The Future of Flying Robots | Vijay Kumar | TED Talks

755,311 views ・ 2015-11-04

TED


Please double-click on the English subtitles below to play the video.

00:13
In my lab, we build autonomous aerial robots
0
13280
3656
00:16
like the one you see flying here.
1
16960
1880
00:20
Unlike the commercially available drones that you can buy today,
2
20720
3696
00:24
this robot doesn't have any GPS on board.
3
24440
2640
00:28
So without GPS,
4
28160
1216
00:29
it's hard for robots like this to determine their position.
5
29400
3280
00:34
This robot uses onboard sensors, cameras and laser scanners,
6
34240
4736
00:39
to scan the environment.
7
39000
1696
00:40
It detects features from the environment,
8
40720
3056
00:43
and it determines where it is relative to those features,
9
43800
2736
00:46
using a method of triangulation.
10
46560
2136
00:48
And then it can assemble all these features into a map,
11
48720
3456
00:52
like you see behind me.
12
52200
1736
00:53
And this map then allows the robot to understand where the obstacles are
13
53960
3936
00:57
and navigate in a collision-free manner.
14
57920
2720
01:01
What I want to show you next
15
61160
2096
01:03
is a set of experiments we did inside our laboratory,
16
63280
3216
01:06
where this robot was able to go for longer distances.
17
66520
3480
01:10
So here you'll see, on the top right, what the robot sees with the camera.
18
70400
5016
01:15
And on the main screen --
19
75440
1216
01:16
and of course this is sped up by a factor of four --
20
76680
2456
01:19
on the main screen you'll see the map that it's building.
21
79160
2667
01:21
So this is a high-resolution map of the corridor around our laboratory.
22
81851
4285
01:26
And in a minute you'll see it enter our lab,
23
86160
2336
01:28
which is recognizable by the clutter that you see.
24
88520
2856
01:31
(Laughter)
25
91400
1016
01:32
But the main point I want to convey to you
26
92440
2007
01:34
is that these robots are capable of building high-resolution maps
27
94472
3584
01:38
at five centimeters resolution,
28
98080
2496
01:40
allowing somebody who is outside the lab, or outside the building
29
100600
4176
01:44
to deploy these without actually going inside,
30
104800
3216
01:48
and trying to infer what happens inside the building.
31
108040
3760
01:52
Now there's one problem with robots like this.
32
112400
2240
01:55
The first problem is it's pretty big.
33
115600
2200
01:58
Because it's big, it's heavy.
34
118120
1680
02:00
And these robots consume about 100 watts per pound.
35
120640
3040
02:04
And this makes for a very short mission life.
36
124360
2280
02:08
The second problem
37
128000
1456
02:09
is that these robots have onboard sensors that end up being very expensive --
38
129480
3896
02:13
a laser scanner, a camera and the processors.
39
133400
3440
02:17
That drives up the cost of this robot.
40
137280
3040
02:21
So we asked ourselves a question:
41
141440
2656
02:24
what consumer product can you buy in an electronics store
42
144120
3776
02:27
that is inexpensive, that's lightweight, that has sensing onboard and computation?
43
147920
6280
02:36
And we invented the flying phone.
44
156080
2656
02:38
(Laughter)
45
158760
1936
02:40
So this robot uses a Samsung Galaxy smartphone that you can buy off the shelf,
46
160720
6176
02:46
and all you need is an app that you can download from our app store.
47
166920
4016
02:50
And you can see this robot reading the letters, "TED" in this case,
48
170960
4216
02:55
looking at the corners of the "T" and the "E"
49
175200
2936
02:58
and then triangulating off of that, flying autonomously.
50
178160
3480
03:02
That joystick is just there to make sure if the robot goes crazy,
51
182720
3256
03:06
Giuseppe can kill it.
52
186000
1416
03:07
(Laughter)
53
187440
1640
03:10
In addition to building these small robots,
54
190920
3816
03:14
we also experiment with aggressive behaviors, like you see here.
