TED2012
Vijay Kumar: Robots that fly ... and cooperate
Roboti koji lete ... i surađuju
Filmed:
Readability: 4.5
5,188,706 views
U svom laboratoriju na Sveučilištu Pennsylvania, Vijay Kumar sa svojim timom gradi leteće kvadratore (quadrotor), male, okretne i samosvjesne robote, koji tvore formacije -- za sklapanje konstrukcija, izviđanje ruševina i mnoge druge zadatke.
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
Dobro jutro.
00:22
I'm here today to talk
1
7000
2000
Danas ću govoriti
00:24
about autonomous, flying beach balls.
2
9000
3000
o samoupravljajućim, letećim... loptama za plažu.
00:27
No, agile aerial robots like this one.
3
12000
4000
Ma ne, o okretnim lebdećim robotima poput ovog.
00:31
I'd like to tell you a little bit about the challenges in building these
4
16000
3000
Nekoliko riječi o izazovima pri građenju robota
00:34
and some of the terrific opportunities
5
19000
2000
kao i mogućnostima koje pruža
00:36
for applying this technology.
6
21000
2000
primjena ove tehnologije.
00:38
So these robots
7
23000
2000
Dakle, ovi roboti
00:40
are related to unmanned aerial vehicles.
8
25000
3000
su svojevrsne bezpilotne letjelice.
00:43
However, the vehicles you see here are big.
9
28000
3000
Međutim ovdje prikazane letjelice su velike.
00:46
They weigh thousands of pounds,
10
31000
2000
Budući da teže stotine kilograma、
00:48
are not by any means agile.
11
33000
2000
pokretljivost im je otežana.
00:50
They're not even autonomous.
12
35000
2000
Štoviše nisu niti samoupravljajuće.
00:52
In fact, many of these vehicles
13
37000
2000
U stvarnosti, mnoge od ovih letjelica
00:54
are operated by flight crews
14
39000
2000
su upravljane stručnim osobljem
00:56
that can include multiple pilots,
15
41000
3000
poput grupe pilota
00:59
operators of sensors
16
44000
2000
kontrolorima senzora
01:01
and mission coordinators.
17
46000
2000
i zapovjednicima misije.
01:03
What we're interested in is developing robots like this --
18
48000
2000
Mi smo se pozabavili razvojem robota poput ovih --
01:05
and here are two other pictures --
19
50000
2000
prikazana su dva primjera
01:07
of robots that you can buy off the shelf.
20
52000
3000
robota koji se mogu kupiti u trgovini.
01:10
So these are helicopters with four rotors
21
55000
3000
Dakle to su helikopteri sa četiri rotora-propelera
01:13
and they're roughly a meter or so in scale
22
58000
4000
otprilike promjera jednog metra
01:17
and weigh several pounds.
23
62000
2000
i teže koji kilogram.
01:19
And so we retrofit these with sensors and processors,
24
64000
3000
Opremili smo ih sa senzorima i procesorima,
01:22
and these robots can fly indoors
25
67000
2000
da bi mogli letjeti u zatvorenom prostoru
01:24
without GPS.
26
69000
2000
bez GPSa.
01:26
The robot I'm holding in my hand
27
71000
2000
Robot kojeg držim
01:28
is this one,
28
73000
2000
...ovaj...
01:30
and it's been created by two students,
29
75000
3000
su kreirala dva studenta.
01:33
Alex and Daniel.
30
78000
2000
Alex i Daniel.
01:35
So this weighs a little more
31
80000
2000
Težak je jedva
01:37
than a tenth of a pound.
32
82000
2000
40-tak grama
01:39
It consumes about 15 watts of power.
33
84000
2000
i pokreće ga snaga od 15 wati.
01:41
And as you can see,
34
86000
2000
Kao što možete ocijeniti
01:43
it's about eight inches in diameter.
35
88000
2000
u promjeru je oko 20 cm.
01:45
So let me give you just a very quick tutorial
36
90000
3000
Dopustite da ukratko pojasnim
01:48
on how these robots work.
37
93000
2000
na koji način ovi roboti rade.
01:50
So it has four rotors.
38
95000
2000
Robot ima četiri propelera.
01:52
If you spin these rotors at the same speed,
39
97000
2000
Ako se svi propeleri okreću istom brzinom
01:54
the robot hovers.
40
99000
2000
robot lebdi u zraku.
01:56
If you increase the speed of each of these rotors,
41
101000
3000
Kada povećamo okretaje propelera
01:59
then the robot flies up, it accelerates up.
