TED2012
Vijay Kumar: Robots that fly ... and cooperate
维杰·库马：善于合作的飞行机器人
Filmed:
Readability: 4.5
5,188,706 views
在宾夕法尼亚的实验室里，维杰·库马和他的团队造出了超小的旋翼飞行器。 这些灵活的机器人能够群组飞行，能感应同伴，从事特别行动——建筑工地，紧急救援等等
Vijay Kumar - Roboticist
As the dean of the University of Pennsylvania's School of Engineering and Applied Science, Vijay Kumar studies the control and coordination of multi-robot formations. Full bio
As the dean of the University of Pennsylvania's School of Engineering and Applied Science, Vijay Kumar studies the control and coordination of multi-robot formations. Full bio
Double-click the English transcript below to play the video.
00:20
Good morning.
0
5000
2000
早上好
00:22
I'm here today to talk
1
7000
2000
我今天想谈谈
00:24
about autonomous, flying beach balls.
2
9000
3000
自主飞行沙滩球
00:27
No, agile aerial robots like this one.
3
12000
4000
其实，是小型飞行器，像这一个
00:31
I'd like to tell you a little bit about the challenges in building these
4
16000
3000
我想和大家谈谈设计这些飞行器时的挑战
00:34
and some of the terrific opportunities
5
19000
2000
和使用这些飞行器能给我们带来的
00:36
for applying this technology.
6
21000
2000
很多用处
00:38
So these robots
7
23000
2000
这些飞行器
00:40
are related to unmanned aerial vehicles.
8
25000
3000
源于无人驾驶的飞行器
00:43
However, the vehicles you see here are big.
9
28000
3000
但是那些都体积很大
00:46
They weigh thousands of pounds,
10
31000
2000
通常上万磅重
00:48
are not by any means agile.
11
33000
2000
毫无灵活型可言
00:50
They're not even autonomous.
12
35000
2000
它们也不是真的自主飞行的
00:52
In fact, many of these vehicles
13
37000
2000
事实上，很多这些飞行器
00:54
are operated by flight crews
14
39000
2000
都是受飞行团队控制的
00:56
that can include multiple pilots,
15
41000
3000
包括好几个飞行员
00:59
operators of sensors
16
44000
2000
感应雷达操作员
01:01
and mission coordinators.
17
46000
2000
和团队协调员
01:03
What we're interested in is developing robots like this --
18
48000
2000
我们想设计的飞行器是这样的——
01:05
and here are two other pictures --
19
50000
2000
这里有两张照片——
01:07
of robots that you can buy off the shelf.
20
52000
3000
是你能够在超市里买到的那种小飞行器
01:10
So these are helicopters with four rotors
21
55000
3000
小型直升机，四个螺旋桨
01:13
and they're roughly a meter or so in scale
22
58000
4000
不超过一米长
01:17
and weigh several pounds.
23
62000
2000
只不过几磅重
01:19
And so we retrofit these with sensors and processors,
24
64000
3000
我们把它们稍微改造一下，加上感应器和处理器，
01:22
and these robots can fly indoors
25
67000
2000
它们就可以在室内飞
01:24
without GPS.
26
69000
2000
用不着导航系统
01:26
The robot I'm holding in my hand
27
71000
2000
我现在拿着的这个飞行器
01:28
is this one,
28
73000
2000
是其中之一
01:30
and it's been created by two students,
29
75000
3000
是两个学生做出来的
01:33
Alex and Daniel.
30
78000
2000
艾利克斯和丹尼尔
01:35
So this weighs a little more
31
80000
2000
这个仅仅比零点一磅
01:37
than a tenth of a pound.
32
82000
2000
稍微重一点
01:39
It consumes about 15 watts of power.
33
84000
2000
只需要大约十五瓦的电源
01:41
And as you can see,
34
86000
2000
你能看到
01:43
it's about eight inches in diameter.
35
88000
2000
它的直径大约只有八个英寸
01:45
So let me give you just a very quick tutorial
36
90000
3000
让我给你们快速解释一下
01:48
on how these robots work.
37
93000
2000
这些飞行器是怎么工作的
01:50
So it has four rotors.
38
95000
2000
它有四个螺旋桨
01:52
If you spin these rotors at the same speed,
39
97000
2000
当四个螺旋桨转速相同
01:54
the robot hovers.
