25:02
TED2012

Regina Dugan: From mach-20 glider to hummingbird drone

Filmed:

"What would you attempt to do if you knew you could not fail?" asks Regina Dugan, then director of DARPA, the Defense Advanced Research Projects Agency. In this breathtaking talk she describes some of the extraordinary projects -- a robotic hummingbird, a prosthetic arm controlled by thought, and, well, the internet -- that her agency has created by not worrying that they might fail. (Followed by a Q&A with TED's Chris Anderson)

- Former director of DARPA
As director of the Defense Advanced Research Projects Agency (DARPA), Regina Dugan oversaw the US armed forces' innovation engine. Now she deploys the same research tactics at Google. Full bio

You should be nice
00:15
to nerds.
00:17
In fact, I'd go so far as to say,
00:19
if you don't already have a nerd in your life,
00:21
you should get one.
00:23
I'm just saying.
00:26
Scientists and engineers
00:28
change the world.
00:30
I'd like to tell you
00:33
about a magical place called DARPA
00:35
where scientists and engineers
00:38
defy the impossible
00:40
and refuse to fear failure.
00:42
Now these two ideas
00:45
are connected more than you may realize,
00:47
because when you remove the fear of failure,
00:50
impossible things
00:54
suddenly become possible.
00:56
If you want to know how,
01:00
ask yourself this question:
01:02
What would you attempt to do
01:05
if you knew you could not fail?
01:07
If you really ask yourself
01:10
this question,
01:12
you can't help but feel uncomfortable.
01:14
I feel a little uncomfortable.
01:17
Because when you ask it,
01:20
you begin to understand
01:22
how the fear of failure constrains you,
01:24
how it keeps us
01:27
from attempting great things,
01:29
and life gets dull,
01:32
amazing things stop happening.
01:35
Sure, good things happen,
01:38
but amazing things
01:40
stop happening.
01:42
Now I should be clear,
01:44
I'm not encouraging failure,
01:47
I'm discouraging
01:50
fear of failure.
01:52
Because it's not failure itself
01:55
that constrains us.
01:57
The path to truly new,
01:59
never-been-done-before things
02:01
always has failure along the way.
02:03
We're tested.
02:06
And in part, that testing feels an appropriate part
02:08
of achieving something great.
02:12
Clemenceau said,
02:15
"Life gets interesting when we fail,
02:17
because it's a sign
02:20
that we've surpassed ourselves."
02:22
In 1895,
02:25
Lord Kelvin declared
02:27
that heavier-than-air flying machines
02:29
were impossible.
02:31
In October of 1903,
02:33
the prevailing opinion
02:35
of expert aerodynamicists
02:37
was that maybe in 10 million years
02:40
we could build an aircraft that would fly.
02:42
And two months later on December 17th,
02:45
Orville Wright powered the first airplane
02:48
across a beach in North Carolina.
02:51
The flight lasted 12 seconds
02:54
and covered 120 feet.
02:57
That was 1903.
03:00
One year later,
03:02
the next declarations of impossibilities began.
03:04
Ferdinand Foch, a French army general
03:07
credited with having one of the most original and subtle minds
03:10
in the French army,
03:14
said, "Airplanes are interesting toys,
03:16
but of no military value."
03:19
40 years later,
03:23
aero experts coined the term transonic.
03:25
They debated, should it have one S or two?
03:28
You see, they were having trouble in this flight regime,
03:33
and it wasn't at all clear
03:36
that we could fly faster than the speed of sound.
03:38
In 1947,
03:42
there was no wind tunnel data
03:44
beyond Mach 0.85.
03:46
And yet,
03:51
on Tuesday, October 14th, 1947,
03:53
Chuck Yeager climbed into the cockpit
03:58
of his Bell X-1
04:01
and he flew
04:03
towards an unknown possibility,
04:05
and in so doing,
04:09
he became the first pilot
04:11
to fly faster than the speed of sound.
04:13
Six of eight Atlas rockets
04:22
blew up on the pad.
04:24
After 11 complete mission failures,
04:26
we got our first images from space.
04:28
And on that first flight
04:30
we got more data
04:32
than in all U-2 missions combined.
04:34
It took a lot of failures
04:39
to get there.
04:41
Since we took to the sky,
04:43
we have wanted to fly
04:45
faster and farther.
04:47
And to do so,
04:50
we've had to believe in impossible things.
04:52
And we've had to refuse
04:55
