TED2018
Will Marshall: The mission to create a searchable database of Earth's surface
Filmed:
Readability: 4.5
1,655,705 views
What if you could search the surface of the Earth the same way you search the internet? Will Marshall and his team at Planet use the world's largest fleet of satellites to image the entire Earth every day. Now they're moving on to a new project: using AI to index all the objects on the planet over time -- which could make ships, trees, houses and everything else on Earth searchable, the same way you search Google. He shares a vision for how this database can become a living record of the immense physical changes happening across the globe. "You can't fix what you can't see," Marshall says. "We want to give people the tools to see change and take action."
Will Marshall - Space scientist
At Planet, Will Marshall leads overall strategy for commercializing new geospatial data and analytics that are disrupting agriculture, mapping, energy, the environment and other vertical markets. Full bio
At Planet, Will Marshall leads overall strategy for commercializing new geospatial data and analytics that are disrupting agriculture, mapping, energy, the environment and other vertical markets. Full bio
Double-click the English transcript below to play the video.
00:12
Four years ago, here at TED,
0
761
2215
00:15
I announced Planet's Mission 1:
1
3000
2336
00:17
to launch a fleet of satellites
2
5360
1856
00:19
that would image
the entire Earth, every day,
the entire Earth, every day,
3
7240
2280
00:22
and to democratize access to it.
4
10560
1640
00:25
The problem we were trying
to solve was simple.
to solve was simple.
5
13520
2216
00:27
Satellite imagery you find online is old,
typically years old,
typically years old,
6
15760
3096
00:30
yet human activity was happening
on days and weeks and months,
on days and weeks and months,
7
18880
3936
00:34
and you can't fix what you can't see.
8
22840
2256
00:37
We wanted to give people the tools
to see that change and take action.
to see that change and take action.
9
25120
3776
00:40
The beautiful Blue Marble image,
taken by the Apollo 17 astronauts in 1972
taken by the Apollo 17 astronauts in 1972
10
28920
4936
00:45
had helped humanity become aware
that we're on a fragile planet.
that we're on a fragile planet.
11
33880
3280
00:49
And we wanted to take it
to the next level,
to the next level,
12
37600
2056
00:51
to give people the tools
to take action, to take care of it.
to take action, to take care of it.
13
39680
3456
00:55
Well, after many
Apollo projects of our own,
Apollo projects of our own,
14
43160
4056
00:59
launching the largest fleet
of satellites in human history,
of satellites in human history,
15
47240
2960
01:03
we have reached our target.
16
51600
1520
01:06
Today, Planet images
the entire Earth, every single day.
the entire Earth, every single day.
17
54080
3656
01:09
Mission accomplished.
18
57760
1216
01:11
(Applause)
19
59000
2536
01:13
Thank you.
20
61560
1200
01:15
It's taken 21 rocket launches --
21
63600
3976
01:19
this animation makes it look
really simple -- it was not.
really simple -- it was not.
22
67600
4160
01:25
And we now have
over 200 satellites in orbit,
over 200 satellites in orbit,
23
73040
3456
01:28
downlinking their data to 31 ground
stations we built around the planet.
stations we built around the planet.
24
76520
3936
01:32
In total, we get 1.5 million 29-megapixel
images of the Earth down each day.
images of the Earth down each day.
25
80480
6296
01:38
And on any one location
of the Earth's surface,
of the Earth's surface,
26
86800
2416
01:41
we now have on average
more than 500 images.
more than 500 images.
27
89240
3496
01:44
A deep stack of data,
documenting immense change.
documenting immense change.
28
92760
3880
01:49
And lots of people are using this imagery.
29
97320
2536
01:51
Agricultural companies are using it
to improve farmers' crop yields.
to improve farmers' crop yields.
30
99880
5136
01:57
Consumer-mapping companies are using it
to improve the maps you find online.
to improve the maps you find online.
31
105040
4176
02:01
Governments are using it
for border security
for border security
32
109240
2096
02:03
or helping with disaster response
after floods and fires and earthquakes.
after floods and fires and earthquakes.
