TED@BCG San Francisco
Dan Berkenstock: The world is one big dataset. Now, how to photograph it ...
Filmed:
Readability: 4.8
888,941 views
We're all familiar with satellite imagery, but what we might not know is that much of it is out of date. That's because satellites are big and expensive, so there aren't that many of them up in space. As he explains in this fascinating talk, Dan Berkenstock and his team came up with a different solution, designing a cheap, lightweight satellite with a radically new approach to photographing what's going on on Earth.
Dan Berkenstock - Satellite designer
Dan Berkenstock and his team at Skybox Imaging are rethinking how to take photographs from space. Full bio
Dan Berkenstock and his team at Skybox Imaging are rethinking how to take photographs from space. Full bio
Double-click the English transcript below to play the video.
00:12
Five years ago, I was a Ph.D. student
0
498
2599
00:15
living two lives.
1
3097
1644
00:16
In one, I used NASA supercomputers
2
4741
2316
00:19
to design next-generation spacecraft,
3
7057
2728
00:21
and in the other I was a data scientist
4
9785
2579
00:24
looking for potential smugglers
5
12364
2419
00:26
of sensitive nuclear technologies.
6
14783
3239
00:30
As a data scientist, I did a lot of analyses,
7
18022
2859
00:32
mostly of facilities,
8
20881
1467
00:34
industrial facilities around the world.
9
22348
2271
00:36
And I was always looking for a better canvas
10
24619
2688
00:39
to tie these all together.
11
27307
2066
00:41
And one day, I was thinking about how
12
29373
1866
00:43
all data has a location,
13
31239
2124
00:45
and I realized that the answer
14
33363
1733
00:47
had been staring me in the face.
15
35096
1888
00:48
Although I was a satellite engineer,
16
36984
2638
00:51
I hadn't thought about using satellite imagery
17
39622
3171
00:54
in my work.
18
42793
1566
00:56
Now, like most of us, I'd been online,
19
44359
1947
00:58
I'd see my house, so I thought,
20
46306
1782
01:00
I'll hop in there and I'll start looking up
21
48088
2298
01:02
some of these facilities.
22
50386
1417
01:03
And what I found really surprised me.
23
51803
1969
01:05
The pictures that I was finding
24
53772
1674
01:07
were years out of date,
25
55446
1901
01:09
and because of that,
26
57347
1068
01:10
it had relatively little relevance
27
58415
1671
01:12
to the work that I was doing today.
28
60086
2460
01:14
But I was intrigued.
29
62546
1461
01:16
I mean, satellite imagery is pretty amazing stuff.
30
64007
3236
01:19
There are millions and millions of sensors
31
67243
2303
01:21
surrounding us today,
32
69546
1444
01:22
but there's still so much we
don't know on a daily basis.
don't know on a daily basis.
33
70990
2960
01:25
How much oil is stored in all of China?
34
73950
3753
01:29
How much corn is being produced?
35
77703
2933
01:32
How many ships are in all of our world's ports?
36
80636
4071
01:36
Now, in theory, all of these questions
37
84707
2559
01:39
could be answered by imagery,
38
87266
2244
01:41
but not if it's old.
39
89510
1727
01:43
And if this data was so valuable,
40
91237
2458
01:45
then how come I couldn't get my hands
41
93695
1932
01:47
on more recent pictures?
42
95627
2550
01:50
So the story begins over 50 years ago
43
98177
3135
01:53
with the launch of the first generation
44
101312
1871
01:55
of U.S. government photo reconnaissance satellites.
45
103183
3600
01:58
And today, there's a handful
46
106783
1696
02:00
of the great, great grandchildren
47
108479
2195
02:02
of these early Cold War machines
48
110674
1665
02:04
which are now operated by private companies
49
112339
2009
02:06
and from which the vast majority of satellite imagery
50
114348
2709
02:09
that you and I see on a daily basis comes.
51
117057
2825
02:11
During this period, launching things into space,
52
119882
2914
02:14
just the rocket to get the satellite up there,
53
122796
2677
02:17
has cost hundreds of millions of dollars each,
54
125473
4601
02:22
and that's created tremendous pressure
55
130074
1752
02:23
to launch things infrequently
56
131826
2289
02:26
and to make sure that when you do,
57
134115
1344
02:27
you cram as much functionality in there as possible.