to fear failure.
04:57
That's still true today.
05:00
Today, we don't talk about flying transonically,
05:03
or even supersonically,
05:07
we talk about flying hypersonically --
05:10
not Mach 2 or Mach 3, Mach 20.
05:13
At Mach 20,
05:17
we can fly from New York to Long Beach
05:19
in 11 minutes and 20 seconds.
05:21
At that speed,
05:24
the surface of the airfoil
05:26
is the temperature of molten steel --
05:28
3,500 degrees Fahrenheit --
05:30
like a blast furnace.
05:33
We are essentially burning the airfoil
05:35
as we fly it.
05:38
And we are flying it,
05:40
or trying to.
05:42
DARPA's hypersonic test vehicle
05:44
is the fastest maneuvering aircraft
05:47
ever built.
05:49
It's boosted to near-space
05:51
atop a Minotaur IV rocket.
05:54
Now the Minotaur IV has too much impulse,
05:56
so we have to bleed it off
05:58
by flying the rocket
06:00
at an 89 degree angle of attack
06:02
for portions of the trajectory.
06:04
That's an unnatural act
06:07
for a rocket.
06:09
The third stage has a camera.
06:11
We call it rocketcam.
06:14
And it's pointed
06:16
at the hypersonic glider.
06:18
This is the actual rocketcam footage
06:21
from flight one.
06:24
Now to conceal the shape, we changed the aspect ratio a little bit.
06:26
But this is what it looks like
06:29
from the third stage of the rocket
06:32
looking at the unmanned glider
06:34
as it heads into the atmosphere
06:36
back towards Earth.
06:39
We've flown twice.
06:41
In the first flight,
06:43
no aerodynamic control of the vehicle.
06:45
But we collected more hypersonic flight data
06:48
than in 30 years
06:51
of ground-based testing combined.
06:53
And in the second flight,
06:56
three minutes of fully-controlled,
06:59
aerodynamic flight
07:02
at Mach 20.
07:04
We must fly again,
07:06
because amazing, never-been-done-before things
07:08
require that you fly.
07:12
You can't learn to fly at Mach 20
07:15
unless you fly.
07:18
And while there's no substitute for speed,
07:21
maneuverability is a very close second.
07:24
If a Mach 20 glider takes 11 minutes and 20 seconds
07:28
to get from New York to Long Beach,
07:31
a hummingbird would take,
07:33
well, days.
07:35
You see, hummingbirds are not hypersonic,
07:38
but they are maneuverable.
07:40
In fact, the hummingbird is the only bird
07:43
that can fly backwards.
07:46
It can fly up, down,
07:48
forwards, backwards,
07:50
even upside-down.
07:52
And so if we wanted to fly in this room
07:55
or places where humans can't go,
07:58
we'd need an aircraft
08:00
small enough and maneuverable enough
08:02
to do so.
08:05
This is a hummingbird drone.
08:07
It can fly in all directions,
08:10
even backwards.
08:12
It can hover and rotate.
08:14
This prototype aircraft
08:17
is equipped with a video camera.
08:19
It weighs less than one AA battery.
08:22
It does not eat nectar.
08:26
In 2008,
08:29
it flew for a whopping 20 seconds,
08:31
a year later, two minutes,
08:34
then six,
08:36
eventually 11.
08:38
Many prototypes crashed -- many.
08:40
But there's no way
08:44
to learn to fly like a hummingbird
08:46
unless you fly.
08:48
(Applause)
08:56
It's beautiful, isn't it.
09:04
Wow.
09:08
It's great.
09:10
Matt is the first ever hummingbird pilot.
09:15
(Applause)
09:18
Failure is part of creating
09:25
new and amazing things.
09:28
We cannot both fear failure
09:30
and make amazing new things --
09:33
like a robot
09:37
with the stability of a dog on rough terrain,
09:39
or maybe even ice;
09:42
a robot that can run like a cheetah,
09:44
or climb stairs like a human
09:46
with the occasional clumsiness of a human.
09:49
Or perhaps, Spider Man
09:54
will one day be Gecko Man.
09:57
A gecko can support
10:00
its entire body weight
10:02
with one toe.
10:04
One square millimeter of a gecko's footpad
10:06
has 14,000 hair-like structures
10:09
called setae.
10:12
They are used to help it grip to surfaces
10:14
using intermolecular forces.
10:17
Today we can manufacture structures
10:20
that mimic the hairs of a gecko's foot.
10:23
The result,
10:26
a four-by-four-inch
10:28
artificial nano-gecko adhesive.
10:30
can support a static load
10:34
of 660 pounds.
10:36
That's enough to stick
10:38
six 42-inch plasma TV's to your wall,
10:40
no nails.
10:43
So much for Velcro, right?
10:45
And it's not just passive structures,
10:48
it's entire machines.