55
194760
4800
03:19
So this robot is now traveling at two to three meters per second,
56
199920
5296
03:25
pitching and rolling aggressively as it changes direction.
57
205240
3496
03:28
The main point is we can have smaller robots that can go faster
58
208760
4256
03:33
and then travel in these very unstructured environments.
59
213040
2960
03:37
And in this next video,
60
217120
2056
03:39
just like you see this bird, an eagle, gracefully coordinating its wings,
61
219200
5896
03:45
its eyes and feet to grab prey out of the water,
62
225120
4296
03:49
our robot can go fishing, too.
63
229440
1896
03:51
(Laughter)
64
231360
1496
03:52
In this case, this is a Philly cheesesteak hoagie that it's grabbing out of thin air.
65
232880
4056
03:56
(Laughter)
66
236960
2400
03:59
So you can see this robot going at about three meters per second,
67
239680
3296
04:03
which is faster than walking speed, coordinating its arms, its claws
68
243000
5136
04:08
and its flight with split-second timing to achieve this maneuver.
69
248160
4120
04:14
In another experiment,
70
254120
1216
04:15
I want to show you how the robot adapts its flight
71
255360
3656
04:19
to control its suspended payload,
72
259040
2376
04:21
whose length is actually larger than the width of the window.
73
261440
3800
04:25
So in order to accomplish this,
74
265680
1696
04:27
it actually has to pitch and adjust the altitude
75
267400
3696
04:31
and swing the payload through.
76
271120
2320
04:38
But of course we want to make these even smaller,
77
278920
2296
04:41
and we're inspired in particular by honeybees.
78
281240
3016
04:44
So if you look at honeybees, and this is a slowed down video,
79
284280
3256
04:47
they're so small, the inertia is so lightweight --
80
287560
3720
04:51
(Laughter)
81
291960
1176
04:53
that they don't care -- they bounce off my hand, for example.
82
293160
3536
04:56
This is a little robot that mimics the honeybee behavior.
83
296720
3160
05:00
And smaller is better,
84
300600
1216
05:01
because along with the small size you get lower inertia.
85
301840
3536
05:05
Along with lower inertia --
86
305400
1536
05:06
(Robot buzzing, laughter)
87
306960
2856
05:09
along with lower inertia, you're resistant to collisions.
88
309840
2816
05:12
And that makes you more robust.
89
312680
1720
05:15
So just like these honeybees, we build small robots.
90
315800
2656
05:18
And this particular one is only 25 grams in weight.
91
318480
3376
05:21
It consumes only six watts of power.
92
321880
2160
05:24
And it can travel up to six meters per second.
93
324440
2536
05:27
So if I normalize that to its size,
94
327000
2336
05:29
it's like a Boeing 787 traveling ten times the speed of sound.
95
329360
3640
05:36
(Laughter)
96
336000
2096
05:38
And I want to show you an example.
97
338120
1920
05:40
This is probably the first planned mid-air collision, at one-twentieth normal speed.
98
340840
5256
05:46
These are going at a relative speed of two meters per second,
99
346120
2858
05:49
and this illustrates the basic principle.
100
349002
2480
05:52
The two-gram carbon fiber cage around it prevents the propellers from entangling,
101
352200
4976
05:57
but essentially the collision is absorbed and the robot responds to the collisions.
102
357200
5296
06:02
And so small also means safe.
103
362520
2560
06:05
In my lab, as we developed these robots,
104
365400
2016
06:07
we start off with these big robots
105
367440
1620
06:09
and then now we're down to these small robots.
106
369084
2812
06:11
And if you plot a histogram of the number of Band-Aids we've ordered
107
371920
3456
06:15
in the past, that sort of tailed off now.
108
375400
2576
06:18
Because these robots are really safe.
109
378000
1960
06:20
The small size has some disadvantages,
110
380760
2456
06:23
and nature has found a number of ways to compensate for these disadvantages.
111
383240
4080
06:27
The basic idea is they aggregate to form large groups, or swarms.
112
387960
4000
06:32
So, similarly, in our lab, we try to create artificial robot swarms.
