green, and blue –
what the eye can see.

So that paints an
exact color to that dot.

And these points clouds,
as we call them, can be generated

either this way or from a laser –
it’s very common today.

And we’ll generate from – for an
example, we’ll fly for a half-hour,

we’ll get about 500 photos that
it’s automatically taking for us.

And we’ll run it through
the software, and we’re generating

this massive point cloud of
several hundred million points.

And the more points we can get,
the more accurate the elevation model

that we can
derive from that.

And then, once you have that
elevation model, you can drape

the imagery over it, stretches it
to the correct position, and now

you have a very usable mapping
product – an orthophoto, as we call it.

It’s a corrected photo.

Now we can start
doing our mapping.

And our ultimate goal is to
just create those data layers

for geographic information
systems or for plotting a map.

These are examples of
some that we’ve already done.

And as I said, it goes into
the geographic data layer.

I’m going to show a
couple videos at the end.

I would show it right here on how these
are launched, but remind me at the end.

We’re having some
formatting problems here.

So what are the rules?

What we had to do in the
past working with the FAA,

in order for us to get a 5-pound airplane
in the air for over a 1-square-mile area

in the middle of nowhere, New Mexico,
we had to file paperwork about

a inch thick, and it took six months.

And that was the process,
and it was painful. Very painful.

They gave us a little
gift about three years ago.

We set up a
memorandum of agreement.

We could fly up to – well, they
actually gave us – well, first of all,

these are some of the rules that
actually are coming out on Monday

I’ll talk about in
just a minute here.

But we have to keep these
in line of sight all the time.

So you really
can’t cover a big area.

Usually about a one – we can see them
out to about a mile, and then we can

have somebody on a radio that
can maybe see them another mile.

We can’t daisy-chain
those observers.

So as long as they have
direct communication

back to the people flying it,
you’re good to go.

In general, if you can stay 5 miles from an
airport, you’re pretty good to go.

They let us – for, like,
grass strips and things like that,

they let us get
a little bit closer.

Not over people.
Not over urban settings.

We have to file
our Notice to Airmen.

But with this memorandum,
we literally were able to file

within about a 48-hour
period and then go.

We don’t have to wait six months and
have a inch tablet of paper with us.

A couple special things we got.

We get to go up to 1,200 feet.

And then we are
allowed to fly at night.

We made the case that we have a lot
of biological studies, mainly for birds,

that it’s safer for us to fly at night,
and easier to see, actually.

It’s a lot easier to see these at night.
They have blinking lights on them,

and we can actually see them
out to about a mile and a half.

So this is the new rule
that’s coming out on Monday.

You’ll hear it referred to as the FAA Part
107 – Small Unmanned Aircraft Rule.

So you can read down through there,
but the key thing is up to 55 pounds.

That’s a pretty healthy system that
is now allowed to fly locally here.

They still have to keep
them light-of-sight.

And these are mainly
for commercial entities

that want to create
a business with this.

The hobbyists, they still go under
the rules they were always allowed to.

So one of your questions might be, I’m
seeing them fly in my park right now.

How come they’re
able to do it?

And they’re probably just
doing it as a hobby.

If it turns into a research project, like
universities or for the USGS scientists,

or if it’s a commercial entity
for a real estate company,

if they’re generating money,
then it falls under different rules.

And these are now the
new rules that are coming out.

Maximum speed – 100 miles an hour.
I’m not sure you ever want to

fly these 100 mile an hour,
but I’m sure they will eventually.

And really about 400 feet is their
top limit, and they really aren’t –

unless they get a special waiver, they’re
not going to let them go above that.

But down on the
bottom is what –

they’re going to have
a basic level of certification.

I call that a driver’s license, but they’ll
get on, and they have to have some

aeronautical knowledge, kind of like
ground school with a private pilot.

They’ll have to –
this may be online,

or they’ll have to go to an actual
FAA testing center to take this.

Has to be vetted by the TSA,
and they have to be 16 years old.

That’s the rules we’re hearing.

And then finally, let me –
let me let you decide

as I plead my case here
that this is just another tool.

Is it really safer? Is it really
more cost-effective, better data?

Let me show you
what we’ve done over

the last five years,
and I’ll let you decide.

Our first project was
using the Raven.

We were down in southwest Colorado
looking at the sandhill cranes.

And we first flew this
during the day because

we couldn’t fly at night,
and it scared them.

That’s a big issue,
actually – scaring wildlife.

And our U.S. Fish & Wildlife
people who we’re working with

are very sensitive to that.
So usually we’ll start out

really high and work our way down
until we do see them disturbed.

But it wasn’t working during the day,
so we got a waiver to work into

civil twilight – 30 minutes before
or 30 minutes after sunset.

And we found that they come back to the
water and roost at night on the water.

And when they’re on their roosts,
they’re very calm.

So we were able to fly clear
down to almost 75 feet above them,

and it didn’t disturb them –
didn’t disturb them at all.

We had our – some of the
Fish & Wildlife biologists

were out there counting
while we were flying.

And then we derived those
data sets afterwards,

and I believe we compared the results,
and we were about 3 to 5% off.

So they did see this as a viable tool that
they can possibly use down the road.

We also used a computer-automated
way to count them instead of

counting them by hand.
But it’s far from unmanned.

You see all the people that
were involved there. [chuckles]

It’s how many people – how many
government employees does it take

to fly a 5-pound airplane?