40
99000
2000
这个飞行器就浮在空中
01:56
If you increase the speed of each of these rotors,
41
101000
3000
当所有螺旋桨的速度提升时
01:59
then the robot flies up, it accelerates up.
42
104000
3000
这个飞行器就加速升高
02:02
Of course, if the robot were tilted,
43
107000
2000
当然了，如果飞行器已经是倾斜的
02:04
inclined to the horizontal,
44
109000
2000
向着地平线侧过来
02:06
then it would accelerate in this direction.
45
111000
3000
就会向这个方向加速
02:09
So to get it to tilt, there's one of two ways of doing it.
46
114000
3000
怎么能让它侧过来呢，有两个途径
02:12
So in this picture
47
117000
2000
从这张照片
02:14
you see that rotor four is spinning faster
48
119000
2000
你能看到四号螺旋桨旋转加速
02:16
and rotor two is spinning slower.
49
121000
2000
同时二号螺旋桨转速变慢
02:18
And when that happens
50
123000
2000
这时
02:20
there's moment that causes this robot to roll.
51
125000
3000
飞行器就能向一边倒
02:23
And the other way around,
52
128000
2000
反之亦然
02:25
if you increase the speed of rotor three
53
130000
3000
当三号螺旋桨加速
02:28
and decrease the speed of rotor one,
54
133000
2000
一号减速时
02:30
then the robot pitches forward.
55
135000
3000
飞行器就向前倒
02:33
And then finally,
56
138000
2000
最后
02:35
if you spin opposite pairs of rotors
57
140000
2000
如果任意两端的螺旋桨的转速
02:37
faster than the other pair,
58
142000
2000
大于另两端的螺旋桨的转速
02:39
then the robot yaws about the vertical axis.
59
144000
2000
飞行器就能原地旋转
02:41
So an on-board processor
60
146000
2000
所以装在飞行器上的处理器
02:43
essentially looks at what motions need to be executed
61
148000
3000
基本上能判断需要执行哪些动作
02:46
and combines these motions
62
151000
2000
然后把它们组合起来
02:48
and figures out what commands to send to the motors
63
153000
3000
决定给螺旋桨下什么指令
02:51
600 times a second.
64
156000
2000
一秒钟六百次
02:53
That's basically how this thing operates.
65
158000
2000
简单地说这些飞行器就是这么工作的
02:55
So one of the advantages of this design
66
160000
2000
这个设计的一个好处
02:57
is, when you scale things down,
67
162000
2000
就是小巧
02:59
the robot naturally becomes agile.
68
164000
3000
这些飞行器很灵活
03:02
So here R
69
167000
2000
这里的R
03:04
is the characteristic length of the robot.
70
169000
2000
是飞行器的长度
03:06
It's actually half the diameter.
71
171000
3000
其实是半径
03:09
And there are lots of physical parameters that change
72
174000
3000
当半径变小时
03:12
as you reduce R.
73
177000
2000
很多物理参数都会变
03:14
The one that's the most important
74
179000
2000
最重要的一个参数是
03:16
is the inertia or the resistance to motion.
75
181000
2000
惯性， 也就是对于运动的阻力
03:18
So it turns out,
76
183000
2000
结果是
03:20
the inertia, which governs angular motion,
77
185000
3000
惯性决定角速度
03:23
scales as a fifth power of R.
78
188000
3000
它是半径的五次方函数
03:26
So the smaller you make R,
79
191000
2000
当半径变得越来越小时
03:28
the more dramatically the inertia reduces.
80
193000
3000
惯性越来越快地减小
03:31
So as a result, the angular acceleration,
81
196000
3000
另一个结果是角速度的加速度
03:34
denoted by Greek letter alpha here,
82
199000
2000
也就是这里的希腊字母alpha
03:36
goes as one over R.
83
201000
2000
等于一除以半径
03:38
It's inversely proportional to R.
84
203000
2000
也就是半径的倒数
03:40
The smaller you make it the more quickly you can turn.
85
205000
3000
当半径越小时飞行器能转弯越快
03:43
So this should be clear in these videos.
86
208000
2000
这个视频清楚地显示
03:45
At the bottom right you see a robot
87
210000
3000
大家看右下角的飞行器
03:48
performing a 360 degree flip
88
213000
2000
正在做一个三百六十度翻转
03:50
in less than half a second.