10:51
This is a spider mite.
10:54
It's one millimeter long,
10:56
but it looks like Godzilla
10:58
next to these micromachines.
11:00
In the world of Godzilla spider mites,
11:03
we can make millions of mirrors,
11:06
each one-fifth the diameter
11:09
of a human hair,
11:11
moving at hundreds of thousands of times per second
11:13
to make large screen displays,
11:16
so that we can watch movies like "Godzilla"
11:19
in high-def.
11:21
And if we can build machines
11:24
at that scale,
11:26
what about Eiffel Tower-like trusses
11:28
at the microscale?
11:31
Today we are making metals
11:33
that are lighter than Styrofoam,
11:36
so light
11:38
they can sit atop a dandelion puff
11:40
and be blown away
11:42
with a wisp of air --
11:44
so light
11:46
that you can make a car that two people can lift,
11:48
but so strong
11:51
that it has the crash-worthiness of an SUV.
11:53
From the smallest wisp of air
11:56
to the powerful forces of nature's storms.
11:59
There are 44 lightning strikes per second
12:02
around the globe.
12:05
Each lightning bolt heats the air
12:07
to 44,000 degrees Fahrenheit --
12:10
hotter than the surface of the Sun.
12:12
What if we could use
12:15
these electromagnetic pulses
12:17
as beacons,
12:19
beacons in a moving network
12:21
of powerful transmitters?
12:24
Experiments suggest
12:27
that lightning could be the next GPS.
12:29
Electrical pulses form the thoughts in our brains.
12:34
Using a grid the size of your thumb,
12:37
with 32 electrodes
12:40
on the surface of his brain,
12:42
Tim uses his thoughts
12:44
to control an advanced prosthetic arm.
12:46
And his thoughts
12:52
made him reach for Katie.
12:54
This is the first time
12:56
a human has controlled a robot
13:00
with thought alone.
13:02
And it is the first time
13:05
that Tim has held Katie's hand
13:07
in seven years.
13:09
That moment mattered
13:11
to Tim and Katie,
13:13
and this green goo
13:15
may someday matter to you.
13:17
This green goo
13:20
is perhaps the vaccine that could save your life.
13:22
It was made in tobacco plants.
13:25
Tobacco plants
13:28
can make millions of doses of vaccine
13:30
in weeks instead of months,
13:33
and it might just be
13:36
the first healthy use of tobacco ever.
13:38
And if it seems far-fetched
13:42
that tobacco plants could make people healthy,
13:44
what about gamers that could solve problems
13:47
that experts can't solve?
13:50
Last September,
13:52
the gamers of Foldit
13:54
solved the three-dimensional structure
13:56
of the retroviral protease
13:58
that contributes to AIDS in rhesus monkeys.
14:00
Now understanding this structure
14:03
is very important for developing treatments.
14:05
For 15 years,
14:09
it was unsolved
14:11
in the scientific community.
14:13
The gamers of Foldit
14:15
solved it in 15 days.
14:17
Now they were able to do so
14:20
by working together.
14:22
They were able to work together
14:24
because they're connected by the Internet.
14:26
And others, also connected to the Internet,
14:28
used it as an instrument of democracy.
14:31
And together
14:35
they changed the fate of their nation.
14:37
The Internet is home to two billion people,
14:40
or 30 percent of the world's population.
14:43
It allows us to contribute
14:46
and to be heard
14:49
as individuals.
14:51
It allows us to amplify
14:53
our voices and our power
14:56
as a group.
14:59
But it too had humble beginnings.
15:01
In 1969, the internet was but a dream,
15:04
a few sketches on a piece of paper.
15:07
And then on October 29th,
15:09
the first packet-switched message was sent
15:12
from UCLA to SRI.
15:14
The first two letters of the word "Login,"
15:17
that's all that made it through --
15:20
an L and an O --
15:22
and then a buffer overflow crashed the system.
15:24
(Laughter)
15:26
Two letters,
15:29
an L and an O,
15:31
now a worldwide force.
15:34
So who are these scientists and engineers
15:38
at a magical place called DARPA?
15:41
They are nerds,
15:44
and they are heroes among us.
15:47
They challenge existing perspectives
15:50
at the edges of science
15:52
and under the most demanding of conditions.
15:54
They remind us
15:58
that we can change the world
16:00
if we defy the impossible
16:02
and we refuse to fear failure.
16:04
They remind us
16:08
that we all have nerd power.
16:10