42
104000
3000
robot se uzdiže.
02:02
Of course, if the robot were tilted,
43
107000
2000
Ako je robot nagnut
02:04
inclined to the horizontal,
44
109000
2000
prema horizontali
02:06
then it would accelerate in this direction.
45
111000
3000
onda će se kretati u tom smjeru.
02:09
So to get it to tilt, there's one of two ways of doing it.
46
114000
3000
Evo jednog od dva načina kako nakrenuti robota.
02:12
So in this picture
47
117000
2000
Na ovoj slici
02:14
you see that rotor four is spinning faster
48
119000
2000
uočite kako se 4. propeler okreće brže,
02:16
and rotor two is spinning slower.
49
121000
2000
a 2. propeler se okreće sporije.
02:18
And when that happens
50
123000
2000
U tom trenutku
02:20
there's moment that causes this robot to roll.
51
125000
3000
robot se nakrene.
02:23
And the other way around,
52
128000
2000
I drugi način je
02:25
if you increase the speed of rotor three
53
130000
3000
da povećate brzinu 3. propelera
02:28
and decrease the speed of rotor one,
54
133000
2000
i smanjite brzinu 1. propelera,
02:30
then the robot pitches forward.
55
135000
3000
čime je robot usmjeren prema naprijed.
02:33
And then finally,
56
138000
2000
Konačno
02:35
if you spin opposite pairs of rotors
57
140000
2000
ako ubrzate dva
02:37
faster than the other pair,
58
142000
2000
nasuprotna propelera
02:39
then the robot yaws about the vertical axis.
59
144000
2000
robot će se okretati oko vertikalne osi.
02:41
So an on-board processor
60
146000
2000
U osnovi, ugrađeni procesor
02:43
essentially looks at what motions need to be executed
61
148000
3000
opaža koje kretnje trebaju biti izvršene,
02:46
and combines these motions
62
151000
2000
usklađuje kretnje
02:48
and figures out what commands to send to the motors
63
153000
3000
te šalje primjerenu naredbu motorima
02:51
600 times a second.
64
156000
2000
600 puta u sekundi.
02:53
That's basically how this thing operates.
65
158000
2000
Tako u osnovi robot radi.
02:55
So one of the advantages of this design
66
160000
2000
Prednost ovakvog dizajna
02:57
is, when you scale things down,
67
162000
2000
je, da s manjim komponentama
02:59
the robot naturally becomes agile.
68
164000
3000
robot postaje okretniji.
03:02
So here R
69
167000
2000
Na ovoj slici R predstavlja
03:04
is the characteristic length of the robot.
70
169000
2000
karakterističnu duljinu robota.
03:06
It's actually half the diameter.
71
171000
3000
U osnovi to je radius robota.
03:09
And there are lots of physical parameters that change
72
174000
3000
Mnogi fizikalni parametri se mijenjaju
03:12
as you reduce R.
73
177000
2000
kako smanjujemo R.
03:14
The one that's the most important
74
179000
2000
Najvažniji parametar
03:16
is the inertia or the resistance to motion.
75
181000
2000
je tromost, ili otpor kretanju.
03:18
So it turns out,
76
183000
2000
Ispada da je
03:20
the inertia, which governs angular motion,
77
185000
3000
tromost, o kojoj ovisi kretanje pod kutom,
03:23
scales as a fifth power of R.
78
188000
3000
proporcionalna R na petu potenciju.
03:26
So the smaller you make R,
79
191000
2000
Sa smanjivanjem R
03:28
the more dramatically the inertia reduces.
80
193000
3000
tromost dramatično pada.
03:31
So as a result, the angular acceleration,
81
196000
3000
To rezultira kutnom akceleracijom
03:34
denoted by Greek letter alpha here,
82
199000
2000
označenom grčkim slovom alfa
03:36
goes as one over R.
83
201000
2000
približno 1/R.
03:38
It's inversely proportional to R.
84
203000
2000
Dakle, obrnuto proporcionalna s R.
03:40
The smaller you make it the more quickly you can turn.
85
205000
3000
Što je robot manji, to je okretniji.
03:43
So this should be clear in these videos.
86
208000
2000
Što je očito iz ovih snimki.
03:45
At the bottom right you see a robot
87
210000
3000
Promotrite robota, desno dolje,
03:48
performing a 360 degree flip
88
213000
2000
kako se okrenuo 360 stupnjeva.
03:50
in less than half a second.
