TED2011
Carlo Ratti: Architecture that senses and responds
Filmed:
Readability: 3.9
746,791 views
With his team at SENSEable City Lab, MIT's Carlo Ratti makes cool things by sensing the data we create. He pulls from passive data sets -- like the calls we make, the garbage we throw away -- to create surprising visualizations of city life. And he and his team create dazzling interactive environments from moving water and flying light, powered by simple gestures caught through sensors.
Carlo Ratti - Architect and engineer
Carlo Ratti directs the MIT SENSEable City Lab, which explores the "real-time city" by studying the way sensors and electronics relate to the built environment. Full bio
Carlo Ratti directs the MIT SENSEable City Lab, which explores the "real-time city" by studying the way sensors and electronics relate to the built environment. Full bio
Double-click the English transcript below to play the video.
00:15
Good afternoon, everybody.
0
0
2000
00:17
I've got something to show you.
1
2000
3000
00:37
(Laughter)
2
22000
2000
00:39
Think about this as a pixel, a flying pixel.
3
24000
3000
00:42
This is what we call, in our lab, sensible design.
4
27000
3000
00:45
Let me tell you a bit about it.
5
30000
2000
00:47
Now if you take this picture -- I'm Italian originally,
6
32000
3000
00:50
and every boy in Italy grows up
7
35000
2000
00:52
with this picture on the wall of his bedroom --
8
37000
2000
00:54
but the reason I'm showing you this
9
39000
2000
00:56
is that something very interesting
10
41000
2000
00:58
happened in Formula 1 racing
11
43000
2000
01:00
over the past couple of decades.
12
45000
2000
01:02
Now some time ago,
13
47000
2000
01:04
if you wanted to win a Formula 1 race,
14
49000
2000
01:06
you take a budget, and you bet your budget
15
51000
2000
01:08
on a good driver and a good car.
16
53000
3000
01:11
And if the car and the driver were good enough, then you'd win the race.
17
56000
3000
01:14
Now today, if you want to win the race,
18
59000
2000
01:16
actually you need also something like this --
19
61000
3000
01:19
something that monitors the car in real time,
20
64000
3000
01:22
has a few thousand sensors
21
67000
2000
01:24
collecting information from the car,
22
69000
2000
01:26
transmitting this information into the system,
23
71000
3000
01:29
and then processing it
24
74000
2000
01:31
and using it in order to go back to the car with decisions
25
76000
3000
01:34
and changing things in real time
26
79000
2000
01:36
as information is collected.
27
81000
2000
01:38
This is what, in engineering terms,
28
83000
2000
01:40
you would call a real time control system.
29
85000
3000
01:43
And basically, it's a system made of two components --
30
88000
3000
01:46
a sensing and an actuating component.
31
91000
2000
01:48
What is interesting today
32
93000
2000
01:50
is that real time control systems
33
95000
2000
01:52
are starting to enter into our lives.
34
97000
3000
01:55
Our cities, over the past few years,
35
100000
3000
01:58
just have been blanketed
36
103000
2000
02:00
with networks, electronics.
37
105000
2000
02:02
They're becoming like computers in open air.
38
107000
2000
02:04
And, as computers in open air,
39
109000
2000
02:06
they're starting to respond in a different way
40
111000
2000
02:08
to be able to be sensed and to be actuated.
41
113000
3000
02:11
If we fix cities, actually it's a big deal.
42
116000
2000
02:13
Just as an aside, I wanted to mention,
43
118000
2000
02:15
cities are only two percent of the Earth's crust,
44
120000
4000
02:19
but they are 50 percent of the world's population.
45
124000
3000
02:22
They are 75 percent of the energy consumption --
46
127000
3000
02:25
up to 80 percent of CO2 emissions.
47
130000
3000
02:28
So if we're able to do something with cities, that's a big deal.
48
133000
3000
02:31
Beyond cities,
49
136000
2000
02:33
all of this sensing and actuating
50
138000
3000
02:36
is entering our everyday objects.
51
141000
2000
02:38
That's from an exhibition that
52
143000
2000
02:40
Paola Antonelli is organizing
53
145000
2000
02:42
at MoMA later this year, during the summer.
54
147000
2000
02:44
It's called "Talk to Me."
55
149000
2000
02:46
Well our objects, our environment
56
151000
2000
02:48
is starting to talk back to us.