Sometimes we just forget.
16:13
You see, there was a time
16:17
when you weren't afraid of failure,
16:19
when you were a great artist or a great dancer
16:21
and you could sing, you were good at math,
16:24
you could build things, you were an astronaut,
16:26
an adventurer, Jacques Cousteau,
16:29
you could jump higher, run faster,
16:31
kick harder than anyone.
16:33
You believed in impossible things
16:36
and you were fearless.
16:38
You were totally and completely in touch
16:40
with your inner superhero.
16:44
Scientists and engineers
16:48
can indeed change the world.
16:51
So can you.
16:54
You were born to.
16:56
So go ahead,
16:58
ask yourself,
17:00
what would you attempt to do
17:02
if you knew you could not fail?
17:05
Now I want to say,
17:07
this is not easy.
17:09
It's hard to hold onto this feeling,
17:12
really hard.
17:14
I guess in some way,
17:16
I sort of believe it's supposed to be hard.
17:18
Doubt and fear always creep in.
17:20
We think someone else, someone smarter than us,
17:23
someone more capable,
17:26
someone with more resources will solve that problem.
17:28
But there isn't anyone else;
17:31
there's just you.
17:33
And if we're lucky,
17:36
in that moment,
17:38
someone steps into that doubt and fear,
17:40
takes a hand and says,
17:43
"Let me help you believe."
17:46
Jason Harley did that for me.
17:49
Jason started at DARPA
17:52
on March 18th, 2010.
17:55
He was with our transportation team.
17:57
I saw Jason nearly every day,
18:01
sometimes twice a day.
18:03
And more so than most,
18:05
he saw the highs and the lows,
18:07
the celebrations and the disappointments.
18:10
And on one particularly dark day for me,
18:14
Jason sat down
18:18
and he wrote an email.
18:20
He was encouraging,
18:23
but firm.
18:25
And when he hit send,
18:27
he probably didn't realize what a difference it would make.
18:29
It mattered to me.
18:32
In that moment
18:34
and still today
18:36
when I doubt,
18:39
when I feel afraid,
18:41
when I need to reconnect
18:44
with that feeling,
18:46
I remember his words,
18:48
they were so powerful.
18:51
Text: "There is only time enough to iron your cape
18:53
and back to the skies for you."
18:56
♫ Superhero, superhero. ♫
19:00
♫ Superhero, superhero. ♫
19:02
♫ Superhero, superhero. ♫
19:07
♫ Superhero, superhero. ♫
19:10
♫ Superhero, superhero. ♫
19:12
Voice: Because that's what being a superhero is all about.
19:18
RD: "There is only time enough
19:24
to iron your cape
19:26
and back to the skies for you."
19:28
And remember,
19:33
be nice to nerds.
19:35
(Applause)
19:37
Thank you. Thank you.
19:52
(Applause)
19:54
Chris Anderson: Regina, thank you.
19:56
I have a couple of questions.
19:58
So that glider of yours,
20:00
the Mach 20 glider,
20:02
the first one, no control, it ended up in the Pacific I think somewhere.
20:04
RD: Yeah, yeah. It did. (CA: What happened on that second flight?)
20:07
Yeah, it also went into the Pacific. (CA: But this time under control?)
20:10
We didn't fly it into the Pacific.
20:14
No, there are multiple portions of the trajectory
20:17
that are demanding
20:20
in terms of really flying at that speed.
20:22
And so in the second flight,
20:25
we were able to get three minutes
20:28
of fully aerodynamic control of the vehicle before we lost it.
20:30
CA: I imagine you're not planning to open up to passenger service
20:34
from New York to Long Beach anytime soon.
20:37
RD: It might be a little warm.
20:39
CA: What do you picture that glider being used for?
20:42
RD: Well our responsibility
20:45
is to develop the technology for this.
20:47
How it's ultimately used
20:49
will be determined by the military.
20:51
Now the purpose of the vehicle though,
20:53
the purpose of the technology,
20:56
is to be able to reach anywhere in the world
20:58
in less than 60 minutes.
21:00
CA: And to carry a payload
21:02
of more than a few pounds? (RD: Yeah.)
21:04
Like what's the payload it could carry?
21:06
RD: Well I don't think we ultimately know what it will be, right.
21:10
We've got to fly it first.
21:13
CA: But not necessarily just a camera?
21:15
RD: No, not necessarily just a camera.
21:17
CA: It's amazing.
21:21
The hummingbird?
21:24
RD: Yeah?
21:26
CA: I'm curious, you started your beautiful sequence on flight
21:28
with a plane kind of trying to flap its wings
21:32