But it was a big deal. It was the first that
we’ve been in the national airspace.

And believe me, there was even
more people than this. So big deal.

And we’ve been down there
many times, and we are allowed to

fly at night,
and it’s very cold.

So this little camper here
was a life saver.

But that camera that I talked about,
the little FLIR, it’s getting better

and better and lighter weight, and we
can fly that, and you can start to see –

a whole lot easier to start counting them
with the resolution that we’re getting.

Using that same type of camera
to do trumpeter swans,

looking at their
nesting sites.

They want to map where
they’re located as well as

just the habitat around them
on where they’re nesting.

Real easy to do when you’ve got a nice,
warm nest, even during the daytime.

And working out this way,
up at Tomalas Bay,

Kern and Pixley
wildlife refuges.

We were working with
Fish & Wildlife on this also,

and our goal here was to try to see
how low we had to come using the

cameras that we had at the time to
start distinguishing species types.

And we cheated a little bit.
Those are actually decoys down there

in the right-hand corner, but it did –
they didn’t move, so it gave us

a real good idea on where it –
how far we’d have to come down.

As well as vegetation –
there’s usually multiple studies

going on
at the same time.

But we – you know, we can – this is a
mosaic of several images put together.

And then the automated counts
of the birds on top of that

for pelicans
in North Dakota.

Same thing in Reno, Nevada, but just a
different way that we can start to display

this imagery – maybe give it a different
perspective for the biologists.

Looking at elk. This was in the
Carrizo Plain just south of here a ways.

But they were – they had a
particular interest because

they normally would do these
population estimates

from a helicopter with
somebody sitting in there.

And I believe they did have some
fatalities on a crash a couple years prior.

So they were very sensitive
to what we were – or actually

very encouraged on using
this as a safer technique.

It’s very difficult
to find elk.

We went up to Washington and tried it,
and we were kind of shut out.

They’re very sneaky.
They can get into trees quite easily.

We happened to know where these were,
so it was a kind of a proof of concept.

The biologists were a little more excited
than we were about it because if they

do have a GPS radio collar on them,
and they tend to hang out in herds,

then we could fly to that location
and start to get our counts that way.

Just this past year, our guys were
flying up in Idaho, and this was

nearly by accident, but with our
better cameras and a higher altitude,

we were able to see
some elk that way.

Usually in the clearings.
Pretty darn hard to see in the trees.

This was using the
little GoPro camera.

One of our first attempts at
just using the little GoPro.

We would put it on still frames.
This was the dam removal up in

Olympic National Park and letting
that river restore back to the way it was.

But we got right out on the sand
and flew right down the river.

And just using the still frames,
the GoPros aren’t ideal

for mapping whatsoever.
But it was a first attempt that we

started to see this resolution,
and we knew that we had something.

This was a contracted file.
I’ll talk about contracting a little bit.

But some of our work
being done by our scientists

out at Cape Cod –
just elevation models.

The idea here was to get a
baseline of imagery before,

then after a hurricane comes in,
or a storm, he can run back out there

with the – actually with the little –
I’m pointing over there,

but the little solo
that I showed you.

He can fly that real quickly and
then start to see just how much

damage was done or how much
sand was lost, whatever.

The mining world is very,
very high on this technology.

I actually was asked two years ago –
two to three years ago

to go to Australia to a
mining conference over there.

And each mine over
there has its own surveyor.

And one of the speakers there, I won’t
forget, he was a 25-year career guy.

And he showed very clearly –
the first statement he said is,

you either learn this technology, or
you’re out of a job, if you’re a surveyor.

And he proved it.

He would walk around with
the surveying instrument,

set up the base station, and walk
all day long with the rover,

and he would show just the
little contour collection

that he was able to do,
you know, for most of the day.

He had his little airplane.

He said, I threw that up in the
air for about an hour and a half.

I had the entire mine mapped,
almost daily if you really wanted to,

and the resolution as well as
the images that came with it,

were better than
what he was collecting.

So he did have a pretty strong point,
but when it comes from

a career guy like that,
you tend to listen.

But this was working with the
Office of Surface Mining – yet another –

and the BLM – another sister
agency within the department.

But this was using our T-Hawk,
where it can stare.

You know, it can just hover there,
and we can point the camera.

And you start to look at,
you know, the head walls,

and we can map
in oblique angles now.

We don’t necessarily
have to be directly overhead.

And this can be measured.

This new software is able to correct
this imagery where we can

measure vertically now instead of
just strictly off of a flat map.

And we’ve done a lot of mining work.
Contracted the job on the left.

That’s in Missouri.
I’ll talk a little more on that.

Some of the abandoned mine lands
up in Colorado and the

reclaiming that’s going on
in West Virginia.

About 2,000 open-pit mines
out there that the

Office of Surface Mining has a
humongous job to try to monitor.

But looking at that Missouri project,
we contracted this job with a –

just a local guy that
had a waiver to fly.

It’s what’s called a FAA Section 333,
and that was the – what we did the

whole past year before these new rules
that are coming out on Monday

that will wipe out all those –
all the need for that waiver.

But anyway, we contracted with
this guy, and we were able to

put two cameras on here.
We were looking at sinkholes.

So just to give you some perspective
on the accuracies we’re getting

with this data, he had the drone,
which is a $2,000 drone –

it’s a DJI Inspire, if you
know the brands out there.