89
215000
2000
只需要不到半秒
03:52
Multiple flips, a little more time.
90
217000
3000
连续翻转，稍微时间长一点
03:55
So here the processes on board
91
220000
2000
这里飞行器上用的处理器
03:57
are getting feedback from accelerometers
92
222000
2000
能够从飞行器上的加速度计
03:59
and gyros on board
93
224000
2000
和陀螺仪得到反馈信息
04:01
and calculating, like I said before,
94
226000
2000
然后算出，就像我刚才讲的
04:03
commands at 600 times a second
95
228000
2000
一秒钟六百个指令
04:05
to stabilize this robot.
96
230000
2000
来稳定控制这个飞行器
04:07
So on the left, you see Daniel throwing this robot up into the air.
97
232000
3000
在左边你能看到丹尼尔把飞行器抛到空中
04:10
And it shows you how robust the control is.
98
235000
2000
你能看到飞行器的控制有多快
04:12
No matter how you throw it,
99
237000
2000
不管你怎么扔
04:14
the robot recovers and comes back to him.
100
239000
4000
飞行器都能恢复平衡飞回来
04:18
So why build robots like this?
101
243000
2000
为什么我们要设计这种飞行器呢？
04:20
Well robots like this have many applications.
102
245000
3000
因为这样的飞行器有很多用处
04:23
You can send them inside buildings like this
103
248000
3000
你能把它们放进像这样的大楼里
04:26
as first responders to look for intruders,
104
251000
3000
作为报警器去寻找入侵者
04:29
maybe look for biochemical leaks,
105
254000
3000
寻找生化泄漏
04:32
gaseous leaks.
106
257000
2000
或者煤气泄漏
04:34
You can also use them
107
259000
2000
你还能用它们
04:36
for applications like construction.
108
261000
2000
建摩天大楼呢
04:38
So here are robots carrying beams, columns
109
263000
4000
这里是飞行器在搬梁运柱
04:42
and assembling cube-like structures.
110
267000
3000
架构一个立方体的建筑
04:45
I'll tell you a little bit more about this.
111
270000
3000
这里我想和大家介绍一下
04:48
The robots can be used for transporting cargo.
112
273000
3000
这些机器人能被用来运货
04:51
So one of the problems with these small robots
113
276000
3000
当然一个问题是这些小飞行器
04:54
is their payload carrying capacity.
114
279000
2000
担不了多少重量
04:56
So you might want to have multiple robots
115
281000
2000
你可能需要很多飞行器
04:58
carry payloads.
116
283000
2000
来搬运重物
05:00
This is a picture of a recent experiment we did --
117
285000
2000
我们新做了个实验——
05:02
actually not so recent anymore --
118
287000
2000
其实不那么新了——
05:04
in Sendai shortly after the earthquake.
119
289000
3000
在日本仙台，地震后不久
05:07
So robots like this could be sent into collapsed buildings
120
292000
3000
我们能把这些飞行器
05:10
to assess the damage after natural disasters,
121
295000
2000
送进倒塌的楼房
05:12
or sent into reactor buildings
122
297000
3000
或者核反应堆大楼
05:15
to map radiation levels.
123
300000
3000
来探测放射性强度
05:19
So one fundamental problem
124
304000
2000
一个根本的问题
05:21
that the robots have to solve if they're to be autonomous
125
306000
3000
是当这些飞行器需要自控飞行，
05:24
is essentially figuring out
126
309000
2000
它们自己得弄明白
05:26
how to get from point A to point B.
127
311000
2000
怎么从一个地点到另一个地点
05:28
So this gets a little challenging
128
313000
2000
这就变得有点难度了
05:30
because the dynamics of this robot are quite complicated.
129
315000
3000
因为这些飞行器的动力学是很复杂的
05:33
In fact, they live in a 12-dimensional space.
130
318000
2000
事实上它们总在对付十二维的空间
05:35
So we use a little trick.
131
320000
2000
这里我们用了一点小技巧
05:37
We take this curved 12-dimensional space
132
322000
3000
我们拿这个十二位的空间
05:40
and transform it
133
325000
2000
把它们转换成
05:42
into a flat four-dimensional space.