113
392320
3976
06:36
And this is quite challenging
114
396320
1381
06:37
because now you have to think about networks of robots.
115
397725
3320
06:41
And within each robot,
116
401360
1296
06:42
you have to think about the interplay of sensing, communication, computation --
117
402680
5616
06:48
and this network then becomes quite difficult to control and manage.
118
408320
4960
06:54
So from nature we take away three organizing principles
119
414160
3296
06:57
that essentially allow us to develop our algorithms.
120
417480
3160
07:01
The first idea is that robots need to be aware of their neighbors.
121
421640
4536
07:06
They need to be able to sense and communicate with their neighbors.
122
426200
3440
07:10
So this video illustrates the basic idea.
123
430040
2656
07:12
You have four robots --
124
432720
1296
07:14
one of the robots has actually been hijacked by a human operator, literally.
125
434040
4240
07:19
But because the robots interact with each other,
126
439217
2239
07:21
they sense their neighbors,
127
441480
1656
07:23
they essentially follow.
128
443160
1296
07:24
And here there's a single person able to lead this network of followers.
129
444480
5360
07:32
So again, it's not because all the robots know where they're supposed to go.
130
452000
5056
07:37
It's because they're just reacting to the positions of their neighbors.
131
457080
4320
07:43
(Laughter)
132
463720
4120
07:48
So the next experiment illustrates the second organizing principle.
133
468280
5240
07:54
And this principle has to do with the principle of anonymity.
134
474920
3800
07:59
Here the key idea is that
135
479400
4296
08:03
the robots are agnostic to the identities of their neighbors.
136
483720
4240
08:08
They're asked to form a circular shape,
137
488440
2616
08:11
and no matter how many robots you introduce into the formation,
138
491080
3296
08:14
or how many robots you pull out,
139
494400
2576
08:17
each robot is simply reacting to its neighbor.
140
497000
3136
08:20
It's aware of the fact that it needs to form the circular shape,
141
500160
4976
08:25
but collaborating with its neighbors
142
505160
1776
08:26
it forms the shape without central coordination.
143
506960
3720
08:31
Now if you put these ideas together,
144
511520
2416
08:33
the third idea is that we essentially give these robots
145
513960
3896
08:37
mathematical descriptions of the shape they need to execute.
146
517880
4296
08:42
And these shapes can be varying as a function of time,
147
522200
3496
08:45
and you'll see these robots start from a circular formation,
148
525720
4496
08:50
change into a rectangular formation, stretch into a straight line,
149
530240
3256
08:53
back into an ellipse.
150
533520
1375
08:54
And they do this with the same kind of split-second coordination
151
534919
3617
08:58
that you see in natural swarms, in nature.
152
538560
3280
09:03
So why work with swarms?
153
543080
2136
09:05
Let me tell you about two applications that we are very interested in.
154
545240
4120
09:10
The first one has to do with agriculture,
155
550160
2376
09:12
which is probably the biggest problem that we're facing worldwide.
156
552560
3360
09:16
As you well know,
157
556760
1256
09:18
one in every seven persons in this earth is malnourished.
158
558040
3520
09:21
Most of the land that we can cultivate has already been cultivated.
159
561920
3480
09:25
And the efficiency of most systems in the world is improving,
160
565960
3216
09:29
but our production system efficiency is actually declining.
161
569200
3520
09:33
And that's mostly because of water shortage, crop diseases, climate change
162
573080
4216
09:37
and a couple of other things.
163
577320
1520
09:39
So what can robots do?
164
579360
1480
09:41
Well, we adopt an approach that's called Precision Farming in the community.
165
581200
4616
09:45
And the basic idea is that we fly aerial robots through orchards,
166
585840
5376
09:51
and then we build precision models of individual plants.
167
591240
3120
09:54
So just like personalized medicine,
168
594829
1667
09:56
while you might imagine wanting to treat every patient individually,
169
596520
4816
10:01
what we'd like to do is build models of individual plants
170
601360
3696
10:05
and then tell the farmer what kind of inputs every plant needs --
171
605080
4136
10:09
the inputs in this case being water, fertilizer and pesticide.