We did put down ground control.
That’s key if you really want to

get down to the centimeter
accuracy with this data.

But he flew that, and I also
gave him one of our

point-and-shoot cameras
to put on there.

So he flew both cameras
at the same time.

So the bottom line is, our accuracy
was 2 centimeters in the horizontal.

I did some testing against
checkpoints that we put out.

And then 4 centimeters
in the vertical.

So let me put
that in perspective.

It’s a $2,000 drone
with a $500 camera,

and we’re able to get down
to the centimeter level.

So that’s what’s generating
an awful lot of interest.

When you get down to
that type of accuracy,

you can start to do volumetric
measurements – in this case, stockpiles.

We’re also using it for disposals
on Department of Energy-type sites

on what’s being
removed or added.

This is pretty interesting.

Working with the
Park Service and BLM.

They have for years and years
modeled dinosaur tracks

down to the
millimeter level.

They’ll actually do what’s called
close-range photogrammetry,

and they set up tripods in the –
in the past 30 years and –

just to get a camera high enough
off the ground over one track.

And then they take photos where
they can get that stereo overlap.

And then they’ll create a 3D model of
the track, so down to the millimeter.

Along came the drones, and that
takes the place of the monopod.

We don’t have to do that.

So in this case down at White Sands,
the sand would blow away

and reveal these woolly
mammoth track ways.

We flew the drone out.
It’s getting a GPS location.

And they threw out some scale bars.
We can also scale this,

so we can get down to the millimeter,
and the BLM guys were able to

model that – a millimeter-type
level on your dinosaur track.

And then, in 30 days or so,
they’re covered back up

with the sand, so you never
know they were ever there.

So the Park Service was pretty excited
about this because now they not only

had a location, but they have an actual
model of the track that’s there.

Pretty neat idea.

Once again, showing some
of the accuracies.

This is in the Platte River near Kearney.
Our hydrologist surveyed across –

just three transacts across the river.
You know how shallow it is out there.

And we flew our drone with just
a little hand-held $300 camera.

And we processed
the two data sets,

compared them, and our
worst error was 6 centimeters.

So getting down to that type of
accuracy is pretty impressive.

Landslides – I’ll show you
a little bit more about this.

But we can start to generate those point
clouds that I was telling you about,

and that data is pretty useful for seeing –
especially over time, when we start to

compare that, we’ll see
centimeter movement

if it’s done correctly and – with just
some free software that we use.

And we – starting to see where it’s
green there is possibly some movement.

They’re interested in looking at it as
it comes towards the interstate there.

And then fault lines, which is
of interest to you out here.

This is flying up in Idaho –
a pretty easy-to-see fault line.

But that’s a point cloud.
That’s not an image.

It’s actually points.
Just hundreds of millions of points.

We flew that at 1,200 feet.

We actually had some laser data
that was collected 10 years earlier,

so we are always asked how – just how
good this is because the Lidar, the laser

imaging and detecting, I believe – that’s
kind of considered the best out there.

So we like to
compare the two.

And we really only had –
where it’s colored there

up in the right-hand corner, a 6- to
12-inch difference in the vertical.

So we consider that quite good.

Because we’re just using
off-the-shelf cameras for that work.

By the way, a laser collection
is extremely expensive to do.

Then down here in the Mojave Desert
in the southern part of California.

I just wanted to point out the
contours that we’re able to generate.

We can – we can generate 6-inch
to 12-inch contours pretty easily.

And I don’t know if you’ve ever looked
for that type of data, but it doesn’t exist.

So we’re starting to
really push the limits.

Let me show you some of
the uses of the multispectral.

Like I said, this is pushing into
the near infrared, but what I mainly

want to show you here is this is
actually one of these cameras.

And all we did is open it up, pull out
a filter, and put a blue filter in there.

So we get a pseudo
color infrared image.

But it’s probably not what we
would use for scientific studies,

but if you compare here, a lot of times,
we will task a satellite –

a high-resolution commercial satellite
that we have access to while

we’re flying just so we can see
what we’re actually getting.

If you look at,
not necessarily the colors there,

but the tone differences,
they’re somewhat similar.

And if I would run an index –
this is just zooming in –

you’re going to see some
pretty similar results from that.

They were taken roughly
a week or two apart.

So the difference being,
on this satellite is a million –

at least a
million-dollar sensor.

On our little airplane is a $300
modified camera. [laughter]

So it may not be the gospel truth
on what you’re getting,

but maybe it’s 80,
90% of your solution.

And I like to show this one
because it’s very common

to use color
infrared imagery.

And we’ll run what’s
called a vegetation index.

This is just a
common one that’s used.

And what it does is
it starts to show moisture

and heat and healthy vegetation –
chlorophyll in some cases.

So what I’ve colored with the rainbow
there – everything in red is good.

That’s a good
healthy growing area.

And the blue, like a sand trap,
is not reflecting much at all.

So what’s probably of more interest
is right in the middle – the green.

And what we’ve found over time
is usually in a color infrared image

will show where
you’re having a problem

about two weeks before
the eye can see it.

So this is what’s really
spurring precision agriculture.

If we can start to get site-specific
on a crop, and the farming world –

I come from a farming background –
it’s very sophisticated now.

It’s all tied to GPS.

Very few people actually drive their
tractors, especially when planting.