134
327000
2000
平的四维空间
05:44
And that four-dimensional space
135
329000
2000
这个四维空间
05:46
consists of X, Y, Z and then the yaw angle.
136
331000
3000
包括了横轴，纵轴和竖轴，还有旋转轴
05:49
And so what the robot does
137
334000
2000
这些飞行器只需要
05:51
is it plans what we call a minimum snap trajectory.
138
336000
4000
计划一件事，我们管它叫最小化加加加速度轨道
05:55
So to remind you of physics,
139
340000
2000
提醒大家一点点物理学
05:57
you have position, derivative, velocity,
140
342000
2000
这里我们有位置向量，导数，速度
05:59
then acceleration,
141
344000
2000
和加速度
06:01
and then comes jerk
142
346000
2000
还有加加速度
06:03
and then comes snap.
143
348000
2000
还有加加加速度
06:05
So this robot minimizes snap.
144
350000
3000
这个飞行器把加加加速度最小化
06:08
So what that effectively does
145
353000
2000
基本上它的工作是
06:10
is produces a smooth and graceful motion.
146
355000
2000
创造一个光滑优雅的运动曲线
06:12
And it does that avoiding obstacles.
147
357000
3000
这样来绕开障碍物
06:15
So these minimum snap trajectories in this flat space
148
360000
3000
所以这个四维平面中，这个飞行器使用
06:18
are then transformed back
149
363000
2000
最小化加加加速度轨道， 然后转换回到
06:20
into this complicated 12-dimensional space,
150
365000
2000
复杂的十二维空间
06:22
which the robot must do
151
367000
2000
飞行器必须这样做来
06:24
for control and then execution.
152
369000
2000
获得控制和执行动作
06:26
So let me show you some examples
153
371000
2000
让我给大家看几个例子
06:28
of what these minimum snap trajectories look like.
154
373000
2000
这些最小化加加加速度轨道是什么样的
06:30
And in the first video,
155
375000
2000
这是第一个视频
06:32
you'll see the robot going from point A to point B
156
377000
2000
这个飞行器从一个地点飞到另一个地点
06:34
through an intermediate point.
157
379000
2000
中间经停一下
06:42
So the robot is obviously capable
158
387000
2000
显然这个飞行器能
06:44
of executing any curve trajectory.
159
389000
2000
飞出一个曲线轨道
06:46
So these are circular trajectories
160
391000
2000
还有这样的打圈的轨道
06:48
where the robot pulls about two G's.
161
393000
3000
这里飞行器对抗两倍的重力
06:52
Here you have overhead motion capture cameras on the top
162
397000
4000
它们上方还有一个动感监控摄像机，每秒一百幅画面
06:56
that tell the robot where it is 100 times a second.
163
401000
3000
来告诉这些飞行器它们的位置
06:59
It also tells the robot where these obstacles are.
164
404000
3000
也能告诉这些飞行器障碍物在哪里
07:02
And the obstacles can be moving.
165
407000
2000
障碍物移动都不要紧
07:04
And here you'll see Daniel throw this hoop into the air,
166
409000
3000
当丹尼尔把套圈扔到空中
07:07
while the robot is calculating the position of the hoop
167
412000
2000
飞行器就开始计算套圈的位置
07:09
and trying to figure out how to best go through the hoop.
168
414000
4000
试图预测怎么才能最有效地钻过去
07:13
So as an academic,
169
418000
2000
作为一个科研人员
07:15
we're always trained to be able to jump through hoops to raise funding for our labs,
170
420000
3000
我们总在试图钻出重重圈套，拿到更多经费
07:18
and we get our robots to do that.
171
423000
3000
甚至训练了我们的飞行器也来做这个
07:21
(Applause)
172
426000
6000
（掌声）
07:27
So another thing the robot can do
173
432000
2000
另一个飞行器能做的事情
07:29
is it remembers pieces of trajectory
174
434000
3000
是当我们预先编入一些轨迹
07:32
that it learns or is pre-programmed.
175
437000
2000
或者它自己学着走过的，它能够记住
07:34
So here you see the robot
176
439000
2000
这里大家能看到
07:36
combining a motion
177
441000
2000
飞行器能够（在预设轨迹上）加上一个动作
07:38
that builds up momentum
178
443000
2000
积聚动量
07:40
and then changes its orientation and then recovers.