57
153000
2000
02:50
In a certain sense, it's almost as if every atom out there
58
155000
3000
02:53
were becoming both a sensor and an actuator.
59
158000
3000
02:56
And that is radically changing the interaction we have as humans
60
161000
3000
02:59
with the environment out there.
61
164000
2000
03:01
In a certain sense,
62
166000
2000
03:03
it's almost as if the old dream of Michelangelo ...
63
168000
3000
03:06
you know, when Michelangelo sculpted the Moses,
64
171000
2000
03:08
at the end it said that he took the hammer, threw it at the Moses --
65
173000
3000
03:11
actually you can still see a small chip underneath --
66
176000
3000
03:14
and said, shouted,
67
179000
2000
03:16
"Perché non parli? Why don't you talk?"
68
181000
2000
03:18
Well today, for the first time,
69
183000
2000
03:20
our environment is starting to talk back to us.
70
185000
3000
03:23
And I'll show just a few examples --
71
188000
2000
03:25
again, with this idea of sensing our environment and actuating it.
72
190000
3000
03:28
Let's starting with sensing.
73
193000
3000
03:31
Well, the first project I wanted to share with you
74
196000
2000
03:33
is actually one of the first projects by our lab.
75
198000
3000
03:36
It was four and a half years ago in Italy.
76
201000
3000
03:39
And what we did there
77
204000
2000
03:41
was actually use a new type of network at the time
78
206000
2000
03:43
that had been deployed all across the world --
79
208000
2000
03:45
that's a cellphone network --
80
210000
2000
03:47
and use anonymous and aggregated information from that network,
81
212000
2000
03:49
that's collected anyway by the operator,
82
214000
2000
03:51
in order to understand
83
216000
2000
03:53
how the city works.
84
218000
2000
03:55
The summer was a lucky summer -- 2006.
85
220000
3000
03:58
It's when Italy won the soccer World Cup.
86
223000
3000
04:01
Some of you might remember, it was Italy and France playing,
87
226000
3000
04:04
and then Zidane at the end, the headbutt.
88
229000
2000
04:06
And anyway, Italy won at the end.
89
231000
2000
04:08
(Laughter)
90
233000
2000
04:10
Now look at what happened that day
91
235000
2000
04:12
just by monitoring activity
92
237000
2000
04:14
happening on the network.
93
239000
2000
04:16
Here you see the city.
94
241000
2000
04:18
You see the Colosseum in the middle,
95
243000
3000
04:21
the river Tiber.
96
246000
3000
04:24
It's morning, before the match.
97
249000
2000
04:26
You see the timeline on the top.
98
251000
2000
04:28
Early afternoon,
99
253000
2000
04:30
people here and there,
100
255000
2000
04:32
making calls and moving.
101
257000
2000
04:34
The match begins -- silence.
102
259000
3000
04:37
France scores. Italy scores.
103
262000
3000
04:40
Halftime, people make a quick call and go to the bathroom.
104
265000
4000
04:44
Second half. End of normal time.
105
269000
2000
04:46
First overtime, second.
106
271000
2000
04:48
Zidane, the headbutt in a moment.
107
273000
3000
04:51
Italy wins. Yeah.
108
276000
2000
04:53
(Laughter)
109
278000
2000
04:55
(Applause)
110
280000
3000
04:58
Well, that night, everybody went to celebrate in the center.
111
283000
2000
05:00
You saw the big peak.
112
285000
2000
05:02
The following day, again everybody went to the center
113
287000
2000
05:04
to meet the winning team
114
289000
3000
05:07
and the prime minister at the time.
115
292000
2000
05:09
And then everybody moved down.
116
294000
2000
05:11
You see the image of the place called Circo Massimo,
117
296000
2000
05:13
where, since Roman times, people go to celebrate,
118
298000
3000
05:16
to have a big party, and you see the peak at the end of the day.
119
301000
3000
05:19
Well, that's just one example of how we can sense the city today
120
304000
2000
05:21
in a way that we couldn't have done
121
306000
2000
05:23
just a few years ago.
122
308000
2000
05:25
Another quick example about sensing:
123
310000
2000
05:27
it's not about people,
124
312000
2000
05:29
but about things we use and consume.
125
314000
2000
05:31
Well today, we know everything
126
316000
2000
05:33
about where our objects come from.
127
318000
3000
05:36
This is a map that shows you
128
321000
2000
05:38
all the chips that form a Mac computer, how they came together.
129
323000
3000
05:41
But we know very little about where things go.