89
215000
2000
u djeliću sekunde.
03:52
Multiple flips, a little more time.
90
217000
3000
Višestruki okreti - malo više vremena.
03:55
So here the processes on board
91
220000
2000
Ugrađeni procesori
03:57
are getting feedback from accelerometers
92
222000
2000
zaprimaju informaciju od ugrađenih
03:59
and gyros on board
93
224000
2000
akcelerometara i žiroskopa
04:01
and calculating, like I said before,
94
226000
2000
te procesiraju, kao što rekoh,
04:03
commands at 600 times a second
95
228000
2000
te informacije 600 puta u sekundi
04:05
to stabilize this robot.
96
230000
2000
stabilizirajući robota.
04:07
So on the left, you see Daniel throwing this robot up into the air.
97
232000
3000
Na dnu lijevo, uočite Daniela kako baca robota u zrak
04:10
And it shows you how robust the control is.
98
235000
2000
demonstrirajući koliko je robot robustan.
04:12
No matter how you throw it,
99
237000
2000
Bez obzira kako ga bacite
04:14
the robot recovers and comes back to him.
100
239000
4000
robot se dočeka na noge i doleti nazad.
04:18
So why build robots like this?
101
243000
2000
Koja je svrha graditi ovakve robote?
04:20
Well robots like this have many applications.
102
245000
3000
Može biti višestruka.
04:23
You can send them inside buildings like this
103
248000
3000
Robote možete poslati u zgrade pod ugrozom
04:26
as first responders to look for intruders,
104
251000
3000
primjerice kao izvidnicu za otkrivanje provalnika,
04:29
maybe look for biochemical leaks,
105
254000
3000
otkrivanje biokemijskog hazarda,
04:32
gaseous leaks.
106
257000
2000
ili curenja plina.
04:34
You can also use them
107
259000
2000
Možete ih koristiti i kao
04:36
for applications like construction.
108
261000
2000
programirane graditelje.
04:38
So here are robots carrying beams, columns
109
263000
4000
Uočite kako roboti nose gredice i stupiće
04:42
and assembling cube-like structures.
110
267000
3000
i sklapaju ih u kockaste konstrukcije.
04:45
I'll tell you a little bit more about this.
111
270000
3000
O tome malo više kasnije.
04:48
The robots can be used for transporting cargo.
112
273000
3000
Ove robote možemo koristiti i za prijenos tereta.
04:51
So one of the problems with these small robots
113
276000
3000
Međutim, s obzirom na njihovu veličinu
04:54
is their payload carrying capacity.
114
279000
2000
kapacitet nosivosti je malen.
04:56
So you might want to have multiple robots
115
281000
2000
Stoga bi bilo dobro udružiti robote
04:58
carry payloads.
116
283000
2000
za prijenos tereta.
05:00
This is a picture of a recent experiment we did --
117
285000
2000
Evo snimke nedavnog pokusa --
05:02
actually not so recent anymore --
118
287000
2000
doduše već ima tome --
05:04
in Sendai shortly after the earthquake.
119
289000
3000
u Sendaiu, netom iza potresa.
05:07
So robots like this could be sent into collapsed buildings
120
292000
3000
Roboti mogu biti poslani u ruševine
05:10
to assess the damage after natural disasters,
121
295000
2000
kako bi se procjenila šteta
05:12
or sent into reactor buildings
122
297000
3000
ili u objekte s reaktorima
05:15
to map radiation levels.
123
300000
3000
da izmjere razinu radijacije.
05:19
So one fundamental problem
124
304000
2000
Osnovni problem
05:21
that the robots have to solve if they're to be autonomous
125
306000
3000
s kojim se roboti moraju nositi, samoinicijativno
05:24
is essentially figuring out
126
309000
2000
je u biti zaključivanje
05:26
how to get from point A to point B.
127
311000
2000
kako doći od točke A do točke B.
05:28
So this gets a little challenging
128
313000
2000
Dakle stvari postaju složenije
05:30
because the dynamics of this robot are quite complicated.
129
315000
3000
s porašću dinamike robota.
05:33
In fact, they live in a 12-dimensional space.
130
318000
2000
U načelu, roboti žive u 12-dimenzionalnom prostoru.
05:35
So we use a little trick.
131
320000
2000
Pa smo morali tome doskočiti.
05:37
We take this curved 12-dimensional space
132
322000
3000
Uzeli smo zakrivljeni 12-dimenzionalni prostor
05:40
and transform it
133
325000
2000
i transformirali ga
05:42
into a flat four-dimensional space.