58
135459
3639
02:31
All of this has only made satellites
59
139098
1611
02:32
bigger and bigger and bigger
60
140709
2417
02:35
and more expensive,
61
143126
1598
02:36
now nearly a billion, with a b, dollars per copy.
62
144724
5037
02:41
Because they are so expensive,
63
149761
1699
02:43
there aren't very many of them.
64
151460
1467
02:44
Because there aren't very many of them,
65
152927
1345
02:46
the pictures that we see on a daily basis
66
154272
2456
02:48
tend to be old.
67
156728
1504
02:50
I think a lot of people actually
understand this anecdotally,
understand this anecdotally,
68
158232
3282
02:53
but in order to visualize just how sparsely
69
161514
2539
02:56
our planet is collected,
70
164053
1884
02:57
some friends and I put together a dataset
71
165937
2066
03:00
of the 30 million pictures that have been gathered
72
168003
2587
03:02
by these satellites between 2000 and 2010.
73
170590
3484
03:06
As you can see in blue, huge areas of our world
74
174074
2755
03:08
are barely seen, less than once a year,
75
176829
2790
03:11
and even the areas that are seen most frequently,
76
179619
2272
03:13
those in red, are seen at best once a quarter.
77
181891
3748
03:17
Now as aerospace engineering grad students,
78
185639
2916
03:20
this chart cried out to us as a challenge.
79
188555
3428
03:23
Why do these things have to be so expensive?
80
191983
3455
03:27
Does a single satellite really have to cost
81
195438
2735
03:30
the equivalent of three 747 jumbo jets?
82
198173
4763
03:34
Wasn't there a way to build a smaller,
83
202936
2163
03:37
simpler, new satellite design that could enable
84
205099
3172
03:40
more timely imaging?
85
208271
2018
03:42
I realize that it does sound a little bit crazy
86
210289
3451
03:45
that we were going to go out and just
87
213740
1511
03:47
begin designing satellites,
88
215251
1847
03:49
but fortunately we had help.
89
217098
1963
03:51
In the late 1990s, a couple of professors
90
219061
2935
03:53
proposed a concept for radically reducing the price
91
221996
3481
03:57
of putting things in space.
92
225477
1772
03:59
This was hitchhiking small satellites
93
227249
2111
04:01
alongside much larger satellites.
94
229360
3016
04:04
This dropped the cost of putting objects up there
95
232376
2909
04:07
by over a factor of 100,
96
235285
1998
04:09
and suddenly we could afford to experiment,
97
237283
3086
04:12
to take a little bit of risk,
98
240369
1468
04:13
and to realize a lot of innovation.
99
241837
2266
04:16
And a new generation of engineers and scientists,
100
244103
3089
04:19
mostly out of universities,
101
247192
1599
04:20
began launching these very small,
102
248791
2343
04:23
breadbox-sized satellites called CubeSats.
103
251134
2591
04:25
And these were built with electronics obtained
104
253725
2403
04:28
from RadioShack instead of Lockheed Martin.
105
256128
3898
04:32
Now it was using the lessons
learned from these early missions
learned from these early missions
106
260026
2840
04:34
that my friends and I began a series of sketches
107
262866
2695
04:37
of our own satellite design.
108
265561
1689
04:39
And I can't remember a specific day
109
267250
2784
04:42
where we made a conscious decision
110
270034
1860
04:43
that we were actually going to
go out and build these things,
go out and build these things,
111
271894
2635
04:46
but once we got that idea in our minds
112
274529
2430
04:48
of the world as a dataset,
113
276959
2161
04:51
of being able to capture millions of data points
114
279120
2490
04:53
on a daily basis describing the global economy,
115
281610
3053
04:56
of being able to unearth billions of connections
116
284663
2520
04:59
between them that had never before been found,
117
287183
3165
05:02
it just seemed boring
118
290348
1648
05:03
to go work on anything else.