130
326000
3000
05:44
So in this project,
131
329000
2000
05:46
we actually developed some small tags
132
331000
2000
05:48
to track trash as it moves through the system.
133
333000
3000
05:51
So we actually started with a number of volunteers
134
336000
3000
05:54
who helped us in Seattle,
135
339000
2000
05:56
just over a year ago,
136
341000
2000
05:58
to tag what they were throwing away --
137
343000
3000
06:01
different types of things, as you can see here --
138
346000
3000
06:04
things they would throw away anyway.
139
349000
2000
06:06
Then we put a little chip, little tag,
140
351000
2000
06:08
onto the trash
141
353000
2000
06:10
and then started following it.
142
355000
2000
06:12
Here are the results we just obtained.
143
357000
3000
06:15
(Music)
144
360000
3000
06:18
From Seattle ...
145
363000
3000
06:26
after one week.
146
371000
2000
06:53
With this information we realized
147
398000
2000
06:55
there's a lot of inefficiencies in the system.
148
400000
2000
06:57
We can actually do the same thing with much less energy.
149
402000
3000
07:00
This data was not available before.
150
405000
2000
07:02
But there's a lot of wasted transportation and convoluted things happening.
151
407000
3000
07:05
But the other thing is that we believe
152
410000
2000
07:07
that if we see every day
153
412000
2000
07:09
that the cup we're throwing away, it doesn't disappear,
154
414000
2000
07:11
it's still somewhere on the planet.
155
416000
2000
07:13
And the plastic bottle we're throwing away every day still stays there.
156
418000
3000
07:16
And if we show that to people,
157
421000
2000
07:18
then we can also promote some behavioral change.
158
423000
2000
07:20
So that was the reason for the project.
159
425000
2000
07:22
My colleague at MIT, Assaf Biderman,
160
427000
2000
07:24
he could tell you much more about sensing
161
429000
2000
07:26
and many other wonderful things we can do with sensing,
162
431000
2000
07:28
but I wanted to go to the second part we discussed at the beginning,
163
433000
3000
07:31
and that's actuating our environment.
164
436000
2000
07:33
And the first project
165
438000
2000
07:35
is something we did a couple of years ago in Zaragoza, Spain.
166
440000
3000
07:38
It started with a question by the mayor of the city,
167
443000
3000
07:41
who came to us saying
168
446000
2000
07:43
that Spain and Southern Europe have a beautiful tradition
169
448000
3000
07:46
of using water in public space, in architecture.
170
451000
3000
07:49
And the question was: How could technology, new technology,
171
454000
2000
07:51
be added to that?
172
456000
2000
07:53
And one of the ideas that was developed at MIT in a workshop
173
458000
3000
07:56
was, imagine this pipe, and you've got valves,
174
461000
3000
07:59
solenoid valves, taps,
175
464000
2000
08:01
opening and closing.
176
466000
2000
08:03
You create like a water curtain with pixels made of water.
177
468000
3000
08:06
If those pixels fall,
178
471000
2000
08:08
you can write on it,
179
473000
2000
08:10
you can show patterns, images, text.
180
475000
2000
08:12
And even you can approach it, and it will open up
181
477000
2000
08:14
to let you jump through,
182
479000
2000
08:16
as you see in this image.
183
481000
2000
08:18
Well, we presented this to Mayor Belloch.
184
483000
2000
08:20
He liked it very much.
185
485000
2000
08:22
And we got a commission to design a building
186
487000
2000
08:24
at the entrance of the expo.
187
489000
2000
08:26
We called it Digital Water Pavilion.
188
491000
2000
08:28
The whole building is made of water.
189
493000
3000
08:33
There's no doors or windows,
190
498000
2000
08:35
but when you approach it,
191
500000
2000
08:37
it will open up to let you in.
192
502000
2000
08:39
(Music)
193
504000
6000
08:52
The roof also is covered with water.
194
517000
3000
08:57
And if there's a bit of wind,
195
522000
2000
08:59
if you want to minimize splashing, you can actually lower the roof.
196
524000
3000
09:04
Or you could close the building,
197
529000
2000
09:06
and the whole architecture will disappear,
198
531000
2000
09:08
like in this case.
199
533000
2000
09:10
You know, these days, you always get images during the winter,
200
535000
2000
09:12
when they take the roof down,
201
537000
2000
09:14
of people who have been there and said, "They demolished the building."
