Origami robots that reshape and transform themselves | Jamie Paik

225,419 views ・ 2019-08-16

TED


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

00:13
As a roboticist, I get asked a lot of questions.
0
13944
3721
00:17
"When we will they start serving me breakfast?"
1
17689
2304
00:21
So I thought the future of robotics would be looking more like us.
2
21049
4676
00:28
I thought they would look like me,
3
28114
1674
00:29
so I built eyes that would simulate my eyes.
4
29812
3747
00:34
I built fingers that are dextrous enough to serve me ...
5
34813
5056
00:39
baseballs.
6
39893
1190
00:43
Classical robots like this
7
43837
2017
00:45
are built and become functional
8
45878
3176
00:49
based on the fixed number of joints and actuators.
9
49078
3198
00:52
And this means their functionality and shape are already fixed
10
52780
4335
00:57
at the moment of their conception.
11
57139
1831
00:59
So even though this arm has a really nice throw --
12
59490
2867
01:02
it even hit the tripod at the end--
13
62381
2431
01:06
it's not meant for cooking you breakfast per se.
14
66015
3163
01:09
It's not really suited for scrambled eggs.
15
69202
3637
01:12
So this was when I was hit by a new vision of future robotics:
16
72863
4325
01:18
the transformers.
17
78101
1760
01:20
They drive, they run, they fly,
18
80989
2547
01:23
all depending on the ever-changing, new environment and task at hand.
19
83560
4869
01:29
To make this a reality,
20
89267
1781
01:31
you really have to rethink how robots are designed.
21
91072
3291
01:35
So, imagine a robotic module in a polygon shape
22
95448
4014
01:39
and using that simple polygon shape
23
99486
2079
01:41
to reconstruct multiple different forms
24
101589
3038
01:44
to create a new form of robot for different tasks.
25
104651
4290
01:49
In CG, computer graphics, it's not any news --
26
109528
3703
01:53
it's been done for a while, and that's how most of the movies are made.
27
113255
3431
01:56
But if you're trying to make a robot that's physically moving,
28
116710
3870
02:00
it's a completely new story.
29
120604
1576
02:02
It's a completely new paradigm.
30
122691
2029
02:06
But you've all done this.
31
126307
1793
02:09
Who hasn't made a paper airplane, paper boat, paper crane?
32
129434
5308
02:15
Origami is a versatile platform for designers.
33
135893
3844
02:19
From a single sheet of paper, you can make multiple shapes,
34
139761
4076
02:23
and if you don't like it, you unfold and fold back again.
35
143861
3394
02:27
Any 3D form can be made from 2D surfaces by folding,
36
147947
6016
02:33
and this is proven mathematically.
37
153987
2529
02:38
And imagine if you were to have an intelligent sheet
38
158555
4408
02:42
that can self-fold into any form it wants,
39
162987
3559
02:46
anytime.
40
166570
1223
02:48
And that's what I've been working on.
41
168322
2254
02:50
I call this robotic origami,
42
170600
3182
02:53
"robogami."
43
173806
1498
02:57
This is our first robogami transformation
44
177387
3520
03:00
that was made by me about 10 years ago.
45
180931
3309
03:04
From a flat-sheeted robot,
46
184264
1767
03:06
it turns into a pyramid and back into a flat sheet
47
186055
2947
03:09
and into a space shuttle.
48
189026
2773
03:12
Quite cute.
49
192822
1241
03:14
Ten years later, with my group of ninja origami robotic researchers --
50
194789
6870
03:21
about 22 of them right now --
51
201683
1898
03:24
we have a new generation of robogamis,
52
204332
3408
03:27
and they're a little more effective and they do more than that.
53
207764
3311
03:32
So the new generation of robogamis actually serve a purpose.
54
212105
3283
03:35
For example, this one actually navigates through different terrains autonomously.
55
215412
5199
03:40
So when it's a dry and flat land, it crawls.
56
220635
3335
03:46
And if it meets sudden rough terrain,
57
226256
2479
03:48
it starts rolling.
58
228759
1237
03:50
It does this -- it's the same robot --
59
230020
2452
03:52
but depending on which terrain it meets,
60
232496
3023
03:55
it activates a different sequence of actuators that's on board.
61
235543
4765
04:02
And once it meets an obstacle, it jumps over it.
62
242459
3280
04:07
It does this by storing energy in each of its legs
63
247485
3376
04:10
and releasing it and catapulting like a slingshot.
64
250885
3876
04:14
And it even does gymnastics.