33
111360
3680
02:08
And lots of NGOs are using it.
34
116320
1536
02:09
So, for tracking
and stopping deforestation.
and stopping deforestation.
35
117880
3416
02:13
Or helping to find the refugees
fleeing Myanmar.
fleeing Myanmar.
36
121320
3536
02:16
Or to track all the activities
in the ongoing crisis in Syria,
in the ongoing crisis in Syria,
37
124880
4376
02:21
holding all sides accountable.
38
129280
1680
02:24
And today, I'm pleased
to announce Planet stories.
to announce Planet stories.
39
132640
3456
02:28
Anyone can go online to planet.com
40
136120
2296
02:30
open an account and see
all of our imagery online.
all of our imagery online.
41
138440
3240
02:34
It's a bit like Google Earth,
except it's up-to-date imagery,
except it's up-to-date imagery,
42
142480
3096
02:37
and you can see back through time.
43
145600
2680
02:41
You can compare any two days
44
149040
1696
02:42
and see the dramatic changes
that happen around our planet.
that happen around our planet.
45
150760
2880
02:46
Or you can create a time lapse
through the 500 images that we have
through the 500 images that we have
46
154560
3400
02:50
and see that change
dramatically over time.
dramatically over time.
47
158600
2560
02:54
And you can share these over social media.
48
162000
2720
02:57
It's pretty cool.
49
165520
1216
02:58
(Applause)
50
166760
1216
03:00
Thank you.
51
168000
1200
03:02
We initially created this tool
for news journalists,
for news journalists,
52
170440
2456
03:04
who wanted to get unbiased information
about world events.
about world events.
53
172920
2736
03:07
But now we've opened it up
for anyone to use,
for anyone to use,
54
175680
2216
03:09
for nonprofit or personal uses.
55
177920
2000
03:12
And we hope it will give people the tools
to find and see the changes on the planet
to find and see the changes on the planet
56
180600
4416
03:17
and take action.
57
185040
1200
03:18
OK, so humanity now has this database
of information about the planet,
of information about the planet,
58
186920
4256
03:23
changing over time.
59
191200
1216
03:24
What's our next mission, what's Mission 2?
60
192440
2056
03:26
In short, it's space plus AI.
61
194520
2440
03:29
What we're doing
with artificial intelligence
with artificial intelligence
62
197720
2176
03:31
is finding the objects
in all the satellite images.
in all the satellite images.
63
199920
3096
03:35
The same AI tools that are used
to find cats in videos online
to find cats in videos online
64
203040
4536
03:39
can also be used to find
information on our pictures.
information on our pictures.
65
207600
3896
03:43
So, imagine if you can say,
this is a ship, this is a tree,
this is a ship, this is a tree,
66
211520
3336
03:46
this is a car, this is a road,
this is a building, this is a truck.
this is a building, this is a truck.
67
214880
4376
03:51
And if you could do that
for all of the millions of images
for all of the millions of images
68
219280
2736
03:54
coming down per day,
69
222040
1256
03:55
then you basically create a database
70
223320
1736
03:57
of all the sizable objects
on the planet, every day.
on the planet, every day.
71
225080
2656
03:59
And that database is searchable.
72
227760
1560
04:02
So that's exactly what we're doing.
73
230520
2096
04:04
Here's a prototype, working on our API.
74
232640
2256
04:06
This is Beijing.
75
234920
1456
04:08
So, imagine if we wanted
to count the planes in the airport.
to count the planes in the airport.
76
236400
2856
04:11
We select the airport,
77
239280
1856
04:13
and it finds the planes in today's image,
78
241160
2376
04:15
and finds the planes
in the whole stack of images before it,
in the whole stack of images before it,
79
243560
3256
04:18
and then outputs this graph of all
the planes in Beijing airport over time.
the planes in Beijing airport over time.
80
246840
4896
04:23
Of course, you could do this
for all the airports around the world.
for all the airports around the world.