202
539000
3000
09:17
No, they didn't demolish it, just when it goes down,
203
542000
2000
09:19
the architecture almost disappears.
204
544000
2000
09:21
Here's the building working.
205
546000
3000
09:24
You see the person puzzled about what was going on inside.
206
549000
3000
09:27
And here was myself trying not to get wet,
207
552000
2000
09:29
testing the sensors that open the water.
208
554000
3000
09:32
Well, I should tell you now what happened one night
209
557000
2000
09:34
when all of the sensors stopped working.
210
559000
3000
09:37
But actually that night, it was even more fun.
211
562000
3000
09:40
All the kids from Zaragoza came to the building,
212
565000
2000
09:42
because the way of engaging with the building became something different.
213
567000
3000
09:45
Not anymore a building that would open up to let you in,
214
570000
3000
09:48
but a building that would still make cuts and holes through the water,
215
573000
3000
09:51
and you had to jump without getting wet.
216
576000
2000
09:53
(Video) (Crowd Noise)
217
578000
13000
10:06
And that was, for us, was very interesting,
218
591000
2000
10:08
because, as architects, as engineers, as designers,
219
593000
3000
10:11
we always think about how people will use the things we design.
220
596000
3000
10:14
But then reality's always unpredictable.
221
599000
3000
10:17
And that's the beauty of doing things
222
602000
2000
10:19
that are used and interact with people.
223
604000
2000
10:21
Here is an image then of the building
224
606000
2000
10:23
with the physical pixels, the pixels made of water,
225
608000
2000
10:25
and then projections on them.
226
610000
3000
10:28
And this is what led us to think about
227
613000
2000
10:30
the following project I'll show you now.
228
615000
2000
10:32
That's, imagine those pixels could actually start flying.
229
617000
3000
10:35
Imagine you could have small helicopters
230
620000
2000
10:37
that move in the air,
231
622000
2000
10:39
and then each of them with a small pixel in changing lights --
232
624000
3000
10:42
almost as a cloud that can move in space.
233
627000
3000
10:45
Here is the video.
234
630000
2000
10:47
(Music)
235
632000
6000
10:53
So imagine one helicopter,
236
638000
3000
10:56
like the one we saw before,
237
641000
3000
11:01
moving with others,
238
646000
3000
11:04
in synchrony.
239
649000
2000
11:06
So you can have this cloud.
240
651000
3000
11:15
You can have a kind of flexible screen or display, like this --
241
660000
4000
11:19
a regular configuration in two dimensions.
242
664000
3000
11:29
Or in regular, but in three dimensions,
243
674000
3000
11:32
where the thing that changes is the light,
244
677000
2000
11:34
not the pixels' position.
245
679000
2000
11:46
You can play with a different type.
246
691000
2000
11:48
Imagine your screen could just appear
247
693000
2000
11:50
in different scales or sizes,
248
695000
3000
11:53
different types of resolution.
249
698000
3000
12:05
But then the whole thing can be
250
710000
2000
12:07
just a 3D cloud of pixels
251
712000
2000
12:09
that you can approach and move through it
252
714000
3000
12:12
and see from many, many directions.
253
717000
3000
12:15
Here is the real Flyfire
254
720000
2000
12:17
control and going down to form the regular grid as before.
255
722000
4000
12:21
When you turn on the light, actually you see this. So the same as we saw before.
256
726000
3000
12:24
And imagine each of them then controlled by people.
257
729000
2000
12:26
You can have each pixel
258
731000
2000
12:28
having an input that comes from people,
259
733000
2000
12:30
from people's movement, or so and so.
260
735000
2000
12:32
I want to show you something here for the first time.
261
737000
3000
12:35
We've been working with Roberto Bolle,
262
740000
2000
12:37
one of today's top ballet dancers --
263
742000
2000
12:39
the étoile at Metropolitan in New York
264
744000
2000
12:41
and La Scala in Milan --
265
746000
2000
12:43
and actually captured his movement in 3D
266
748000
2000
12:45
in order to use it as an input for Flyfire.
267
750000
3000
12:48
And here you can see Roberto dancing.
268
753000
3000
12:53
You see on the left the pixels,
269
758000
2000
12:55
the different resolutions being captured.
270
760000
2000
12:57
It's both 3D scanning in real time
271
762000
2000
12:59
and motion capture.
272
764000
3000
13:03
So you can reconstruct a whole movement.
273
768000
3000
13:10
You can go all the way through.