119
291996
2907
05:06
And so we moved into a cramped,
120
294903
2973
05:09
windowless office in Palo Alto,
121
297876
2979
05:12
and began working to take our design
122
300855
1967
05:14
from the drawing board into the lab.
123
302822
2938
05:17
The first major question we had to tackle
124
305760
2658
05:20
was just how big to build this thing.
125
308418
2334
05:22
In space, size drives cost,
126
310752
3183
05:25
and we had worked with these very small,
127
313935
2013
05:27
breadbox-sized satellites in school,
128
315948
2094
05:30
but as we began to better
understand the laws of physics,
understand the laws of physics,
129
318042
2551
05:32
we found that the quality of pictures
130
320593
2122
05:34
those satellites could take was very limited,
131
322715
3084
05:37
because the laws of physics dictate
132
325799
1704
05:39
that the best picture you
can take through a telescope
can take through a telescope
133
327503
3106
05:42
is a function of the diameter of that telescope,
134
330609
2239
05:44
and these satellites had a very small,
135
332848
1662
05:46
very constrained volume.
136
334510
1706
05:48
And we found that the best picture we would
137
336216
1795
05:50
have been able to get looked something like this.
138
338011
2485
05:52
Although this was the low-cost option,
139
340496
2245
05:54
quite frankly it was just too blurry
140
342741
1673
05:56
to see the things that make
satellite imagery valuable.
satellite imagery valuable.
141
344414
3569
05:59
So about three or four weeks later,
142
347983
2146
06:02
we met a group of engineers randomly
143
350129
2817
06:04
who had worked on the first
144
352946
1641
06:06
private imaging satellite ever developed,
145
354587
2441
06:09
and they told us that back in the 1970s,
146
357028
2118
06:11
the U.S. government had found a powerful
147
359146
2182
06:13
optimal tradeoff --
148
361328
1643
06:14
that in taking pictures at right
about one meter resolution,
about one meter resolution,
149
362971
3088
06:18
being able to see objects one meter in size,
150
366059
2884
06:20
they had found that they could not
just get very high-quality images,
just get very high-quality images,
151
368943
2691
06:23
but get a lot of them at an affordable price.
152
371634
2893
06:26
From our own computer simulations,
153
374527
1976
06:28
we quickly found that one meter really was
154
376503
2005
06:30
the minimum viable product
155
378508
1764
06:32
to be able to see the drivers of our global economy,
156
380272
3079
06:35
for the first time, being able to count
157
383351
1561
06:36
the ships and cars and shipping
containers and trucks
containers and trucks
158
384912
2721
06:39
that move around our world on a daily basis,
159
387633
2824
06:42
while conveniently still not
being able to see individuals.
being able to see individuals.
160
390457
3804
06:46
We had found our compromise.
161
394261
1621
06:47
We would have to build something larger
162
395882
1604
06:49
than the original breadbox,
163
397486
1563
06:51
now more like a mini-fridge,
164
399049
1651
06:52
but we still wouldn't have to build a pickup truck.
165
400700
2761
06:55
So now we had our constraint.
166
403461
2624
06:58
The laws of physics dictated
167
406085
1757
06:59
the absolute minimum-sized
telescope that we could build.
telescope that we could build.
168
407842
3324
07:03
What came next was making the rest of the satellite
169
411166
3054
07:06
as small and as simple as possible,
170
414220
1741
07:07
basically a flying telescope with four walls
171
415961
2652
07:10
and a set of electronics smaller than a phone book
172
418613
2941
07:13
that used less power than a 100 watt lightbulb.
173
421554
3438
07:16
The big challenge became actually taking
174
424992
1900
07:18
the pictures through that telescope.
175
426892
2198
07:21
Traditional imaging satellites use a line scanner,
176
429090
3044
07:24
similar to a Xerox machine,
177
432134
1761
07:25
and as they traverse the Earth, they take pictures,
178
433895
2400
07:28
scanning row by row by row
179
436295
2196
07:30
to build the complete image.
180
438491
1675
07:32
Now people use these because they get a lot of light,
181
440166
3004
07:35
which means less of the noise you see
182
443170
1831
07:37
in a low-cost cell phone image.