179
445000
3000
改变它的定向，再回到预设轨迹上来
07:43
So it has to do this because this gap in the window
180
448000
3000
它必须这样做因为这个窗上的缝隙
07:46
is only slightly larger than the width of the robot.
181
451000
4000
只比它的宽度大一点点
07:50
So just like a diver stands on a springboard
182
455000
3000
所以就像是一个跳水运动员
07:53
and then jumps off it to gain momentum,
183
458000
2000
从跳板上起跳，聚集动量，
07:55
and then does this pirouette, this two and a half somersault through
184
460000
3000
做个旋转，两圈半
07:58
and then gracefully recovers,
185
463000
2000
然后优雅地回到平衡
08:00
this robot is basically doing that.
186
465000
2000
这个飞行器是自主这样做的
08:02
So it knows how to combine little bits and pieces of trajectories
187
467000
3000
它知道怎么把小段的轨迹组合起来
08:05
to do these fairly difficult tasks.
188
470000
4000
来做这些高难度的技巧
08:09
So I want change gears.
189
474000
2000
现在我想换个话题谈谈这些小型飞行器
08:11
So one of the disadvantages of these small robots is its size.
190
476000
3000
的不足之处，就是体积小
08:14
And I told you earlier
191
479000
2000
我已经提过
08:16
that we may want to employ lots and lots of robots
192
481000
2000
我们需要使用很多飞行器
08:18
to overcome the limitations of size.
193
483000
3000
来克服体积小的不便
08:21
So one difficulty
194
486000
2000
一个难点是
08:23
is how do you coordinate lots of these robots?
195
488000
3000
怎么使得这些飞行器集体飞行？
08:26
And so here we looked to nature.
196
491000
2000
我们在大自然中寻找答案
08:28
So I want to show you a clip
197
493000
2000
我想给大家看一个视频
08:30
of Aphaenogaster desert ants
198
495000
2000
是关于Aphaenogaster沙漠蚁的
08:32
in Professor Stephen Pratt's lab carrying an object.
199
497000
3000
在史狄文·普热特教授的实验室里，这些蚂蚁一起搬运重物
08:35
So this is actually a piece of fig.
200
500000
2000
这是一个无花果
08:37
Actually you take any object coated with fig juice
201
502000
2000
事实上无论什么东西，只要蘸上无花果汁
08:39
and the ants will carry them back to the nest.
202
504000
3000
这些蚂蚁都会把它们带回巢去
08:42
So these ants don't have any central coordinator.
203
507000
3000
这些蚂蚁没有任何中央调控
08:45
They sense their neighbors.
204
510000
2000
它们是靠感应邻近的蚂蚁
08:47
There's no explicit communication.
205
512000
2000
它们也没有明确的交流
08:49
But because they sense the neighbors
206
514000
2000
但是因为它们能够感应邻近的蚂蚁
08:51
and because they sense the object,
207
516000
2000
也能感应抬着的重物
08:53
they have implicit coordination across the group.
208
518000
3000
整群的蚂蚁有默契
08:56
So this is the kind of coordination
209
521000
2000
这样的协调
08:58
we want our robots to have.
210
523000
3000
正是飞行器需要的
09:01
So when we have a robot
211
526000
2000
当一个飞行器
09:03
which is surrounded by neighbors --
212
528000
2000
被其他飞行器环绕时——
09:05
and let's look at robot I and robot J --
213
530000
2000
让我们注意 I 和 J 这两个——
09:07
what we want the robots to do
214
532000
2000
当它们成群飞行时
09:09
is to monitor the separation between them
215
534000
3000
我们希望这两个飞行器
09:12
as they fly in formation.
216
537000
2000
能够监控它们之间的距离
09:14
And then you want to make sure
217
539000
2000
我们需要确定
09:16
that this separation is within acceptable levels.
218
541000
2000
这个距离是在可接受的范围里的
09:18
So again the robots monitor this error
219
543000
3000
飞行器要检测这个变化
09:21
and calculate the control commands
220
546000
2000
在控制指令中计算进去
09:23
100 times a second,
221
548000
2000
也是每秒一百次
09:25
which then translates to the motor commands 600 times a second.
222
550000
3000
这个控制指令每秒会被送到马达六百次
09:28
So this also has to be done
223
553000
2000
所以这个程序
09:30
in a decentralized way.