274
775000
3000
13:16
But then, once we have the pixels, then you can play with them
275
781000
2000
13:18
and play with color and movement
276
783000
3000
13:21
and gravity and rotation.
277
786000
3000
13:24
So we want to use this as one of the possible inputs
278
789000
2000
13:26
for Flyfire.
279
791000
2000
13:47
I wanted to show you the last project we are working on.
280
812000
2000
13:49
It's something we're working on for the London Olympics.
281
814000
2000
13:51
It's called The Cloud.
282
816000
2000
13:53
And the idea here is, imagine, again,
283
818000
2000
13:55
we can involve people
284
820000
2000
13:57
in doing something and changing our environment --
285
822000
3000
14:00
almost to impart what we call cloud raising --
286
825000
2000
14:02
like barn raising, but with a cloud.
287
827000
2000
14:04
Imagine you can have everybody make a small donation for one pixel.
288
829000
4000
14:08
And I think what is remarkable
289
833000
2000
14:10
that has happened over the past couple of years
290
835000
2000
14:12
is that, over the past couple of decades,
291
837000
2000
14:14
we went from the physical world to the digital one.
292
839000
3000
14:17
This has been digitizing everything, knowledge,
293
842000
2000
14:19
and making that accessible through the Internet.
294
844000
2000
14:21
Now today, for the first time --
295
846000
2000
14:23
and the Obama campaign showed us this --
296
848000
2000
14:25
we can go from the digital world,
297
850000
2000
14:27
from the self-organizing power of networks,
298
852000
2000
14:29
to the physical one.
299
854000
2000
14:31
This can be, in our case,
300
856000
2000
14:33
we want to use it for designing and doing a symbol.
301
858000
2000
14:35
That means something built in a city.
302
860000
2000
14:37
But tomorrow it can be,
303
862000
2000
14:39
in order to tackle today's pressing challenges --
304
864000
3000
14:42
think about climate change or CO2 emissions --
305
867000
2000
14:44
how we can go from the digital world to the physical one.
306
869000
3000
14:47
So the idea that we can actually involve people
307
872000
2000
14:49
in doing this thing together, collectively.
308
874000
2000
14:51
The cloud is a cloud, again, made of pixels,
309
876000
3000
14:54
in the same way as the real cloud
310
879000
2000
14:56
is a cloud made of particles.
311
881000
2000
14:58
And those particles are water,
312
883000
2000
15:00
where our cloud is a cloud of pixels.
313
885000
2000
15:02
It's a physical structure in London, but covered with pixels.
314
887000
3000
15:05
You can move inside, have different types of experiences.
315
890000
2000
15:07
You can actually see from underneath,
316
892000
2000
15:09
sharing the main moments
317
894000
2000
15:11
for the Olympics in 2012 and beyond,
318
896000
3000
15:14
and really using it as a way to connect with the community.
319
899000
4000
15:18
So both the physical cloud in the sky
320
903000
4000
15:22
and something you can go to the top [of],
321
907000
3000
15:25
like London's new mountaintop.
322
910000
2000
15:27
You can enter inside it.
323
912000
2000
15:29
And a kind of new digital beacon for the night --
324
914000
3000
15:32
but most importantly,
325
917000
2000
15:34
a new type of experience for anybody who will go to the top.
326
919000
3000
15:37
Thank you.
327
922000
2000
15:39
(Applause)
328
924000
2000
ABOUT THE SPEAKER
Carlo Ratti - Architect and engineerCarlo Ratti directs the MIT SENSEable City Lab, which explores the "real-time city" by studying the way sensors and electronics relate to the built environment.
Why you should listen
Carlo Ratti is a civil engineer and architect who teaches at the Massachusetts Institute of Technology, where he directs the SENSEable City Laboratory. This lab studies the built environment of cities -- from street grids to plumbing and garbage systems -- using new kinds of sensors and hand-held electronics that have transformed the way we can describe and understand cities.
Other projects flip this equation -- using data gathered from sensors to actually create dazzling new environments. The Digital Water Pavilion, for instance, reacts to visitors by parting a stream of water to let them visit. And a project for the 2012 Olympics in London turns a pavilion building into a cloud of blinking interactive art. He's opening a research center in Singapore as part of an MIT-led initiative on the Future of Urban Mobility.
For more information on the projects in this talk, visit SENSEable @ TED >>
Carlo Ratti | Speaker | TED.com