183
445016
2736
07:39
The problem with them is they require
184
447752
2361
07:42
very sophisticated pointing.
185
450113
2240
07:44
You have to stay focused on a 50-centimeter target
186
452353
2546
07:46
from over 600 miles away
187
454899
1752
07:48
while moving at more than
seven kilometers a second,
seven kilometers a second,
188
456651
2232
07:50
which requires an awesome degree of complexity.
189
458883
2749
07:53
So instead, we turned to a new
generation of video sensors,
generation of video sensors,
190
461632
3437
07:57
originally created for use in night vision goggles.
191
465069
3091
08:00
Instead of taking a single, high quality image,
192
468160
2899
08:03
we could take a videostream
193
471059
1387
08:04
of individually noisier frames,
194
472446
2923
08:07
but then we could recombine
195
475369
1744
08:09
all of those frames together
196
477113
1466
08:10
into very high-quality images
197
478579
1996
08:12
using sophisticated pixel processing techniques
198
480575
2593
08:15
here on the ground,
199
483168
1521
08:16
at a cost of one one hundredth a traditional system.
200
484689
3137
08:19
And we applied this technique
201
487826
1153
08:20
to many of the other systems on the satellite as well,
202
488979
2974
08:23
and day by day, our design evolved
203
491953
2840
08:26
from CAD to prototypes
204
494793
3489
08:30
to production units.
205
498282
2888
08:33
A few short weeks ago,
206
501170
1657
08:34
we packed up SkySat 1,
207
502827
1851
08:36
put our signatures on it,
208
504678
1773
08:38
and waved goodbye for the last time on Earth.
209
506451
2223
08:40
Today, it's sitting in its final launch configuration
210
508674
3481
08:44
ready to blast off in a few short weeks.
211
512155
3061
08:47
And soon, we'll turn our attention to launching
212
515216
2342
08:49
a constellation of 24 or more of these satellites
213
517558
3031
08:52
and beginning to build the scalable analytics
214
520589
2550
08:55
that will allow us to unearth the insights
215
523139
2339
08:57
in the petabytes of data we will collect.
216
525478
3189
09:00
So why do all of this? Why build these satellites?
217
528667
3973
09:04
Well, it turns out imaging satellites
218
532640
2611
09:07
have a unique ability to provide global transparency,
219
535251
3710
09:10
and providing that transparency on a timely basis
220
538961
2626
09:13
is simply an idea whose time has come.
221
541587
3411
09:16
We see ourselves as pioneers of a new frontier,
222
544998
3523
09:20
and beyond economic data,
223
548521
1764
09:22
unlocking the human story, moment by moment.
224
550285
3601
09:25
For a data scientist
225
553886
1563
09:27
that just happened to go to space camp as a kid,
226
555449
2750
09:30
it just doesn't get much better than that.
227
558199
2715
09:32
Thank you.
228
560914
2215
09:35
(Applause)
229
563129
4134
ABOUT THE SPEAKER
Dan Berkenstock - Satellite designerDan Berkenstock and his team at Skybox Imaging are rethinking how to take photographs from space.
Why you should listen
Dan Berkenstock is an entrepreneur and engineer from Chicago, who fell into a classic tale of Silicon Valley innovation while taking a graduate entrepreneurship course at Stanford. That class led him and some others to found Skybox Imaging, of which Berkenstock is now executive vice president and chief product officer.
Skybox's mission is simple, if bold: they're working to design and launch small satellites that "hitchhike" to space in an effort to revolutionize the satellite imaging business. In 2013, SkySat-1, the first such satellite, was launched and is now beaming back images that are high-enough resolution to show the real-time state of global commerce. The idea: to "revolutionize the ways that consumers, businesses, and governments make decisions in their day-to-day lives."
In a previous life, Berkenstock worked in the Advanced Supercomputing Division at NASA's Ames Research Center, and also worked as a counterproliferation analyst at Lawrence Livermore National Laboratory, where he focused on trying to find and thwart potential potential smugglers of nuclear technologies. He is currently on leave from the Ph.D. program in aeronautics and astronautics at Stanford University.
Dan Berkenstock | Speaker | TED.com