224
555000
2000
是分散化执行的
09:32
Again, if you have lots and lots of robots,
225
557000
2000
再有，如果你有很多很多飞行器
09:34
it's impossible to coordinate all this information centrally
226
559000
4000
要完成集体飞行任务，能足够快地集中协调所有这些信息
09:38
fast enough in order for the robots to accomplish the task.
227
563000
3000
是几乎不可能的
09:41
Plus the robots have to base their actions
228
566000
2000
加上这些飞行器只能
09:43
only on local information,
229
568000
2000
依靠局部的信息来决定做什么动作
09:45
what they sense from their neighbors.
230
570000
2000
也就是要靠感应邻近的飞行器
09:47
And then finally,
231
572000
2000
最后
09:49
we insist that the robots be agnostic
232
574000
2000
我们希望这些机器人
09:51
to who their neighbors are.
233
576000
2000
不知道它们的邻居是谁
09:53
So this is what we call anonymity.
234
578000
3000
也就是匿名飞行
09:56
So what I want to show you next
235
581000
2000
下一个我想给大家展示的
09:58
is a video
236
583000
2000
是这段视频
10:00
of 20 of these little robots
237
585000
3000
这二十个小型飞行器
10:03
flying in formation.
238
588000
2000
成群飞行
10:05
They're monitoring their neighbors' position.
239
590000
3000
它们在监测邻居的位置
10:08
They're maintaining formation.
240
593000
2000
维持群队
10:10
The formations can change.
241
595000
2000
群队的形状还能变
10:12
They can be planar formations,
242
597000
2000
它们可以在一个平面上飞
10:14
they can be three-dimensional formations.
243
599000
2000
也可以上中下地飞
10:16
As you can see here,
244
601000
2000
大家可以看到
10:18
they collapse from a three-dimensional formation into planar formation.
245
603000
3000
它们能从上中下的群队变成平面的
10:21
And to fly through obstacles
246
606000
2000
在飞越障碍物的时候
10:23
they can adapt the formations on the fly.
247
608000
4000
它们能边飞边变换队形
10:27
So again, these robots come really close together.
248
612000
3000
我想强调，这些飞行器距离都很近
10:30
As you can see in this figure-eight flight,
249
615000
2000
比如这个群队，八架飞行器
10:32
they come within inches of each other.
250
617000
2000
相互距离不过几英寸
10:34
And despite the aerodynamic interactions
251
619000
3000
尽管在空气动力学上
10:37
of these propeller blades,
252
622000
2000
这些螺旋桨相互干扰
10:39
they're able to maintain stable flight.
253
624000
2000
它们还是能够维持平稳飞行
10:41
(Applause)
254
626000
7000
（掌声）
10:48
So once you know how to fly in formation,
255
633000
2000
现在它们会成群飞了
10:50
you can actually pick up objects cooperatively.
256
635000
2000
它们就可以合作抬重物
10:52
So this just shows
257
637000
2000
这里展示的是
10:54
that we can double, triple, quadruple
258
639000
3000
我们能够把飞行器的能力
10:57
the robot strength
259
642000
2000
翻倍，翻三倍，四倍
10:59
by just getting them to team with neighbors, as you can see here.
260
644000
2000
仅仅通过让它们和邻居合作，大家可以看到
11:01
One of the disadvantages of doing that
261
646000
3000
这样做的一个不便之处
11:04
is, as you scale things up --
262
649000
2000
就是当加大数量时——
11:06
so if you have lots of robots carrying the same thing,
263
651000
2000
比如使用很多飞行器来抬一个物体
11:08
you're essentially effectively increasing the inertia,
264
653000
3000
你其实是加大了惯性
11:11
and therefore you pay a price; they're not as agile.
265
656000
3000
这样它们就不够灵活了，这是一个代价
11:14
But you do gain in terms of payload carrying capacity.
266
659000
3000
但是你可以增加载荷承载量
11:17
Another application I want to show you --
267
662000
2000
另一个我想给大家展示的用处是——
11:19
again, this is in our lab.
268
664000
2000
这是在我们实验室
11:21
This is work done by Quentin Lindsey who's a graduate student.