That’s why all the rows
are perfectly straight now.

You can’t tell a good farmer from
a bad farmer anymore. [laughter]

But the applicators that they’re pulling
behind are also GPS-controlled.

Each one of those nozzles that are
putting on chemicals or water

or fertilizers, whatever,
those are GPS-controlled.

If you use this type of data,
which has really been our missing link,

and you grid that, you just – literally
what I’ve seen is take a thumb drive,

plug that into the tractor computer
system there, and it will control –

as it’s going along, based on GPS,
it’ll vary the rate that that chemical

is being put on, or the water,
or the fertilizer.

And it varies as it’s
going across the field.

So what that – what that does,
these chemicals are expensive.

And it saves the farmer
right away the amount of –

in my – when I was a kid,
it was blanketly applied.

So the environmentalists are happy.
The farmer’s happy because he’s

saving money, and it’s increasing
his yield, so he’s doubly happy.

So everybody’s happy.
So it’s a – it’s been our missing link.

And the drones are very –
what’s the word –

really going to town
in the farming world.

Using our five-channel –
the little multispectral camera –

this, by the way,
is about $5,000.

It’s a – it’s a fancy little camera,
but really lightweight.

We were up in Teddy Roosevelt
National Monument.

This was the first chance
we really got to use that.

But you can see these are just
a little Sony A5100,

and we’ll land it and put in
this multispectral camera.

We fly the same areas,
and this is what we’re getting.

But very narrow bandwidths,
and we can do all kinds of

stuff with this
type of data.

But you can start to see the
resolutions that we’re getting.

And even with the little – the little
drone you can buy down at Best Buy,

this is the type of images you’re
going to be getting off of that.

Getting a little fancier –
this is our heavy-lifting aircraft,

but we can start to put on that laser
scanner that I was telling you about.

A pretty expensive sensor,
but nonetheless, we can fly that

at low altitude and get these –
the laser collections off of that.

But also what we can do is put on
these other lightweight cameras,

so we can collect simultaneous
data at the same time,

which is very valuable
to the scientists.

Albeit very small areas.

We have to keep these line-of-sight,
and the battery power –

these are all battery-powered,
and that’s still a big limit.

So very small areas.

Working with the Florida –
with the hyperspectral-type cameras.

And I won’t get
too much into that.

And then this one – I’m almost
done here, but it’s very interesting.

It’s what we call
a pseudo-satellite.

So these are – I believe Google has
looked into this also as a way to – like,

in Third World countries to be able
to provide internet services.

But they’re usually solar-powered
or some other type of power

that they can stay up for very
long periods, like months, even years.

And it’s called a pseudo-satellite because
it can hover at very high altitudes.

We’re talking 6,000 –
or, 60,000 feet and higher.

But it’ll stay up there for
long periods of time,

and we can put multiple
sensors on there.

If there’s a forest fire going on,
or a flood, we can loiter that

instead of waiting for a satellite
to rotate around the Earth.

So it’s becoming a
pseudo-satellite, so to speak.

The other – the other advantage to it,
which has happened many times –

the sensors on the
satellites will go bad.

And there’s really not a
cost-effective way to get those back.

This, you just land it,
fix the sensor,

and put it right back up.

So it makes a lot of sense, really.

We’re working with NOAA
and the NASA folks here

to actually put this to work
for us in a couple years here.

And then the amount of data,
like I said, it’s overwhelming at times.

But we do have
a way to archive this.

It becomes just another way
the USGS always wants to

get data into the public domain
so anybody out there can use it.

And it fits right in because
it’s all geospatially corrected,

and we have a way
to distribute that.

And for our scientists, can the
data be published?

And yes, it can.
It already has.

Not a problem.

And then just
the basic costs.

So this, once again, is why you see
so much of the entrepreneurial world,

especially out here,
going to town with this.

If you – you can really pay as much
as you want for these platforms.

It just depends how
sophisticated you want to be.

But really –
and the sensors as well,

and then your labor
and the processing software.

But if you really just go right
across the top, you’re functional.

So $800 for a drone, $300 for a camera,
however much you want to charge for

your labor, and then the processing
software is really not that bad anymore.

In the good old days, the software
was $50,000 to $75,000

to be able to process
these overlapping images.

It’s now $3,000.
Pretty affordable.

So what if you don’t want to do – what if
you don’t want to deal with the platform?

What if you don’t want to
deal with the cameras?

What if you don’t want to
collect your data?

Or what if you don’t want to
process your data? [laughter]

Any one of these piece
can be contracted out,

and that’s what
we’re starting to do.

So as I – as I said,
what’s coming out on Monday

is opening the gate
or all these companies.

But what’s – that really took –
started back about a year ago right now.

And some of the
first companies to

get these waivers were
Hollywood, of course.

Real estate agents
were getting the waivers.

The mapping didn’t really
come in until about December,

so I couldn’t reach out and
contact any of these guys.

Where it’s really cost-effective
is if they’re in the area.

So I’ll actually use this map
right here and start querying,

for example,
Cape Cod.

You know, to get a person
out there is fairly expensive.

Travel is one of our biggest costs.
So I can just tap a company that’s there

as long as they know what they’re
doing in the mapping world.

They fly the data for us, send that
back to us, and we process it.

Some of these will
go full-service.