134
327000
2000
u ravninski 4-dimenzionalni prostor.
05:44
And that four-dimensional space
135
329000
2000
Ovaj 4-dimenzionalni prostor
05:46
consists of X, Y, Z and then the yaw angle.
136
331000
3000
se sastoji od X, Y, Z i osnog kuta.
05:49
And so what the robot does
137
334000
2000
Što je robotu činiti
05:51
is it plans what we call a minimum snap trajectory.
138
336000
4000
je isplanirati putanju najmanje akceleracije.
05:55
So to remind you of physics,
139
340000
2000
Da se prisjetimo fizike,
05:57
you have position, derivative, velocity,
140
342000
2000
imamo položaj, te 1. derivaciju položaja - brzinu,
05:59
then acceleration,
141
344000
2000
zatim akceleraciju - 2. vremenska derivacija položaja (d2x/dt2)
06:01
and then comes jerk
142
346000
2000
impuls - 3. derivacija položaja (d3x/dt3)
06:03
and then comes snap.
143
348000
2000
'cimanje' - 4. derivacija položaja (d4x/dt4).
06:05
So this robot minimizes snap.
144
350000
3000
Robot minimizira cimanje.
06:08
So what that effectively does
145
353000
2000
Rezultat toga je
06:10
is produces a smooth and graceful motion.
146
355000
2000
glatko, ravnomjerno kretanje robota.
06:12
And it does that avoiding obstacles.
147
357000
3000
I to pri zaobilaženju zapreka.
06:15
So these minimum snap trajectories in this flat space
148
360000
3000
Dakle, putanje najmanje akceleracije u ovom ravninskom prostoru
06:18
are then transformed back
149
363000
2000
potom transformiramo nazad
06:20
into this complicated 12-dimensional space,
150
365000
2000
u složeni 12-dimenzionalni prostor,
06:22
which the robot must do
151
367000
2000
koje robot mora slijediti
06:24
for control and then execution.
152
369000
2000
radi provjere i potom izvršavanja.
06:26
So let me show you some examples
153
371000
2000
Evo nekoliko primjera
06:28
of what these minimum snap trajectories look like.
154
373000
2000
kako putanje najmanje akceleracije izgledaju.
06:30
And in the first video,
155
375000
2000
U prvoj snimci
06:32
you'll see the robot going from point A to point B
156
377000
2000
vidjet ćete kretanje robota od točke A do točke B
06:34
through an intermediate point.
157
379000
2000
ali kroz zadanu točku.
06:42
So the robot is obviously capable
158
387000
2000
Robot je sposoban
06:44
of executing any curve trajectory.
159
389000
2000
pratiti zakrivljene putanje.
06:46
So these are circular trajectories
160
391000
2000
Ovo su kružne putanje
06:48
where the robot pulls about two G's.
161
393000
3000
na kojima robot potegne 2G.
06:52
Here you have overhead motion capture cameras on the top
162
397000
4000
Ovdje imamo stropne kamere
06:56
that tell the robot where it is 100 times a second.
163
401000
3000
koje signaliziraju robotu njegov položaj 100 puta u sekundi.
06:59
It also tells the robot where these obstacles are.
164
404000
3000
Također daju robotu do znanja gdje su prepreke.
07:02
And the obstacles can be moving.
165
407000
2000
Te prepreke mogu biti pokretne.
07:04
And here you'll see Daniel throw this hoop into the air,
166
409000
3000
Ovdje vidite Daniela kako baca kolut u zrak,
07:07
while the robot is calculating the position of the hoop
167
412000
2000
a robot računa položaj koluta
07:09
and trying to figure out how to best go through the hoop.
168
414000
4000
pokušavajući odrediti pravi trenutak za prolazak kroz kolut.
07:13
So as an academic,
169
418000
2000
Kao znanstvenici, primorani smo izvoditi
07:15
we're always trained to be able to jump through hoops to raise funding for our labs,
170
420000
3000
akrobacije kako bi fondovi za naše projekte bili odobreni,
07:18
and we get our robots to do that.
171
423000
3000
pa smo odlučili i robote naučiti akrobacijama.
07:21
(Applause)
172
426000
6000
(Pljesak)
07:27
So another thing the robot can do
173
432000
2000
Dodatna stvar koju robot može učiniti
07:29
is it remembers pieces of trajectory
174
434000
3000
je memoriranje dijelova putanje
07:32
that it learns or is pre-programmed.