269
666000
2000
这是研究生昆汀·林夕的工作
11:23
So his algorithm essentially tells these robots
270
668000
3000
他的算法程序告诉这些飞行器
11:26
how to autonomously build
271
671000
2000
怎么使用桁架结构
11:28
cubic structures
272
673000
2000
自动建造
11:30
from truss-like elements.
273
675000
3000
一个立方体
11:33
So his algorithm tells the robot
274
678000
2000
他的算法程序告诉这些机器人
11:35
what part to pick up,
275
680000
2000
该用哪一块
11:37
when and where to place it.
276
682000
2000
什么时候用，用在哪里
11:39
So in this video you see --
277
684000
2000
从这个视频我们可以看到——
11:41
and it's sped up 10, 14 times --
278
686000
2000
这个视频是十倍或者十四倍速度播放的——
11:43
you see three different structures being built by these robots.
279
688000
3000
大家可以看到飞行器在搭建很不一样的构架
11:46
And again, everything is autonomous,
280
691000
2000
并且，所有的运动都是自主的
11:48
and all Quentin has to do
281
693000
2000
昆汀仅仅是
11:50
is to get them a blueprint
282
695000
2000
给它们一个蓝图
11:52
of the design that he wants to build.
283
697000
4000
也就是他想建的设计
11:56
So all these experiments you've seen thus far,
284
701000
3000
所有这里展示的实验
11:59
all these demonstrations,
285
704000
2000
所有这些演习
12:01
have been done with the help of motion capture systems.
286
706000
3000
都是靠着它们自己的动感检测摄像机完成的
12:04
So what happens when you leave your lab
287
709000
2000
那么，当它们离开实验室
12:06
and you go outside into the real world?
288
711000
3000
来到真实世界的时候，又怎么样呢？
12:09
And what if there's no GPS?
289
714000
3000
没有卫星导航会怎么样？
12:12
So this robot
290
717000
2000
这个飞行器
12:14
is actually equipped with a camera
291
719000
2000
其实装有一个摄像机
12:16
and a laser rangefinder, laser scanner.
292
721000
3000
和一个激光测距仪，一个激光扫描仪
12:19
And it uses these sensors
293
724000
2000
它可以使用这些探测装置
12:21
to build a map of the environment.
294
726000
2000
来描绘周围的环境的地图
12:23
What that map consists of are features --
295
728000
3000
这个地图包括很多细节——
12:26
like doorways, windows,
296
731000
2000
玄关，窗户
12:28
people, furniture --
297
733000
2000
人，家具——
12:30
and it then figures out where its position is
298
735000
2000
还能弄清楚相对于这些东西
12:32
with respect to the features.
299
737000
2000
它自己在哪里
12:34
So there is no global coordinate system.
300
739000
2000
所以这里没有整体的协调系统
12:36
The coordinate system is defined based on the robot,
301
741000
3000
这个协调系统是靠飞行器自己来完成的
12:39
where it is and what it's looking at.
302
744000
3000
它自己在哪里，前面有什么
12:42
And it navigates with respect to those features.
303
747000
3000
还能利用周围环境为自己找到出路
12:45
So I want to show you a clip
304
750000
2000
这里我想给大家再看一段视频
12:47
of algorithms developed by Frank Shen
305
752000
2000
这个算法程序是法兰克·沈
12:49
and Professor Nathan Michael
306
754000
2000
和南希·麦克教授编的
12:51
that shows this robot entering a building for the very first time
307
756000
4000
当这个飞行器第一次飞入一个建筑
12:55
and creating this map on the fly.
308
760000
3000
它是怎么边飞边画地图的
12:58
So the robot then figures out what the features are.
309
763000
3000
这个飞行器弄明白了这些细节
13:01
It builds the map.
310
766000
2000
开始画地图
13:03
It figures out where it is with respect to the features
311
768000
2000
弄明白了相对这些细节，自己在哪里，
13:05
and then estimates its position
312
770000
2000
然后自我定位
13:07
100 times a second
313
772000
2000
全以每秒一百次的速度发生
13:09
allowing us to use the control algorithms
314
774000
2000
这就给我们一个机会来控制这些算法
13:11
that I described to you earlier.
315
776000
2000
像我之前讲过的
13:13
So this robot is actually being commanded
316
778000
2000
所以这个机器人其实是
13:15
remotely by Frank.