They’ll – or even the companies –
if you did want to fly your own data,

but you don’t know – you don’t really
want to learn the processing of it,

you can upload that magically
to the cloud, and the next morning,

you get an email
with the products.

There’s a couple companies out here that
are pretty – doing quite well with that.

But anyway, any piece of that
can be contracted.

I have – that map is showing
about 3,000 companies.

The last count
I saw was 7,500.

I would imagine, on Monday,
that will increase quite a bit.

So I rest my case.

Will it become
just another tool in the toolbox?

I’ll let you decide, but I think –
I think I made my case, and I got to

get my little drone flying again.

And I do want to show you
some of the movies if you –

if you’re still wanting
to stay around.

A lot of times,
the people come out with us,

and all they really want to
see is it take off and land.

When we’re flying, it’s pretty boring.
Because it actually flies itself.

It’ll actually take off
and land on its own.

So it’s really pretty boring. But other
than the takeoff and landing,

we think boring is great.

What we don’t like is when it
gets exciting, so … [laughter]

So let me show you,
first of all, the platforms.

Whoop, what am I doing?

Oh, I’m sorry.
There we go.

So just how we
launch some of these.

And that’s me standing
in the back of a pickup.

And I don’t have
any volume with it.

But just a
hand launch.

And that’s the little GoPro collecting
images right out of the nose.

This was done in the Carrizo Plain
area of California.

And then, to land it, it actually is
a couple buttons on our controller,

but it will put itself – you’ll watch
the nose pitch up here.

And it puts itself into a stall.

And then it’s meant to crash land
and actually break apart.

It’s meant to do that, and you just
snap it back together. [laughter]

This one’s a
little more unique.

It’s, like I said, fueled with
100 low-lead for you aviation guys.

But if you can imagine,
right now, it sounds like

a lawnmower or
a chain saw flying.

It’s extremely loud.

Very hard to
use with wildlife, obviously.

But it’s still cool, you know?

And it was free. Not to the military.
[laughs] It was very expensive.

But we used these for
a good four years, five years.

We just retired this one
and the Raven, so we’ve

moved on to the –
to the new generation.

This is actually old technology
in the realm of drone world.

You know, this is probably
seven years old by now.

The Raven is about
10 to 12 years old.

This is our new fixed-wing.
This is one we’re using.

We are – this doesn’t always go,
you know, perfectly as expected.

We are having some problems, mainly
with GPS, on some of these systems.

So it’s not always working
just swell and dandy.


Now that’s a pretty
smart way to land it, actually,

because that protects not only
our camera but the aircraft too.

Where it doesn’t work so well
is if it’s windy, it’ll grab that

parachute and drag our camera
right across the terrain.

And that’s the hover.
That’s the one you see over there.

The beauty of those two –
like I said, we can just 3D print

with the new 3D printers,
or a vendor will do that.

He’ll print the mount around
the camera that we want,

and then we can
attach it that way.

But we can interchange it between
our fixed-wing and this hover.

So we like that idea.

In this case, I think the one I have
has the camera on the end.

This has the
camera underneath.

And these are usually gimbled,
so if your – the wind is throwing this

all around, that camera stays
pointing down real nice for us.

This is a little 3DR,
3DR being 3D Robotics.

I believe they’re a company
out here actually.

What you can buy
down at Best Buy.

But we see a lot of
utility in this little aircraft.

Like I said, we’re working on five
different cameras to interchange on this.

Real easy to learn to fly.

And it’s pretty reliable.

The downfall with that was it’s probably
only got a 15-minute time to fly.

This was working with the –
our University of Florida,

who we’ve worked
with for years.

They were kind of pioneers in a lot
of this, but this is called an octocopter.

It’s eight blades.
It’s not a bad idea.

If one of those goes out,
you can still get it back

because there’s plenty
of other propellers.

But that’s lifting a pretty heavy payload
there – probably 5 pounds, I would say.

This is an Inspire, a very
common Chinese-made aircraft.

That’s – just like you see,
the DJI Phantom.

It’s a very common – those
white ones that you’ll see a lot of,

this is the next jump up.
It’s about a $2,000 drone.

And like I said there,
we have two cameras on that one.

And this is a healthy aircraft.

We zoomed in to video this,
but we keep our distance.

Because if that blade comes off,
it’s not pretty at all.

And we don’t want to
find out what it’ll do.

It’s an expensive aircraft.

So we’re very careful flying this.

But it’s impressive.

It’s a nice little system,
and it’ll carry a lot of weight,

so we have a lot of – we have
a lot of thoughts on this one.

It’ll probably become
one of our work horses.

And these land themselves.
So in some cases, we’ll get it

close to the ground and then hit
a button, and it takes over and lands it.

In this case, you literally can
fly the whole thing by itself.

Just set up GPS points.

- Can you digitally transfer the
data to some central location?

- Yeah. Yeah, we’ve done that before.
- So it’s not just [inaudible].

- So on the low end, I’ve actually
live-streamed from the field back to

an office or to a school.
We’ve done that.

So that’s a cheap way to do it.
However, you can go clear up to

the really fancy systems
that beam that down.

We worked with the border
patrol on their Predators

to actually fly some
fires down in Arizona.

They were flying it – oddly enough, this
is how it worked, that they took off from

Grand – how was it – they took off
from Arizona – or Corpus Christi, Texas.