and failing horribly,
21:34
and there haven't been that many planes built since
21:36
that flap wings.
21:39
Why did we think that this was the time to go biomimicry
21:41
and copy a hummingbird?
21:44
Isn't that a very expensive solution
21:46
for a small maneuverable flying object?
21:48
RD: So I mean, in part,
21:52
we wondered if it was possible to do it.
21:54
And you have to revisit these questions
21:56
over time.
21:58
The folks at AeroVironment
22:00
tried 300 or more different wing designs,
22:02
12 different forms of the avionics.
22:05
It took them 10 full prototypes
22:08
to get something that would actually fly.
22:10
But there's something really interesting
22:12
about a flying machine
22:15
that looks like something you'd recognize.
22:17
So we often talk about stealth
22:20
as a means for avoiding any type of sensing,
22:22
but when things looks just natural,
22:25
you also don't see them.
22:28
CA: Ah. So it's not necessarily just the performance.
22:30
It's partly the look. (RD: Sure.)
22:33
It's actually, "Look at that cute hummingbird
22:35
flying into my headquarters."
22:37
(Laughter)
22:39
Because I think, as well as the awe of looking at that,
22:43
I'm sure some people here are thinking,
22:45
technology catches up so quick,
22:47
how long is it
22:49
before some crazed geek with a little remote control
22:51
flies one through a window of the White House?
22:53
I mean, do you worry about the Pandora's box issue here?
22:55
RD: Well look, our singular mission
22:59
is the creation and prevention of strategic surprise.
23:03
That's what we do.
23:06
It would be inconceivable
23:08
for us to do that work
23:10
if we didn't make people excited and uncomfortable with the things that we do
23:13
at the same time.
23:16
It's just the nature of what we do.
23:18
Now our responsibility
23:21
is to push that edge.
23:23
And we have to be, of course, mindful and responsible
23:25
of how the technology is developed
23:29
and ultimately used,
23:31
but we can't simply close our eyes
23:33
and pretend that it isn't advancing; it's advancing.
23:35
CA: I mean, you're clearly a really inspiring leader.
23:38
And you persuade people
23:42
to go to these great feats of invention,
23:44
but at a personal level,
23:46
in a way I can't imagine doing your job.
23:48
Do you wake up in the night sometimes,
23:50
just asking questions
23:52
about the possibly unintended consequences
23:54
of your team's brilliance?
23:56
RD: Sure.
23:58
I think you couldn't be human
24:00
if you didn't ask those questions.
24:03
CA: How do you answer them?
24:05
RD: Well I don't always have answers for them, right.
24:07
I think that we learn
24:11
as time goes on.
24:14
My job is one of the most exhilarating jobs you could have.
24:16
I work with some of the most amazing people.
24:22
And with that exhilaration,
24:26
comes a really deep sense
24:28
of responsibility.
24:30
And so you have on the one hand
24:32
this tremendous lift
24:34
of what's possible
24:37
and this tremendous seriousness
24:40
of what it means.
24:43
CA: Regina, that was jaw-dropping, as they say.
24:45
Thank you so much for coming to TED. (RD: Thank you.)
24:48
(Applause)
24:51

▲Back to top

About the Speaker:

Regina Dugan - Former director of DARPA
As director of the Defense Advanced Research Projects Agency (DARPA), Regina Dugan oversaw the US armed forces' innovation engine. Now she deploys the same research tactics at Google.

Why you should listen

Businesswoman and technology developer Regina Dugan achieved national prominence when she became the first woman in charge of Darpa, the Pentagon's research arm. Dugan earned a reputation as a motivator and creative thinker, spurring non-traditional projects like a nationwide contest to find hidden balloons in order to test the power of social networks for intelligence gathering.

In a previous stint at the agency, Dugan investigated mine-detection technology, and her work wasn't limited to the lab. In Mozambique she drove a mine-clearing vehicle during a live detection exercise. Dugan currently is senior vice president at Motorola, where her mission is to advance the company's technology, which one day may include password tattoos. Ask her and she may show you hers.

More profile about the speaker
Regina Dugan | Speaker | TED.com