175
437000
2000
koju je savladao ili mu je predprogramirana.
07:34
So here you see the robot
176
439000
2000
Ovdje vidite robota
07:36
combining a motion
177
441000
2000
kako kombinira kretnje
07:38
that builds up momentum
178
443000
2000
i vreba pravi trenutak
07:40
and then changes its orientation and then recovers.
179
445000
3000
u kojem će se nagnuti i potom vratiti u početni položaj.
07:43
So it has to do this because this gap in the window
180
448000
3000
Na to je prisiljen jer svjetli otvor prozora
07:46
is only slightly larger than the width of the robot.
181
451000
4000
je tijesan za njegovu širinu.
07:50
So just like a diver stands on a springboard
182
455000
3000
Poput skakača s daske
07:53
and then jumps off it to gain momentum,
183
458000
2000
kada odskoči da bi dobio zamah,
07:55
and then does this pirouette, this two and a half somersault through
184
460000
3000
potreban za piruetu, i dvostruki salto
07:58
and then gracefully recovers,
185
463000
2000
da bi se na kraju ponovo ispružio,
08:00
this robot is basically doing that.
186
465000
2000
ovaj robot u osnovi čini isto.
08:02
So it knows how to combine little bits and pieces of trajectories
187
467000
3000
Dakle zna kako usuglasiti više manjih putanja
08:05
to do these fairly difficult tasks.
188
470000
4000
u jednu znatno složeniju.
08:09
So I want change gears.
189
474000
2000
Promjenimo temu.
08:11
So one of the disadvantages of these small robots is its size.
190
476000
3000
Jedan od nedostataka robota je veličina.
08:14
And I told you earlier
191
479000
2000
I kao što rekoh
08:16
that we may want to employ lots and lots of robots
192
481000
2000
trebat ćemo upregnuti veliki broj robota
08:18
to overcome the limitations of size.
193
483000
3000
da bi nadomjestili taj nedostatak.
08:21
So one difficulty
194
486000
2000
Prva poteškoća je
08:23
is how do you coordinate lots of these robots?
195
488000
3000
kako koordinirati jato robota?
08:26
And so here we looked to nature.
196
491000
2000
Okrenuli smo se majci prirodi.
08:28
So I want to show you a clip
197
493000
2000
Pogledajmo slijedeći prilog
08:30
of Aphaenogaster desert ants
198
495000
2000
s pustinjskim mravima Aphaenogaster.
08:32
in Professor Stephen Pratt's lab carrying an object.
199
497000
3000
kako nose teret, u laboratoriju profesora Stephen Pratta.
08:35
So this is actually a piece of fig.
200
500000
2000
Ovo je komadić smokve.
08:37
Actually you take any object coated with fig juice
201
502000
2000
Ako im podmetnete bilo koji predmet preliven smokvinim sokom
08:39
and the ants will carry them back to the nest.
202
504000
3000
mravi će ga odvući u mravinjak.
08:42
So these ants don't have any central coordinator.
203
507000
3000
Mravi očito nemaju koordinatora.
08:45
They sense their neighbors.
204
510000
2000
Ali osjećaju ostale iz skupine.
08:47
There's no explicit communication.
205
512000
2000
Ne postoji izravna komunikacija.
08:49
But because they sense the neighbors
206
514000
2000
Ali kako osjećaju ostale u skupini
08:51
and because they sense the object,
207
516000
2000
i kako osjećaju predmet,
08:53
they have implicit coordination across the group.
208
518000
3000
koordinacija je ostvarena posredno.
08:56
So this is the kind of coordination
209
521000
2000
Takvu sličnu koordinaciju
08:58
we want our robots to have.
210
523000
3000
smo željeli ostvariti kod robota.
09:01
So when we have a robot
211
526000
2000
Ako imamo robota
09:03
which is surrounded by neighbors --
212
528000
2000
grupiranog u jato
09:05
and let's look at robot I and robot J --
213
530000
2000
promotrimo robota I i robota J --
09:07
what we want the robots to do
214
532000
2000
želimo robote osvijestiti
09:09
is to monitor the separation between them
215
534000
3000
da paze na međusobni razmak
09:12
as they fly in formation.
216
537000
2000
kada tvore formaciju.
09:14
And then you want to make sure
217
539000
2000
Želimo da se taj razmak
09:16
that this separation is within acceptable levels.
218
541000
2000
održava unutar određene tolerancije.