They flew it out of Arizona,
and they beamed the imagery

back to us in our office in Denver
as well as in Washington, D.C.

And we could communicate
with them to point their sensor.

But, yeah, it gets pretty
sophisticated sometimes.

So that’s the
takeoffs and landings,

and now you don’t need to
come out with us. [laughter]

And then let me show you yet another
data – whoops, keep hitting that –

just a fly-through of the modeling
that we’re able to do.

So the first one is over the
landslide area, if it’s going to cooperate.

Oh, I’m going to hit that.
There we go.

So this is that landslide.
Now, this isn’t a video.

This isn’t just a
image that we took.

It’s stitching all those photos
together and then draping them

over an elevation model
that we created.

But as we zoom in here,
you can start to see the accuracies

we’re getting down to
the individual rocks.

And that’s what the
capability is giving to us

because we can fly
at these low altitudes.

And the imagery and the processing
software that gives us this capability.

So we’re – this is kind of
show-and-tell-type stuff in a 3D model,

but this is – these are individual data
layers that we can put into our

geographic information system, or plot
it out if you want to work with paper.

Nothing different than
we’ve ever done before.

Just the resolution
is a whole lot better.

And a lot cheaper, actually. We’re using
just handheld cameras to do this.

If we go into Colorado,
this is the Red Rocks Amphitheater,

pretty close
to our office.

And if you’re a photogrammetrist,
this is very difficult to make these

models, but when we can fly this low,
and it’s able to handle the processing

the way it does now, we can get some
very difficult work accomplished.

And that’s the elevation
that it’s creating –

just a triangulated
little network.

And then this will be
the final one, just to look at

those pelicans from
a different perspective.

This is, once again, mosaicked
all together individual images.

With the pelicans, you can start to
see them as little white spots there,

but maybe gives the biologist a
little different angle on

how these birds are nesting,
their habitats, maybe the vegetation.

And this was
down in Nevada.

So there you have it. I believe that’s
all I have, other than –

I do have to give
credit to the folks

that I’m the project leader for,
but I’ll tell you,

that’s cream-of-the-crop guys
there that make this all work.

So we also have people in Flagstaff;
Bozeman; Fort Collins, Colorado.

And our goal within USGS and
all of Interior is to start distributing

these out so that the
scientists have them available.

Say, for instance, a park or
a wildlife refuge, they have

one available to them, and it’s
a beautiful day like it was today.

Pop it up in the air
and get some data.

They don’t have to coordinate with us
coming out or contracting or whatever.

But they can have it on-site
and use it for what it’s worth.

So that’s all I have.
Thank you.

[ Applause ]

- Thank you, Jeff.

I know many of you have questions.
You always do.

Please use one of the microphones
in either of the two aisles, not only

so that we in this room can hear your
questions, but for people who may

be watching it streamed online
can also hear your questions.

So just help yourself,
get in line, stand up by

the microphone,
and you can ask questions.

If you’re not able to do that,
wave at me, and I’ll bring you

a microphone, but I’d appreciate it
if you’d try those ones in the aisles.

Oh, good. This always
entices people to go.

Why don’t you go ahead, please?

- I don’t know. Do they ever – do they
ever use – will they ever use drones,

like, in the motion picture – for motion
pictures and movie industry?

- Yeah. As a matter of fact,
they were one of the first companies

to ask for a waiver to fly. I believe
they were granted the first waiver.

- Also, I wanted to ask
another question is,

I have a friend who once purchased
a drone at the Makers Faire,

and do you have to be –
do they have – do you have to

have a license to have to register
them for just recreational use?

- No.
Not for recreational use.

No, but you do need to
follow some of those rules.

You know, hobbyist rules
is what they’re called.

- Because I remember just looking at –
I just looked at – I just remember,

you know, seeing those –
like, I was thinking, you know –

again, I was thinking about, well,
when you’re talking about drones,

whether they would use them
in the motion picture industry.

And sometimes I wonder
whether even they’ll do –

they’ll use them for, like, the mail,
to deliver mail or [chuckles] …

- Well, yeah, possibly.
One of the most interesting uses

I heard of was from a Mayo Clinic
surgeon that actually made a

good case for transporting blood
across Rochester, Minnesota.

You know, they said when
a big trauma comes in,

they can clean out all
the local hospitals very quickly.

And the way they transport it
right now is get it in the back

of a police car and get it
there as fast as they can.

So he had proposed the idea they
set up corridors at a certain height –

maybe above a telephone line or
something like that, that this becomes

a regular transport of not only blood,
but different types of medication.

- Okay.
- It was a very good idea.

And I think we’ll
see that someday.


- I thought that was a great talk.
- Thank you.

- So thanks very much.

One very quick question, then I’ll –
well, two, actually, but – so the –

like, Red Rocks flyover and other
point maps you did, all the shading

and lighting, that’s all
totally artificial, right?

Because it’s just a point map
that you then rendered?

- The – it’s actually –
just like you take a photograph,

that’s what’s
being captured.

So your ISO settings,
your focal lengths, all that,

just like you do in a normal camera,
that’s being collected.

We’re just using that to
render the 3D model,

and then we drape
that image on top.

So it totally depends on the sunlight.
- Right.

- The condition that
you have during that day.

- Okay. And then the second one is,
you said batteries were a real problem.

Why not fly with fuel?

- For us, we have to treat it
as hazmat, and it’s a pain.