65
254785
1887
04:17
Yay.
66
257688
1171
04:18
(Laughter)
67
258883
1437
04:20
So I just showed you what a single robogami can do.
68
260828
4177
04:25
Imagine what they can do as a group.
69
265029
2785
04:27
They can join forces to tackle more complex tasks.
70
267838
4022
04:31
Each module, either active or passive,
71
271884
3162
04:35
we can assemble them to create different shapes.
72
275070
3567
04:38
Not only that, by controlling the folding joints,
73
278661
2749
04:41
we're able to create and attack different tasks.
74
281434
4548
04:46
The form is making new task space.
75
286006
3175
04:49
And this time, what's most important is the assembly.
76
289736
3839
04:54
They need to autonomously find each other in a different space,
77
294192
4184
04:58
attach and detach, depending on the environment and task.
78
298400
4290
05:03
And we can do this now.
79
303616
2059
05:06
So what's next?
80
306412
1427
05:07
Our imagination.
81
307863
1310
05:09
This is a simulation of what you can achieve
82
309704
2401
05:12
with this type of module.
83
312129
1657
05:13
We decided that we were going to have a four-legged crawler
84
313810
3172
05:18
turn into a little dog and make small gaits.
85
318870
3179
05:22
With the same module, we can actually make it do something else:
86
322073
3834
05:25
a manipulator, a typical, classical robotic task.
87
325931
3418
05:29
So with a manipulator, it can pick up an object.
88
329373
2699
05:32
Of course, you can add more modules to make the manipulator legs longer
89
332096
4044
05:36
to attack or pick up objects that are bigger or smaller,
90
336164
3723
05:39
or even have a third arm.
91
339911
1766
05:43
For robogamis, there's no one fixed shape nor task.
92
343545
4361
05:48
They can transform into anything, anywhere, anytime.
93
348628
4300
05:54
So how do you make them?
94
354408
2676
05:57
The biggest technical challenge of robogami is keeping them super thin,
95
357108
5317
06:02
flexible,
96
362449
1151
06:03
but still remaining functional.
97
363624
2214
06:06
They're composed of multiple layers of circuits, motors,
98
366562
3812
06:10
microcontrollers and sensors,
99
370398
2418
06:12
all in the single body,
100
372840
1854
06:14
and when you control individual folding joints,
101
374718
3321
06:18
you'll be able to achieve soft motions like that
102
378063
3526
06:21
upon your command.
103
381613
1480
06:26
Instead of being a single robot that is specifically made for a single task,
104
386013
4859
06:30
robogamis are optimized to do multi-tasks.
105
390896
3895
06:35
And this is quite important
106
395366
1757
06:37
for the difficult and unique environments on the Earth
107
397147
3831
06:41
as well as in space.
108
401002
2965
06:45
Space is a perfect environment for robogamis.
109
405782
3174
06:49
You cannot afford to have one robot for one task.
110
409673
4133
06:54
Who knows how many tasks you will encounter in space?
111
414966
3336
06:58
What you want is a single robotic platform that can transform to do multi-tasks.
112
418846
6959
07:07
What we want is a deck of thin robogami modules
113
427188
5112
07:12
that can transform to do multiples of performing tasks.
114
432324
4371
07:18
And don't take my word for it,
115
438322
3210
07:21
because the European Space Agency and Swiss Space Center
116
441556
3036
07:24
are sponsoring this exact concept.
117
444616
2260
07:27
So here you see a couple of images of reconfiguration of robogamis,
118
447562
5045
07:32
exploring the foreign land aboveground, on the surface,
119
452631
3778
07:36
as well as digging into the surface.
120
456433
1922
07:39
It's not just exploration.
121
459117
2079
07:41
For astronauts, they need additional help,
122
461220
2588
07:43
because you cannot afford to bring interns up there, either.
123
463832
2834
07:46
(Laughter)
124
466690
1182
07:48
They have to do every tedious task.
125
468357
2881
07:51
They may be simple,
126
471262
1218
07:52
but super interactive.
127
472504
1611
07:54
So you need robots to facilitate their experiments,
128
474762
3412
07:58
assisting them with the communications
129
478198
2311
08:00
and just docking onto surfaces to be their third arm holding different tools.
130
480533
5252
08:07
But how will they be able to control robogamis, for example,
131
487245
3133
08:10
outside the space station?
132
490402
1553
08:11
In this case, I show a robogami that is holding space debris.
133
491979
4037
08:16
You can work with your vision so that you can control them,
134
496040
3604
08:19
but what would be better is having the sensation of touch
135
499668
4462
08:24
directly transported onto the hands of the astronauts.
