0:00:01.317,0:00:03.809
I am an astrophysicist.

0:00:03.833,0:00:06.888
I research stellar explosions[br]across the universe.

0:00:07.603,0:00:08.803
But I have a flaw:

0:00:09.325,0:00:11.808
I'm restless, and I get bored easily.

0:00:12.356,0:00:15.552
And although as an astrophysicist,[br]I have the incredible opportunity

0:00:15.576,0:00:17.508
to study the entire universe,

0:00:17.532,0:00:20.880
the thought of doing[br]only that, always that,

0:00:20.904,0:00:23.135
makes me feel caged and limited.

0:00:24.762,0:00:28.676
What if my issues with[br]keeping attention and getting bored

0:00:28.700,0:00:30.182
were not a flaw, though?

0:00:30.206,0:00:32.873
What if I could turn them into an asset?

0:00:33.830,0:00:36.475
An astrophysicist cannot[br]touch or interact with

0:00:36.499,0:00:38.037
the things that she studies.

0:00:38.061,0:00:41.774
No way to explode a star in a lab[br]to figure out why or how it blew up.

0:00:42.164,0:00:44.694
Just pictures and movies of the sky.

0:00:45.339,0:00:47.560
Everything we know about the universe,

0:00:47.584,0:00:50.675
from the big bang[br]that originated space and time,

0:00:50.699,0:00:53.493
to the formation and evolution[br]of stars and galaxies,

0:00:53.517,0:00:55.977
to the structure of our own solar system,

0:00:56.001,0:00:58.801
we figured out studying images of the sky.

0:01:00.006,0:01:03.958
And to study a system[br]as complex as the entire universe,

0:01:03.982,0:01:08.687
astrophysicists are experts[br]at extracting simple models and solutions

0:01:08.711,0:01:11.050
from large and complex data sets.

0:01:11.705,0:01:14.130
So what else can I do with this expertise?

0:01:16.030,0:01:20.125
What if we turned the camera[br]around towards us?

0:01:21.057,0:01:24.048
At the Urban Observatory,[br]that is exactly what we are doing.

0:01:24.072,0:01:26.564
Greg Dobler, also an astrophysicist

0:01:26.588,0:01:27.755
and my husband,

0:01:27.779,0:01:31.739
created the first urban observatory[br]in New York University in 2013,

0:01:31.763,0:01:33.319
and I joined in 2015.

0:01:33.752,0:01:35.681
Here are some of the things that we do.

0:01:36.196,0:01:38.466
We take pictures of the city at night

0:01:38.490,0:01:41.079
and study city lights like stars.

0:01:41.514,0:01:43.526
By studying how light changes over time

0:01:43.550,0:01:45.624
and the color of astronomical lights,

0:01:45.648,0:01:48.461
I gain insight about the nature[br]of exploding stars.

0:01:48.934,0:01:51.204
By studying city lights the same way,

0:01:51.228,0:01:55.910
we can measure and predict how much energy[br]the city needs and consumes

0:01:55.934,0:01:57.781
and help build a resilient grid

0:01:57.805,0:02:01.104
that will support the needs[br]or growing urban environments.

0:02:02.283,0:02:05.640
In daytime images,[br]we capture plumes of pollution.

0:02:06.274,0:02:09.744
Seventy-five percent[br]of greenhouse gases in New York City

0:02:09.768,0:02:13.400
come from a building like this one,[br]burning oil for heat.

0:02:14.477,0:02:16.872
You can measure pollution[br]with air quality sensors.

0:02:16.896,0:02:20.738
But imagine putting a sensor[br]on each New York City building,

0:02:20.762,0:02:23.468
reading in data from a million monitors.

0:02:23.492,0:02:24.818
Imagine the cost.

0:02:26.048,0:02:29.476
With a team of NYU students,[br]we built a mathematical model,

0:02:29.500,0:02:32.889
a neural network that can detect[br]and track these plumes

0:02:32.913,0:02:34.603
over the New York City skyline.