That’s probably
the biggest use.

And then trying to find it –
we have to be on-site.

In this – in this – the case of the T-Hawk
that we had, we had to find aviation fuel.

[chuckles] So we have to beg
the little airport, can we just go get,

you know, five gallons of fuel.
And sometimes they say yeah,

and sometimes they go, no, why
would you ever want to do that?

So then you got to explain
that you’re flying drones.

And so we just prefer not to.

But it also is – for us,
was a noise-type thing.

Batteries are just easy to work with.
Rechargeable – we can recharge them.

But there is use
for fuel, certainly.

Some of those long endurance,
we’re kind of wondering,

you know, solar isn’t going to be
the actual total answer to that.

There has to be some other type
of fuel source, I would think.

But I don’t know enough about that.
Yes? Oh, hi.

- How big can drones be?

- How big can they be?

The biggest I’ve seen
would fill half this room.

So that’s the Global Hawk that
you may have heard about.

It’s made by Northrop Grumman,
but it’s used by the military.

It’s also – actually, NASA has one
and NOAA has one, and they use –

they actually are using them to fly into
hurricanes – fly at very high altitudes.

But it’s a – it’s a pretty
impressive little –

or, not a little airplane,
a big airplane.

Yeah. They get pretty big.


- There’s really two exciting currents,
technologically speaking, going on now.

UAVs, UASs, drones,
whatever, is one.

The other is a big explosion
in small satellites in terms of

costs coming down,
the technology flying on – is going up.

Do you see how – I mean, this is
really putting you on the spot,

but can you see sort of a balance
of how that’s going to play out for –

at least in your particular remote
sensing/mapping part of the world?

- Yeah. Boy. I mean, there’s
going to be a lot of stuff flying.

Those nanosatellites as you’re
referring to – they’re kind of

intriguing because they’re cheap,
and any country can now get them.

I would say probably – and the
tradeoff is maybe in resolution.

Like we – like we said, the USGS
sees a role for the really small ones.

We don’t necessarily see the role
for those big ones like what – the

missions that NASA and NOAA have.
They have a little more need for that.

But for us, I mean, just to get down
on the ground and make it easily –

you know, we can steer into a canyon,
or we can steer underneath trees

in a river or things like that that the
satellite won’t ever be able to see into.

So I’m – that’s probably really
not a good answer for all that,

but there’s a lot
of data coming.

- How small can drones be?
- Ah.

- When you think of the size of a camera
in a cell phone, it could be pretty small.

- Yeah. The smallest I’ve
seen was a little Dragonfly.

It mimicked a dragonfly.

I think I’ve even seen smaller than
that since, but that was a military

application for checking out bad guys,
and you don’t even know it’s there.

I’ve seen some that look
like owls or mimic a bird.

And you don’t even know it’s flying, but
it’s going along, but yeah, very small.

You know, the cameras – even the
cameras that we use in these, the little –

you see them at the hobby shops.
They’re just tiny.

They’re like the camera
that’s in your iPhone, so …

- Right. So the ones that –
the DJIs and so on, they look as if

they might become dinosaurs
if they’re the size they are.

Is the consumer version going to
go much smaller, do you think?

- The problem with
smaller is wind, really.

I mean, they’ll go
all over the place.

And I’ve – they’ll probably figure
that out too, but right now, you know,

that’s one of the reasons we go to
a little bigger platform is stability.

The stiller we can keep our sensors,
the better we can derive that data.

So that’s one reason.

But, yeah,
they’ll probably figure that out.

- Okay.
- Stabilize imagery. Hi.

- I have two questions –
one technical, one topical.

- Okay.
- The technical is, do they make

accelerometers smaller – small and
light enough now to go into drones

to record in real time the pitch
and yaw characteristics?

- Yes.
- They do.

- As a matter of fact, we – that’s one
of the main things we were looking at.

Especially for when you
get into these more

sophisticated sensors,
like the laser.

You have to have a –
what he’s talking about is a –

I call it a IMU –
initial measurement unit.

But it measures how that
airplane or that drone is flying.

So if – take a fixed-wing.
If it’s flying like this, and it does a roll,

it records that angle.
If it does a pitch, it records that angle.

And then the yaw as it’s crabbing
along – it will record all those angles.

And we can use that in our software.
And that’s what tightens up

all that imagery is where we can
get down to the centimeter level.

But yes. The answer is yes.
They have little, tiny …

- And you still use
ground control in addition?

- Yeah. We do.
We still – there’s a big push

right now for what
they call RTK drones.

So if you see, out on the road,
surveys being done,

they’ll usually set up
a base station.

And then another guys goes around –
usually the poor guy that has to

walk around with the rover.
He’s collecting the data.

And then that – the satellites are
beaming down to the base station,

and then beaming that out and
doing a real-time correction.

So that guy that’s
walking around with the rover

is getting a very accurate
coordinate as – real-time.

So what these drones are doing is
replacing the rover that’s walking –

or that poor guy that’s walking around.
It now becomes the rover.

And we can transmit that
correction up to the airplane.

And the by the time it lands, it already –
we already have a corrected image.

And there’s a lot of interest in it
because then we can possibly

eliminate ground control.
That’s a lot of work to put that down.

As well as, when we
process in the backend,

we have to pick
that ground control.

So there’s two labor savings there,
so it’s got a lot of attention.