136
504154
3603
08:28
And what you need is a haptic device,
137
508248
2423
08:30
a haptic interface that recreates the sensation of touch.
138
510695
3708
08:35
And using robogamis, we can do this.
139
515051
2524
08:39
This is the world's smallest haptic interface
140
519276
4148
08:44
that can recreate a sensation of touch just underneath your fingertip.
141
524316
5189
08:50
We do this by moving the robogami
142
530104
2577
08:52
by microscopic and macroscopic movements at the stage.
143
532705
4511
08:57
And by having this, not only will you be able to feel
144
537812
3630
09:01
how big the object is,
145
541466
1739
09:03
the roundness and the lines,
146
543229
2886
09:06
but also the stiffness and the texture.
147
546139
3563
09:11
Alex has this interface just underneath his thumb,
148
551019
3989
09:15
and if he were to use this with VR goggles and hand controllers,
149
555032
4812
09:19
now the virtual reality is no longer virtual.
150
559868
3629
09:23
It becomes a tangible reality.
151
563521
2268
09:28
The blue ball, red ball and black ball that he's looking at
152
568529
3687
09:32
is no longer differentiated by colors.
153
572240
2466
09:34
Now it is a rubber blue ball, sponge red ball and billiard black ball.
154
574730
5311
09:40
This is now possible.
155
580642
1710
09:43
Let me show you.
156
583263
1202
09:46
This is really the first time this is shown live
157
586150
4169
09:50
in front of a public grand audience,
158
590343
2846
09:53
so hopefully this works.
159
593213
1828
09:55
So what you see here is an atlas of anatomy
160
595668
4301
09:59
and the robogami haptic interface.
161
599993
2793
10:02
So, like all the other reconfigurable robots,
162
602810
2545
10:05
it multitasks.
163
605379
1175
10:06
Not only is it going to serve as a mouse,
164
606578
2023
10:08
but also a haptic interface.
165
608625
2365
10:11
So for example, we have a white background where there is no object.
166
611381
3774
10:15
That means there is nothing to feel,
167
615179
2052
10:17
so we can have a very, very flexible interface.
168
617255
3712
10:21
Now, I use this as a mouse to approach skin,
169
621352
3152
10:24
a muscular arm,
170
624528
1263
10:25
so now let's feel his biceps,
171
625815
2016
10:27
or shoulders.
172
627855
1488
10:29
So now you see how much stiffer it becomes.
173
629367
3010
10:32
Let's explore even more.
174
632401
1371
10:33
Let's approach the ribcage.
175
633796
2910
10:36
And as soon as I move on top of the ribcage
176
636730
2562
10:39
and between the intercostal muscles,
177
639316
2285
10:41
which is softer and harder,
178
641625
1549
10:43
I can feel the difference of the stiffness.
179
643198
2236
10:45
Take my word for it.
180
645458
1326
10:46
So now you see, it's much stiffer in terms of the force
181
646808
4203
10:51
it's giving back to my fingertip.
182
651035
1845
10:53
So I showed you the surfaces that aren't moving.
183
653822
3978
10:57
How about if I were to approach something that moves,
184
657824
3553
11:01
for example, like a beating heart?
185
661401
2067
11:03
What would I feel?
186
663492
1395
11:11
(Applause)
187
671573
6146
11:19
This can be your beating heart.
188
679158
2293
11:22
This can actually be inside your pocket
189
682010
3610
11:25
while you're shopping online.
190
685644
1765
11:28
Now you'll be able to feel the difference of the sweater that you're buying,
191
688361
3789
11:32
how soft it is,
192
692174
1222
11:33
if it's actually cashmere or not,
193
693420
2836
11:36
or the bagel that you're trying to buy,
194
696280
2056
11:38
how hard it is or how crispy it is.
195
698360
3068
11:42
This is now possible.
196
702229
1736
11:46
The robotics technology is advancing to be more personalized and adaptive,
197
706774
5812
11:52
to adapt to our everyday needs.
198
712610
3104
11:56
This unique specie of reconfigurable robotics
199
716457
3625
12:00
is actually the platform to provide this invisible, intuitive interface
200
720106
5978
12:06
to meet our exact needs.
201
726108
2492
12:10
These robots will no longer look like the characters from the movies.
202
730169
4006
12:14
Instead, they will be whatever you want them to be.
203
734843
3848
12:19
Thank you.
204
739223
1206
12:20
(Applause)
205
740453
3598
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