The Future of Flying Robots | Vijay Kumar | TED Talks

755,740 views ใƒป 2015-11-04

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

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

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7