0:02:34.627,0:02:36.109
We can classify them --

0:02:36.133,0:02:39.148
harmless steam plumes,[br]white and evanescent;

0:02:39.172,0:02:42.688
polluting smokestacks,[br]dark and persistent --

0:02:42.712,0:02:46.471
and provide policy makers[br]with a map of neighborhood pollution.

0:02:47.777,0:02:51.666
This cross-disciplinary project[br]created transformational solutions.

0:02:53.942,0:02:56.855
But the data analysis methodologies[br]we use in astrophysics

0:02:56.879,0:02:58.840
can be applied to all sorts of data,

0:02:58.864,0:03:00.014
not just images.

0:03:00.450,0:03:02.934
We were asked to help[br]a California district attorney

0:03:02.958,0:03:06.379
understand prosecutorial delays[br]in their jurisdiction.

0:03:06.839,0:03:09.506
There are people on probation[br]or sitting in jail,

0:03:09.530,0:03:12.141
awaiting for trial sometimes for years.

0:03:12.165,0:03:14.609
They wanted to know[br]what kind of cases dragged on,

0:03:14.633,0:03:17.757
and they had a massive data set[br]to explore to understand it,

0:03:17.781,0:03:19.193
but didn't have the expertise

0:03:19.217,0:03:21.765
or the instruments[br]in their office to do so.

0:03:21.789,0:03:23.535
And that's where we came in.

0:03:23.559,0:03:26.956
I worked with my colleague,[br]public policy professor Angela Hawken,

0:03:26.980,0:03:30.466
and our team first created[br]a visual dashboard

0:03:30.490,0:03:34.390
for DAs to see and better understand[br]the prosecution process.

0:03:34.997,0:03:37.926
But also, we ourselves[br]analyzed their data,

0:03:37.950,0:03:40.470
looking to see if the duration[br]of the process

0:03:40.494,0:03:43.684
suffered from social inequalities[br]in their jurisdiction.

0:03:44.367,0:03:45.909
We did so using methods

0:03:45.933,0:03:48.906
that I would use to classify[br]thousands of stellar explosions,

0:03:48.930,0:03:51.585
applied to thousands of court cases.

0:03:51.609,0:03:52.760
And in doing so,

0:03:52.784,0:03:55.617
we built a model that can be applied[br]to other jurisdictions

0:03:55.641,0:03:57.831
who are willing to explore their biases.

0:03:57.855,0:04:01.101
These collaborations between[br]domain experts and astrophysicists

0:04:01.125,0:04:03.180
created transformational solutions

0:04:03.204,0:04:05.604
to help improve people's quality of life.

0:04:07.426,0:04:08.910
But it is a two-way road.

0:04:08.934,0:04:11.477
I bring my astrophysics background[br]to urban science,

0:04:11.501,0:04:15.342
and I bring what I learn in urban science[br]back to astrophysics.

0:04:15.930,0:04:17.782
Light echoes:

0:04:18.461,0:04:23.028
the reflections of stellar explosions[br]onto interstellar dust.

0:04:24.046,0:04:29.831
In our images, these reflections appear[br]as white, evanescent, moving features,

0:04:29.855,0:04:31.005
just like plumes.

0:04:31.363,0:04:35.258
I am adapting the same models[br]that detect plumes in city images

0:04:35.282,0:04:38.210
to detect light echoes[br]in images of the sky.

0:04:40.290,0:04:43.583
By exploring the things[br]that interest and excite me,

0:04:43.607,0:04:45.403
reaching outside of my domain,

0:04:45.427,0:04:48.303
I did turn my restlessness into an asset.

0:04:49.031,0:04:54.078
We, you, all have a unique perspective[br]that can generate new insight

0:04:54.102,0:04:58.285
and lead to new, unexpected,[br]transformational solutions.

0:04:58.944,0:05:00.106
Thank you.

0:05:00.130,0:05:04.288
(Applause)