0:00:01.713,0:00:03.079
Can I get a show of hands --

0:00:03.103,0:00:06.805
how many of you in this room[br]have been on a plane in this past year?

0:00:08.258,0:00:09.411
That's pretty good.

0:00:09.435,0:00:12.320
Well, it turns out that you[br]share that experience

0:00:12.344,0:00:15.179
with more than three billion[br]people every year.

0:00:15.203,0:00:18.340
And when we put so many people[br]in all these metal tubes

0:00:18.364,0:00:19.950
that fly all over the world,

0:00:19.974,0:00:22.647
sometimes, things like this can happen

0:00:22.671,0:00:24.558
and you get a disease epidemic.

0:00:25.116,0:00:27.067
I first actually got into this topic

0:00:27.091,0:00:29.723
when I heard about the Ebola[br]outbreak last year.

0:00:29.747,0:00:31.207
And it turns out that,

0:00:31.231,0:00:34.061
although Ebola spreads[br]through these more range-limited,

0:00:34.085,0:00:35.419
large-droplet routes,

0:00:35.443,0:00:37.421
there's all these other sorts of diseases

0:00:37.445,0:00:39.421
that can be spread in the airplane cabin.

0:00:39.445,0:00:42.629
The worst part is, when we take[br]a look at some of the numbers,

0:00:42.653,0:00:44.066
it's pretty scary.

0:00:44.090,0:00:45.850
So with H1N1,

0:00:45.874,0:00:48.263
there was this guy that decided[br]to go on the plane

0:00:48.287,0:00:50.058
and in the matter of a single flight

0:00:50.082,0:00:52.336
actually spread the disease[br]to 17 other people.

0:00:52.360,0:00:54.488
And then there was this[br]other guy with SARS,

0:00:54.512,0:00:56.614
who managed to go on a three-hour flight

0:00:56.638,0:00:59.480
and spread the disease to 22 other people.

0:00:59.504,0:01:02.912
That's not exactly my idea[br]of a great superpower.

0:01:03.658,0:01:06.222
When we take a look at this,[br]what we also find

0:01:06.246,0:01:09.222
is that it's very difficult[br]to pre-screen for these diseases.

0:01:09.619,0:01:11.710
So when someone actually[br]goes on a plane,

0:01:11.734,0:01:12.940
they could be sick

0:01:12.964,0:01:15.359
and they could actually[br]be in this latency period

0:01:15.383,0:01:17.542
in which they could actually[br]have the disease

0:01:17.566,0:01:19.138
but not exhibit any symptoms,

0:01:19.162,0:01:21.353
and they could, in turn,[br]spread the disease

0:01:21.377,0:01:23.050
to many other people in the cabin.

0:01:23.074,0:01:25.156
How that actually works is that right now

0:01:25.180,0:01:27.466
we've got air coming in[br]from the top of the cabin

0:01:27.490,0:01:29.916
and from the side of the cabin,[br]as you see in blue.

0:01:29.940,0:01:34.142
And then also, that air goes out[br]through these very efficient filters

0:01:34.166,0:01:38.714
that eliminate 99.97 percent[br]of pathogens near the outlets.

0:01:39.444,0:01:40.921
What happens right now, though,

0:01:40.945,0:01:43.012
is that we have this[br]mixing airflow pattern.

0:01:43.036,0:01:44.868
So if someone were to actually sneeze,

0:01:44.892,0:01:47.596
that air would get swirled[br]around multiple times

0:01:47.620,0:01:50.865
before it even has a chance[br]to go out through the filter.

0:01:51.785,0:01:54.998
So I thought: clearly, this[br]is a pretty serious problem.

0:01:55.022,0:01:58.755
I didn't have the money[br]to go out and buy a plane,

0:01:58.779,0:02:01.017
so I decided to build a computer instead.

0:02:01.041,0:02:04.313
It actually turns out that[br]with computational fluid dynamics,

0:02:04.337,0:02:06.938
what we're able to do[br]is create these simulations

0:02:06.962,0:02:08.756
that give us higher resolutions

0:02:08.780,0:02:12.400
than actually physically going[br]in and taking readings in the plane.

0:02:12.836,0:02:15.850
And so how, essentially, this works[br]is you would start out

0:02:15.874,0:02:17.546
with these 2D drawings --

0:02:17.570,0:02:20.698
these are floating around[br]in technical papers around the Internet.

0:02:20.722,0:02:23.615
I take that and then I put it[br]into this 3D-modeling software,

0:02:23.639,0:02:25.418
really building that 3D model.

0:02:25.442,0:02:29.901
And then I divide that model[br]that I just built into these tiny pieces,

0:02:29.925,0:02:33.502
essentially meshing it so that[br]the computer can better understand it.

0:02:33.526,0:02:37.247
And then I tell the computer where[br]the air goes in and out of the cabin,

0:02:37.271,0:02:38.770
throw in a bunch of physics

0:02:38.794,0:02:43.015
and basically sit there and wait until[br]the computer calculates the simulation.

0:02:44.015,0:02:47.642
So what we get, actually,[br]with the conventional cabin is this:

0:02:47.666,0:02:49.913
you'll notice the middle person sneezing,

0:02:50.767,0:02:54.159
and we go "Splat!" -- it goes[br]right into people's faces.

0:02:54.882,0:02:56.703
It's pretty disgusting.

0:02:56.727,0:02:59.075
From the front, you'll notice[br]those two passengers

0:02:59.099,0:03:00.885
sitting next to the central passenger

0:03:00.909,0:03:02.615
not exactly having a great time.