09:18
So again the robots monitor this error
219
543000
3000
Da ponovimo, roboti prate ovu toleranciju
09:21
and calculate the control commands
220
546000
2000
i proračunavaju kontrolne naredbe
09:23
100 times a second,
221
548000
2000
100 puta u sekundi,
09:25
which then translates to the motor commands 600 times a second.
222
550000
3000
što znači da motor zaprima 600 naredbi u sekundi.
09:28
So this also has to be done
223
553000
2000
I to treba biti učinjeno
09:30
in a decentralized way.
224
555000
2000
decentralizirano.
09:32
Again, if you have lots and lots of robots,
225
557000
2000
Ako imate jato robota,
09:34
it's impossible to coordinate all this information centrally
226
559000
4000
bilo bi nemoguće koordinirati ih centralno
09:38
fast enough in order for the robots to accomplish the task.
227
563000
3000
zadovoljavajućom brzinom za ostvarenje zadaće.
09:41
Plus the robots have to base their actions
228
566000
2000
Dodatno, roboti moraju zasnivati kretnje
09:43
only on local information,
229
568000
2000
isključivo na lokalnim informacijama
09:45
what they sense from their neighbors.
230
570000
2000
koje zaprimaju od susjednih robota.
09:47
And then finally,
231
572000
2000
I napokon
09:49
we insist that the robots be agnostic
232
574000
2000
inzistirali smo da su roboti agnostični
09:51
to who their neighbors are.
233
576000
2000
prema susjedima.
09:53
So this is what we call anonymity.
234
578000
3000
To smo nazvali anonimnost.
09:56
So what I want to show you next
235
581000
2000
Sada ću vam pokazati
09:58
is a video
236
583000
2000
snimak
10:00
of 20 of these little robots
237
585000
3000
sa 20 robota
10:03
flying in formation.
238
588000
2000
koji lete u formaciji.
10:05
They're monitoring their neighbors' position.
239
590000
3000
Svaki robot promatra položaj susjeda.
10:08
They're maintaining formation.
240
593000
2000
I tako održavaju formaciju.
10:10
The formations can change.
241
595000
2000
Formacija se može mijenjati.
10:12
They can be planar formations,
242
597000
2000
Može biti ravninska
10:14
they can be three-dimensional formations.
243
599000
2000
ili prostorna.
10:16
As you can see here,
244
601000
2000
Kao što ovdje vidite,
10:18
they collapse from a three-dimensional formation into planar formation.
245
603000
3000
roboti iz prostorne formacije tvore ravninsku.
10:21
And to fly through obstacles
246
606000
2000
I da bi letjeli kroz prepreke
10:23
they can adapt the formations on the fly.
247
608000
4000
sposobni su prilagoditi formaciju u letu.
10:27
So again, these robots come really close together.
248
612000
3000
Roboti mogu prići blizu jedan drugom.
10:30
As you can see in this figure-eight flight,
249
615000
2000
Kao što vidite u ovoj formaciji u obliku osmice,
10:32
they come within inches of each other.
250
617000
2000
roboti prilaze jedni drugima unutar par centimetara.
10:34
And despite the aerodynamic interactions
251
619000
3000
I unatoč aerodinamičnom utjecaju
10:37
of these propeller blades,
252
622000
2000
propelerskih elisa
10:39
they're able to maintain stable flight.
253
624000
2000
održavaju stabilan let.
10:41
(Applause)
254
626000
7000
(Pljesak)
10:48
So once you know how to fly in formation,
255
633000
2000
Kada ih jednom naučite letjeti u formaciji
10:50
you can actually pick up objects cooperatively.
256
635000
2000
možete podići predmete surađujući.
10:52
So this just shows
257
637000
2000
To ujedno znači
10:54
that we can double, triple, quadruple
258
639000
3000
da možemo udvostručiti, utrostručiti ili učetverostručiti
10:57
the robot strength
259
642000
2000
snagu robota
10:59
by just getting them to team with neighbors, as you can see here.
260
644000
2000
usklađujući njihovo djelovanje, kao što je prikazano.
11:01
One of the disadvantages of doing that
261
646000
3000
Jedan od nedostataka ovakvog pristupa
11:04
is, as you scale things up --
262
649000
2000
je, kako teret postaje teži...
11:06
so if you have lots of robots carrying the same thing,
263
651000
2000
i ako uposlite jato robota za prijenos tereta
11:08
you're essentially effectively increasing the inertia,
264
653000
3000
u osnovi povećavate tromost,
11:11
and therefore you pay a price; they're not as agile.