It’s not there yet.
It’s not down to the level that they’re –

a lot of the salesmen will claim.
But it’s coming.

- Okay. And then the topical one
is that yesterday I was seeing

images over central Italy
that were being taken from drones

to analyze the damage to structures.
- The flooding, yeah.

- And I can foresee this, you know,
being a use that USGS seismologists

and geologists – you know, on the
sort of strike force teams that go into

Nepal or something like that right away.
- Oh, definitely.

- How do you coordinate that flying
with the relief helicopters and the –

and the, you know, medical
evacuation, things like that?

- Right. And you guys probably
know better than anything,

especially with the wildfires.
When a drone – somebody

pops one up – the Forest Service
or the DOI – they have to

shut down all their operations.
And it’s happened way too often.

And it’s got a lot of attention,
but the same thing – over a flood,

or over any type of disaster,
that’s a pretty busy airspace.

And we’re, at times,
reluctant to get in there.

We have the – during the Colorado
flooding, we tried to get in there,

and we knew we could help,
but it’s just not worth it.

If we crash into –
if we bring down an aircraft,

that would be the end
of what we’re doing.

- So you’re not operating on
radio frequencies where

you can communicate?
- Well, yeah. We could.

What’s happened just lately –
there was a White House meeting that –

for the good use of drones.
And actually, there was some charges

given out to the DOI specific to wildfire,
but also any other disaster-type relief.

What they did policy-wise is
set up a blanket certificate

of authorization so that we can
literally file within a 24-hour period

and then possibly get out to
the location and be able to fly.

So they’re paving the way to do that.
But yeah, it’s a big coordination effort.

They’ll set up, you know,
temporary restricted flight zones.

That’s one way to operate.
But you really have to coordinate

with everybody there because the
pilots that are – that are flying don’t

necessarily like those drones. [chuckles]
You know, and I don’t blame them.

- Sure. Yeah, no, can’t blame them.

- I was just going to
add a comment to that.

In all of those kind of natural disasters –
actually, they’re not natural disasters,

but hazards – volcanoes, earthquakes,
floods – USGS scientists are

not the first people to
rush in and want to help.

And that’s because search-and-rescue
efforts are always first.

And we don’t want to
get into that – in their way.

So we don’t come until
we’re invited, and we’re definitely

not the first ones in there,
and that’s very intentional.

- Just curious on your
consumer-grade cameras

that you tried to use
for photogrammetry.

Do you, before you fly them,
calibrate them in any way?

You know, take shots against
known grids and look for

aberrations and what not?
- Yep. That’s a good question.

Because in the past,
the USGS held a

camera calibration lab
at Sioux Falls, South Dakota.

So all the metric cameras –
all the mapping cameras are

usually sent there
annually to get a calibration.

They were getting
radiometric corrections

across that lens to
incredibly accurate levels.

This new world order does a
camera calibration on the fly.

It’s spectacular.

It will actually calibrate that camera
an opposite way, and it’s getting –

when we run our software,
we’ll get a report that will

give us our camera calibration.
And it’s very accurate.

And it’s actually better than a lab setting.
So some arguments are being made

that we now have a better camera
calibration, even with these little guys.

- Have you started taking repeated
flyovers of areas where landforms

are changing, and particularly rapidly,
like, for example, you know,

looking at the earthquake fault
movements, loss of wetlands

in Louisiana, you know,
urban development,

you know, erosion,
et cetera, et cetera?

Making motion pictures
of these kinds of processes.

- Yeah. Two examples –
or three that I can think of.

Like the Cape Cod – that’s exactly
what they’re trying to do.

- Okay.
- Get a baseline – base photo.

And then they’ll come back in
right after a storm and then

compare it – that type of stuff.
- Okay.

- Our landslide – that’s what
we’re doing there.

We’ve been out there four times.
But that’s just a few months apart.

What we really want to do
is a year or two apart, but …

- Right.

- … in general, almost everybody
wants us back out there because

that’s exactly what
they want to do.

They want to see land
change over time.

- Okay, thanks.
- Yep.

- My question is about the power
source you use on the drones.

- Mm-hmm.
- Are they mostly, like, battery?

And how long
does it usually fly?

And maybe the current status and the
future outlook on the power source.

- Yeah. All the vehicles we have now
are battery-powered – lithium-ion.

So we recharge them in the field.
So we take a lot of batteries with us.

Normally, we – for a hover-type
airplane, one of those that’s

a quadcopter-type, you’re lucky
if you get a half-hour.

For a fixed-wing, usually you
can push up to an hour.

Our big helicopter,
we’re about 45 minutes.

So still you can do a lot of work,
but once again, that’s going to be

a small area that
you’re working with.

Batteries – whoever figures
out a long-duration battery

is going to make
a lot of money.

- What’s the percentage of the battery
weight compared – I guess it varies.

- Yeah. Weight is always
our biggest issue.

So if you go – after we’re done here,
go lift that – the Raven.

Half that weight is battery.

And you ought to see the size of
the batteries that we have with our –

with our Pulse –
the helicopter.

I mean, it’s literally two stacks of
batteries this heavy.

It’s very heavy.
But that’s our power, so …

- Thank you.
- Yes.

- Thank you for a wonderful
round of questions.

Thank you, Jeff.
- Yes.

- Excellent talk again.
And please join us next month.

[ Applause ]

[inaudible background conversations]