0:03:02.639,0:03:04.825
And when we take a look[br]at that from the side,

0:03:04.849,0:03:08.842
you'll also notice those pathogens[br]spreading across the length of the cabin.

0:03:10.017,0:03:12.367
The first thing I thought was,[br]"This is no good."

0:03:12.391,0:03:15.899
So I actually conducted[br]more than 32 different simulations

0:03:15.923,0:03:19.288
and ultimately, I came up[br]with this solution right here.

0:03:19.312,0:03:22.828
This is what I call a -- patent pending --[br]Global Inlet Director.

0:03:22.852,0:03:25.418
With this, we're able to reduce[br]pathogen transmission

0:03:25.442,0:03:27.210
by about 55 times,

0:03:27.234,0:03:30.387
and increase fresh-air inhalation[br]by about 190 percent.

0:03:30.411,0:03:32.015
So how this actually works

0:03:32.039,0:03:35.168
is we would install this piece[br]of composite material

0:03:35.192,0:03:38.160
into these existing spots[br]that are already in the plane.

0:03:38.184,0:03:40.185
So it's very cost-effective to install

0:03:40.209,0:03:42.057
and we can do this directly overnight.

0:03:42.081,0:03:45.629
All we have to do is put a couple[br]of screws in there and you're good to go.

0:03:45.653,0:03:48.512
And the results that we get[br]are absolutely amazing.

0:03:48.536,0:03:52.072
Instead of having those problematic[br]swirling airflow patterns,

0:03:52.096,0:03:53.838
we can create these walls of air

0:03:53.862,0:03:56.032
that come down in-between the passengers

0:03:56.056,0:03:57.954
to create personalized breathing zones.

0:03:57.978,0:04:00.963
So you'll notice the middle passenger[br]here is sneezing again,

0:04:00.987,0:04:03.700
but this time, we're able[br]to effectively push that down

0:04:03.724,0:04:06.437
to the filters for elimination.

0:04:06.461,0:04:07.857
And same thing from the side,

0:04:07.881,0:04:11.110
you'll notice we're able to directly[br]push those pathogens down.

0:04:11.682,0:04:15.176
So if you take a look again now[br]at the same scenario

0:04:15.200,0:04:16.889
but with this innovation installed,

0:04:16.913,0:04:18.929
you'll notice the middle[br]passenger sneezes,

0:04:18.953,0:04:22.026
and this time, we're pushing[br]that straight down into the outlet

0:04:22.050,0:04:25.771
before it gets a chance[br]to infect any other people.

0:04:25.795,0:04:28.919
So you'll notice the two passengers[br]sitting next to the middle guy

0:04:28.943,0:04:31.176
are breathing virtually[br]no pathogens at all.

0:04:31.200,0:04:33.728
Take a look at that from the side as well,

0:04:33.752,0:04:35.307
you see a very efficient system.

0:04:35.331,0:04:37.953
And in short, with this system, we win.

0:04:39.255,0:04:42.144
When we take a look at what this means,

0:04:42.168,0:04:45.640
what we see is that this not only works[br]if the middle passenger sneezes,

0:04:45.664,0:04:48.438
but also if the window-seat[br]passenger sneezes

0:04:48.462,0:04:50.557
or if the aisle-seat passenger sneezes.

0:04:51.167,0:04:54.254
And so with this solution, what does[br]this mean for the world?

0:04:54.278,0:04:57.792
Well, when we take a look at this

0:04:57.816,0:05:00.385
from the computer simulation[br]into real life,

0:05:00.409,0:05:03.171
we can see with this 3D model[br]that I built over here,

0:05:03.195,0:05:05.283
essentially using 3D printing,

0:05:05.307,0:05:08.266
we can see those same[br]airflow patterns coming down,

0:05:08.290,0:05:09.876
right to the passengers.

0:05:10.920,0:05:13.990
In the past, the SARS epidemic[br]actually cost the world

0:05:14.014,0:05:15.943
about 40 billion dollars.

0:05:15.967,0:05:17.126
And in the future,

0:05:17.150,0:05:19.696
a big disease outbreak[br]could actually cost the world

0:05:19.720,0:05:21.578
in excess of three trillion dollars.

0:05:21.942,0:05:25.419
So before, it used to be that you had[br]to take an airplane out of service

0:05:25.443,0:05:27.315
for one to two months,

0:05:27.339,0:05:30.911
spend tens of thousands of man hours[br]and several million dollars

0:05:30.935,0:05:32.258
to try to change something.

0:05:32.282,0:05:35.793
But now, we're able to install[br]something essentially overnight

0:05:35.817,0:05:37.544
and see results right away.

0:05:37.568,0:05:40.774
So it's really now a matter of taking[br]this through to certification,

0:05:40.798,0:05:41.988
flight testing,

0:05:42.012,0:05:45.004
and going through all of these[br]regulatory approvals processes.

0:05:45.028,0:05:48.092
But it just really goes to show[br]that sometimes the best solutions

0:05:48.116,0:05:49.554
are the simplest solutions.

0:05:49.935,0:05:53.125
And two years ago, even,

0:05:53.149,0:05:54.918
this project would not have happened,

0:05:54.942,0:05:57.768
just because the technology then[br]wouldn't have supported it.

0:05:57.792,0:06:00.261
But now with advanced computing

0:06:00.285,0:06:02.471
and how developed our Internet is,

0:06:02.495,0:06:05.134
it's really the golden era for innovation.

0:06:05.158,0:06:08.401
And so the question I ask all[br]of you today is: why wait?

0:06:08.425,0:06:10.746
Together, we can build the future today.

0:06:11.123,0:06:12.274
Thanks.

0:06:12.298,0:06:15.404
(Applause)