265
656000
3000
čime naravno jato postaje manje okretno.
11:14
But you do gain in terms of payload carrying capacity.
266
659000
3000
Ali tako možete prenašati teže objekte.
11:17
Another application I want to show you --
267
662000
2000
Još jedna primjena koju želim pokazati --
11:19
again, this is in our lab.
268
664000
2000
ponovo, u našem laboratoriju.
11:21
This is work done by Quentin Lindsey who's a graduate student.
269
666000
2000
Ovo je napravio naš postdiplomac Quentin Lindsey.
11:23
So his algorithm essentially tells these robots
270
668000
3000
Njegov algoritam u osnovi naređuje robotima
11:26
how to autonomously build
271
671000
2000
kako da samostalno sklope
11:28
cubic structures
272
673000
2000
prostorne rešetkaste konstrukcije
11:30
from truss-like elements.
273
675000
3000
iz gredica.
11:33
So his algorithm tells the robot
274
678000
2000
Njegov algoritam govori robotu
11:35
what part to pick up,
275
680000
2000
koji dio podići,
11:37
when and where to place it.
276
682000
2000
te kada i gdje ga ugraditi.
11:39
So in this video you see --
277
684000
2000
Na ovoj snimci možete vidjeti --
11:41
and it's sped up 10, 14 times --
278
686000
2000
ubrzano 10 - 14 puta --
11:43
you see three different structures being built by these robots.
279
688000
3000
tri različite konstrukcije sklapane robotima.
11:46
And again, everything is autonomous,
280
691000
2000
Roboti su samoupravljajući,
11:48
and all Quentin has to do
281
693000
2000
i sve što Quentin treba učiniti
11:50
is to get them a blueprint
282
695000
2000
je dostaviti im nacrt
11:52
of the design that he wants to build.
283
697000
4000
konstrukcije koja treba biti sklopljena.
11:56
So all these experiments you've seen thus far,
284
701000
3000
Dakle, svi ovi pokusi koje ste vidjeli,
11:59
all these demonstrations,
285
704000
2000
sve ove demonstracije,
12:01
have been done with the help of motion capture systems.
286
706000
3000
su ostvarene uz pomoć sustava za praćenje kretnji.
12:04
So what happens when you leave your lab
287
709000
2000
Međutim, što se događa kada napustite laboratorij
12:06
and you go outside into the real world?
288
711000
3000
i otisnete se u stvaran svijet?
12:09
And what if there's no GPS?
289
714000
3000
I što kada nema GPSa?
12:12
So this robot
290
717000
2000
Primjerice, ovaj robot
12:14
is actually equipped with a camera
291
719000
2000
je opremljen kamerom
12:16
and a laser rangefinder, laser scanner.
292
721000
3000
i laserskim daljinomjerom, laserskim skenerom.
12:19
And it uses these sensors
293
724000
2000
I koristi ove senzore
12:21
to build a map of the environment.
294
726000
2000
kako bi izgradio kartu okruženja.
12:23
What that map consists of are features --
295
728000
3000
Takva karta sadrži objekte --
12:26
like doorways, windows,
296
731000
2000
poput vrata, prozora,
12:28
people, furniture --
297
733000
2000
ljudi, namještaja --
12:30
and it then figures out where its position is
298
735000
2000
i onda robot procjenjuje gdje se nalazi
12:32
with respect to the features.
299
737000
2000
u odnosu na te objekte.
12:34
So there is no global coordinate system.
300
739000
2000
Sve bez globalnog koordinatnog sustava (GPS).
12:36
The coordinate system is defined based on the robot,
301
741000
3000
Koordinatni sustav je definiran robotom,
12:39
where it is and what it's looking at.
302
744000
3000
njegovim položajem, i time što promatra.
12:42
And it navigates with respect to those features.
303
747000
3000
I navođenje se odvija u odnosu na te objekte.
12:45
So I want to show you a clip
304
750000
2000
Evo isječka
12:47
of algorithms developed by Frank Shen
305
752000
2000
s algoritmima koje je razvio Frank Shen
12:49
and Professor Nathan Michael
306
754000
2000
i profesor Nathan Michael
12:51
that shows this robot entering a building for the very first time
307
756000
4000
koji prikazuje robota kako ulazi u zgradu po prvi puta
12:55
and creating this map on the fly.
308
760000
3000
i stvara kartu u letu.
12:58
So the robot then figures out what the features are.