317
780000
2000
被法兰克遥控的
13:17
But the robot can also figure out
318
782000
2000
但是它自己也可以弄明白
13:19
where to go on its own.
319
784000
2000
怎么飞
13:21
So suppose I were to send this into a building
320
786000
2000
假设我想放一个这样的飞行器进一幢楼
13:23
and I had no idea what this building looked like,
321
788000
2000
我并不知道里面是什么样的
13:25
I can ask this robot to go in,
322
790000
2000
我可以让它飞进去
13:27
create a map
323
792000
2000
创造一个地图
13:29
and then come back and tell me what the building looks like.
324
794000
3000
然后飞回来告诉我里面是什么样的
13:32
So here, the robot is not only solving the problem,
325
797000
3000
所以，这个飞行器不仅仅解决了
13:35
how to go from point A to point B in this map,
326
800000
3000
怎么从一点到另一点的问题
13:38
but it's figuring out
327
803000
2000
还能够随时知道
13:40
what the best point B is at every time.
328
805000
2000
最好的目标在哪里
13:42
So essentially it knows where to go
329
807000
3000
基本上，它知道该去搜索哪里
13:45
to look for places that have the least information.
330
810000
2000
因为那里的信息是最“未知”的
13:47
And that's how it populates this map.
331
812000
3000
这就是它怎么填充这个地图
13:50
So I want to leave you
332
815000
2000
这里我想展示给大家
13:52
with one last application.
333
817000
2000
最后一个用途
13:54
And there are many applications of this technology.
334
819000
3000
当然这个技术有很多很多用途
13:57
I'm a professor, and we're passionate about education.
335
822000
2000
我是个教授，我们很关心教育
13:59
Robots like this can really change the way
336
824000
2000
这样的飞行器其实可以改变
14:01
we do K through 12 education.
337
826000
2000
我们的小学和中学教育
14:03
But we're in Southern California,
338
828000
2000
我们在南加州
14:05
close to Los Angeles,
339
830000
2000
离洛杉矶很近
14:07
so I have to conclude
340
832000
2000
所以我不得不
14:09
with something focused on entertainment.
341
834000
2000
放点娱乐元素进去
14:11
I want to conclude with a music video.
342
836000
2000
我想给大家看一个音乐视频
14:13
I want to introduce the creators, Alex and Daniel,
343
838000
3000
我想向你们介绍艾利克斯和丹尼尔，
14:16
who created this video.
344
841000
2000
他们是导演兼制作
14:18
(Applause)
345
843000
7000
（掌声）
14:25
So before I play this video,
346
850000
2000
在我播放这个视频前
14:27
I want to tell you that they created it in the last three days
347
852000
3000
我想告诉大家这是他们在过去三天做出来的
14:30
after getting a call from Chris.
348
855000
2000
因为主持人克瑞斯给我打了个电话
14:32
And the robots that play the video
349
857000
2000
在这个视频中表演的飞行器
14:34
are completely autonomous.
350
859000
2000
全是靠自控表演的
14:36
You will see nine robots play six different instruments.
351
861000
3000
你能看到九个机器人，演奏六种不同乐器
14:39
And of course, it's made exclusively for TED 2012.
352
864000
4000
当然了，这是为了今年的TED2012特别制作的
14:43
Let's watch.
353
868000
3000
请欣赏
15:19
(Music)
354
904000
10000
（音乐）
16:23
(Applause)
355
968000
17000
（掌声）
ABOUT THE SPEAKER
Vijay Kumar - RoboticistAs the dean of the University of Pennsylvania's School of Engineering and Applied Science, Vijay Kumar studies the control and coordination of multi-robot formations.
Why you should listen
At the General Robotics, Automation, Sensing and Perception (GRASP) Lab at the University of Pennsylvania, flying quadrotor robots move together in eerie formation, tightening themselves into perfect battalions, even filling in the gap when one of their own drops out. You might have seen viral videos of the quads zipping around the netting-draped GRASP Lab (they juggle! they fly through a hula hoop!). Vijay Kumar headed this lab from 1998-2004. He's now the dean of the School of Engineering and Applied Science at the University of Pennsylvania in Philadelphia, where he continues his work in robotics, blending computer science and mechanical engineering to create the next generation of robotic wonders.
Vijay Kumar | Speaker | TED.com