172
609240
4440
10:14
Here you'll see robots traveling through an apple orchard,
173
614640
3616
10:18
and in a minute you'll see two of its companions
174
618280
2256
10:20
doing the same thing on the left side.
175
620560
1810
10:22
And what they're doing is essentially building a map of the orchard.
176
622800
3656
10:26
Within the map is a map of every plant in this orchard.
177
626480
2816
10:29
(Robot buzzing)
178
629320
1656
10:31
Let's see what those maps look like.
179
631000
1896
10:32
In the next video, you'll see the cameras that are being used on this robot.
180
632920
4296
10:37
On the top-left is essentially a standard color camera.
181
637240
3240
10:41
On the left-center is an infrared camera.
182
641640
3296
10:44
And on the bottom-left is a thermal camera.
183
644960
3776
10:48
And on the main panel, you're seeing a three-dimensional reconstruction
184
648760
3336
10:52
of every tree in the orchard as the sensors fly right past the trees.
185
652120
6120
10:59
Armed with information like this, we can do several things.
186
659640
4040
11:04
The first and possibly the most important thing we can do is very simple:
187
664200
4256
11:08
count the number of fruits on every tree.
188
668480
2440
11:11
By doing this, you tell the farmer how many fruits she has in every tree
189
671520
4536
11:16
and allow her to estimate the yield in the orchard,
190
676080
4256
11:20
optimizing the production chain downstream.
191
680360
2840
11:23
The second thing we can do
192
683640
1616
11:25
is take models of plants, construct three-dimensional reconstructions,
193
685280
4496
11:29
and from that estimate the canopy size,
194
689800
2536
11:32
and then correlate the canopy size to the amount of leaf area on every plant.
195
692360
3776
11:36
And this is called the leaf area index.
196
696160
2176
11:38
So if you know this leaf area index,
197
698360
1936
11:40
you essentially have a measure of how much photosynthesis is possible in every plant,
198
700320
5456
11:45
which again tells you how healthy each plant is.
199
705800
2880
11:49
By combining visual and infrared information,
200
709520
4216
11:53
we can also compute indices such as NDVI.
201
713760
3296
11:57
And in this particular case, you can essentially see
202
717080
2816
11:59
there are some crops that are not doing as well as other crops.
203
719920
3016
12:02
This is easily discernible from imagery,
204
722960
4056
12:07
not just visual imagery but combining
205
727040
2216
12:09
both visual imagery and infrared imagery.
206
729280
2776
12:12
And then lastly,
207
732080
1336
12:13
one thing we're interested in doing is detecting the early onset of chlorosis --
208
733440
4016
12:17
and this is an orange tree --
209
737480
1496
12:19
which is essentially seen by yellowing of leaves.
210
739000
2560
12:21
But robots flying overhead can easily spot this autonomously
211
741880
3896
12:25
and then report to the farmer that he or she has a problem
212
745800
2936
12:28
in this section of the orchard.
213
748760
1520
12:30
Systems like this can really help,
214
750800
2696
12:33
and we're projecting yields that can improve by about ten percent
215
753520
5816
12:39
and, more importantly, decrease the amount of inputs such as water
216
759360
3216
12:42
by 25 percent by using aerial robot swarms.
217
762600
3280
12:47
Lastly, I want you to applaud the people who actually create the future,
218
767200
5736
12:52
Yash Mulgaonkar, Sikang Liu and Giuseppe Loianno,
219
772960
4920
12:57
who are responsible for the three demonstrations that you saw.
220
777920
3496
13:01
Thank you.
221
781440
1176
13:02
(Applause)
222
782640
5920
About this website

This site will introduce you to YouTube videos that are useful for learning English. You will see English lessons taught by top-notch teachers from around the world. Double-click on the English subtitles displayed on each video page to play the video from there. The subtitles scroll in sync with the video playback. If you have any comments or requests, please contact us using this contact form.

https://forms.gle/WvT1wiN1qDtmnspy7