309
763000
3000
Robot razaznaje objekte i
13:01
It builds the map.
310
766000
2000
iscrtava kartu (model).
13:03
It figures out where it is with respect to the features
311
768000
2000
Izračunavajući udaljenosti do objekata
13:05
and then estimates its position
312
770000
2000
robot određuje svoj položaj
13:07
100 times a second
313
772000
2000
100 puta u sekundi
13:09
allowing us to use the control algorithms
314
774000
2000
uz korištenje kontrolnog algoritma
13:11
that I described to you earlier.
315
776000
2000
koji sam opisao ranije.
13:13
So this robot is actually being commanded
316
778000
2000
Dakle ovaj robot zaprima
13:15
remotely by Frank.
317
780000
2000
Frankove naredbe sa udaljene lokacije.
13:17
But the robot can also figure out
318
782000
2000
Ali robot također može
13:19
where to go on its own.
319
784000
2000
samostalno procijeniti gdje treba ići.
13:21
So suppose I were to send this into a building
320
786000
2000
Recimo da želim poslati robota u zgradu
13:23
and I had no idea what this building looked like,
321
788000
2000
za koju nemam predodžbu o unutrašnjosti.
13:25
I can ask this robot to go in,
322
790000
2000
Jednostavno pošaljem robota da
13:27
create a map
323
792000
2000
kreira kartu (model)
13:29
and then come back and tell me what the building looks like.
324
794000
3000
i pozovem ga nazad da mi predoči što je zabilježio.
13:32
So here, the robot is not only solving the problem,
325
797000
3000
Dakle, robot ne samo da rješava problem
13:35
how to go from point A to point B in this map,
326
800000
3000
kako stići od točke A do točke B na ovoj karti,
13:38
but it's figuring out
327
803000
2000
nego i procjenjuje
13:40
what the best point B is at every time.
328
805000
2000
koja je optimalna točka B.
13:42
So essentially it knows where to go
329
807000
3000
U osnovi, robot zna da treba ići
13:45
to look for places that have the least information.
330
810000
2000
prema lokacijama o kojima ima najmanje informacija.
13:47
And that's how it populates this map.
331
812000
3000
I na taj način iscrtava kartu (model).
13:50
So I want to leave you
332
815000
2000
I da zaključim
13:52
with one last application.
333
817000
2000
s još jednom primjenom robota.
13:54
And there are many applications of this technology.
334
819000
3000
A postoje i mnoge druge primjene.
13:57
I'm a professor, and we're passionate about education.
335
822000
2000
Ja sam profesor, i kao takav strastven za edukaciju.
13:59
Robots like this can really change the way
336
824000
2000
Ovakvi roboti mogu promijeniti
14:01
we do K through 12 education.
337
826000
2000
način na koji se odvija nastava.
14:03
But we're in Southern California,
338
828000
2000
Ali s obzirom da smo u Južnoj Kaliforniji,
14:05
close to Los Angeles,
339
830000
2000
blizu Los Angelesa,
14:07
so I have to conclude
340
832000
2000
morat ću zaključiti
14:09
with something focused on entertainment.
341
834000
2000
u holivudskom tonu.
14:11
I want to conclude with a music video.
342
836000
2000
Neka to bude glazbeni video.
14:13
I want to introduce the creators, Alex and Daniel,
343
838000
3000
Želim predstaviti Alexa i Daniela,
14:16
who created this video.
344
841000
2000
kreatore videa.
14:18
(Applause)
345
843000
7000
(Pljesak)
14:25
So before I play this video,
346
850000
2000
No prije nego prikažemo video,
14:27
I want to tell you that they created it in the last three days
347
852000
3000
želim napomenuti da su ga snimili u zadnja tri dana
14:30
after getting a call from Chris.
348
855000
2000
nakon poziva od Chrisa (TED kuratora).
14:32
And the robots that play the video
349
857000
2000
I roboti koji sviraju na snimci
14:34
are completely autonomous.
350
859000
2000
su potpuno samostalni.
14:36
You will see nine robots play six different instruments.
351
861000
3000
Vidjet ćete 9 robota kako svira 6 instrumenata.
14:39
And of course, it's made exclusively for TED 2012.
352
864000
4000
Naravno, eksluzivno za TED 2012.
14:43
Let's watch.
353
868000
3000
Pogledajmo.
15:19
(Music)
354
904000
10000
(Glazba)
16:23
(Applause)
355
968000
17000
(Pljesak)
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