81
251760
3576
04:27
And let's look here
in the port of Vancouver.
in the port of Vancouver.
82
255360
2936
04:30
So, we would do the same,
but this time we would look for vessels.
but this time we would look for vessels.
83
258320
3536
04:33
So, we zoom in on Vancouver,
we select the area,
we select the area,
84
261880
4136
04:38
and we search for ships.
85
266040
2056
04:40
And it outputs where all the ships are.
86
268120
1858
04:42
Now, imagine how useful this would be
to people in coast guards
to people in coast guards
87
270002
3214
04:45
who are trying to track
and stop illegal fishing.
and stop illegal fishing.
88
273240
2736
04:48
See, legal fishing vessels
89
276000
2056
04:50
transmit their locations
using AIS beacons.
using AIS beacons.
90
278080
2936
04:53
But we frequently find ships
that are not doing that.
that are not doing that.
91
281040
3616
04:56
The pictures don't lie.
92
284680
1776
04:58
And so, coast guards could use that
93
286480
1696
05:00
and go and find
those illegal fishing vessels.
those illegal fishing vessels.
94
288200
2176
05:02
And soon we'll add
not just ships and planes
not just ships and planes
95
290400
2176
05:04
but all the other objects,
96
292600
1296
05:05
and we can output data feeds
97
293920
1896
05:07
of those locations
of all these objects over time
of all these objects over time
98
295840
2536
05:10
that can be integrated digitally
from people's work flows.
from people's work flows.
99
298400
3056
05:13
In time, we could get
more sophisticated browsers
more sophisticated browsers
100
301480
3056
05:16
that people pull in
from different sources.
from different sources.
101
304560
2336
05:18
But ultimately, I can imagine us
abstracting out the imagery entirely
abstracting out the imagery entirely
102
306920
4696
05:23
and just having a queryable
interface to the Earth.
interface to the Earth.
103
311640
2416
05:26
Imagine if we could just ask,
104
314080
1416
05:27
"Hey, how many houses
are there in Pakistan?
are there in Pakistan?
105
315520
2536
05:30
Give me a plot of that versus time."
106
318080
1936
05:32
"How many trees are there in the Amazon
107
320040
2176
05:34
and can you tell me the locations
of the trees that have been felled
of the trees that have been felled
108
322240
3216
05:37
between this week and last week?"
109
325480
1656
05:39
Wouldn't that be great?
110
327160
1216
05:40
Well, that's what
we're trying to go towards,
we're trying to go towards,
111
328400
2136
05:42
and we call it "Queryable Earth."
112
330560
1856
05:44
So, Planet's Mission 1 was
to image the whole planet every day
to image the whole planet every day
113
332440
3896
05:48
and make it accessible.
114
336360
2336
05:50
Planet's Mission 2 is to index
all the objects on the planet over time
all the objects on the planet over time
115
338720
3816
05:54
and make it queryable.
116
342560
1240
05:56
Let me leave you with an analogy.
117
344760
2136
05:58
Google indexed what's on the internet
and made it searchable.
and made it searchable.
118
346920
3400
06:03
Well, we're indexing what's on the Earth
and making it searchable.
and making it searchable.
119
351080
3256
06:06
Thank you very much.
120
354360
1336
06:07
(Applause)
121
355720
4520
ABOUT THE SPEAKER
Will Marshall - Space scientistAt Planet, Will Marshall leads overall strategy for commercializing new geospatial data and analytics that are disrupting agriculture, mapping, energy, the environment and other vertical markets.
Why you should listen
Will Marshall is the co-founder and CEO of Planet. Prior to Planet, he was a Scientist at NASA/USRA where he worked on missions "LADEE" and "LCROSS," served as co-principal investigator on PhoneSat, and was the technical lead on research projects in space debris remediation.
Marshall received his PhD in Physics from the University of Oxford and his Masters in Physics with Space Science and Technology from the University of Leicester. He was a Postdoctoral Fellow at George Washington University and Harvard.
Will Marshall | Speaker | TED.com