Flocking together: the physics of sheep herding and pedestrian flows

1
00:00:04,719 --> 00:00:07,040
Hello, and welcome to the Physics World Stories

2
00:00:07,040 --> 00:00:09,859
podcast. I'm Andre Glesser. And in this episode,

3
00:00:09,919 --> 00:00:12,660
we're gonna be exploring the physics of crowds.

4
00:00:13,814 --> 00:00:14,314
Specifically,

5
00:00:14,695 --> 00:00:17,254
or at least to begin with, looking at

6
00:00:17,254 --> 00:00:19,355
the physics of sheep movement.

7
00:00:19,814 --> 00:00:22,295
And whilst this is a slightly wooly topic,

8
00:00:22,295 --> 00:00:24,875
and you will find the odd sheep based

9
00:00:24,934 --> 00:00:25,914
joke in here,

10
00:00:26,429 --> 00:00:29,410
there is also some really intriguing physics

11
00:00:29,789 --> 00:00:31,949
not just a value to those who are

12
00:00:31,949 --> 00:00:35,070
animal herders but anyone interested in animal or

13
00:00:35,070 --> 00:00:36,130
human behavior

14
00:00:36,670 --> 00:00:38,829
later in the podcast we'll hear from someone

15
00:00:38,829 --> 00:00:41,825
who won for their work on human crowds

16
00:00:41,885 --> 00:00:43,265
the ignobel prize

17
00:00:43,645 --> 00:00:46,365
the prize given for research that makes you

18
00:00:46,365 --> 00:00:49,325
laugh and then think but on the physics

19
00:00:49,325 --> 00:00:51,804
world website you'll find an article by Philip

20
00:00:51,804 --> 00:00:55,265
Ball a science writer based in London entitled

21
00:00:55,760 --> 00:00:58,560
field work, the physics of sheep from phase

22
00:00:58,560 --> 00:00:59,060
transitions

23
00:00:59,520 --> 00:01:00,820
to collective motion.

24
00:01:01,280 --> 00:01:03,119
And through a window out of which I

25
00:01:03,119 --> 00:01:06,579
often gaze, there is a flock of sheep.

26
00:01:06,799 --> 00:01:07,780
I must admit

27
00:01:08,125 --> 00:01:10,125
at no point when I've looked at these

28
00:01:10,125 --> 00:01:12,765
sheep have I thought of physics. So I

29
00:01:12,765 --> 00:01:13,265
wondered

30
00:01:13,724 --> 00:01:16,784
quite why it had occurred to Philip Paul.

31
00:01:16,844 --> 00:01:19,004
So gazing out of that same window at

32
00:01:19,004 --> 00:01:20,765
that flock of sheep, I gave him a

33
00:01:20,765 --> 00:01:21,265
call.

34
00:01:22,420 --> 00:01:25,380
I've had a long interest in how ideas

35
00:01:25,380 --> 00:01:26,920
from physics can be,

36
00:01:27,859 --> 00:01:29,319
can be used to understand

37
00:01:29,780 --> 00:01:32,040
systems that don't sound at all like physics.

38
00:01:32,100 --> 00:01:32,600
And

39
00:01:32,980 --> 00:01:35,640
animal motion is one of them, actually, including

40
00:01:35,780 --> 00:01:37,000
the motion of people.

41
00:01:38,105 --> 00:01:40,584
So actually, years ago, 20 years ago, I

42
00:01:40,584 --> 00:01:41,564
wrote a book about,

43
00:01:42,025 --> 00:01:44,025
the this topic of how we could use

44
00:01:44,025 --> 00:01:45,724
ideas from physics to understand

45
00:01:46,105 --> 00:01:47,645
aspects of human society.

46
00:01:48,825 --> 00:01:50,025
And, you know, as I say in the

47
00:01:50,025 --> 00:01:52,444
piece, compared to trying to do that,

48
00:01:53,149 --> 00:01:55,729
trying to understand sheep might seem like,

49
00:01:56,030 --> 00:01:58,129
you know, an easy thing to do because,

50
00:01:58,670 --> 00:02:01,329
they have somewhat less, sophistication

51
00:02:02,030 --> 00:02:03,329
than, than humans.

52
00:02:04,759 --> 00:02:05,259
But

53
00:02:06,435 --> 00:02:09,235
it it's actually it's a very interesting problem

54
00:02:09,235 --> 00:02:11,555
to understand how sheep herd, but also there's

55
00:02:11,555 --> 00:02:14,435
this added twist that sheep are herded by

56
00:02:14,435 --> 00:02:14,935
dogs.

57
00:02:15,235 --> 00:02:17,495
You know? And that seems like a completely

58
00:02:17,794 --> 00:02:18,534
kind of

59
00:02:19,319 --> 00:02:22,040
idiosyncratic thing that is miles away from anything

60
00:02:22,040 --> 00:02:23,879
physics could handle, but it turns out that

61
00:02:23,879 --> 00:02:24,540
it isn't.

62
00:02:25,319 --> 00:02:27,719
And these models that are used, they're ones

63
00:02:27,719 --> 00:02:28,699
that have,

64
00:02:29,159 --> 00:02:31,580
that sort of have their roots in work

65
00:02:31,639 --> 00:02:32,139
on

66
00:02:32,504 --> 00:02:34,604
particularly on more obviously

67
00:02:34,985 --> 00:02:38,025
sort of collective flocking motions like, flocking of

68
00:02:38,025 --> 00:02:40,985
birds, like the schooling of fish. Things that,

69
00:02:40,985 --> 00:02:42,604
you know, when we look at them,

70
00:02:43,544 --> 00:02:45,724
there seems we we just kind of intuit

71
00:02:45,784 --> 00:02:47,324
that there has to be some

72
00:02:48,030 --> 00:02:50,669
deeper principle going on that gives them this

73
00:02:50,669 --> 00:02:51,569
kind of coherence,

74
00:02:52,750 --> 00:02:54,990
that, you know, doesn't seem to come from

75
00:02:54,990 --> 00:02:55,729
from individual

76
00:02:56,110 --> 00:02:56,610
animals.

77
00:02:57,229 --> 00:02:59,389
And in fact, I'd seen before I even

78
00:02:59,389 --> 00:03:01,250
encountered these papers, I'd seen,

79
00:03:01,790 --> 00:03:04,014
just, you know, video footage on on the

80
00:03:04,014 --> 00:03:04,915
web of

81
00:03:05,294 --> 00:03:05,794
sheep

82
00:03:06,495 --> 00:03:08,814
fluids. So it's got the kind of these

83
00:03:08,814 --> 00:03:11,474
aerial photos, probably drone photos of sheep,

84
00:03:12,415 --> 00:03:15,294
her you know, wandering around in fields, speeded

85
00:03:15,294 --> 00:03:15,794
up

86
00:03:16,349 --> 00:03:18,669
and kind of going through gates and, you

87
00:03:18,669 --> 00:03:20,990
know, going around obstacles and so on. And

88
00:03:20,990 --> 00:03:23,150
it's really weird to see it because it

89
00:03:23,150 --> 00:03:25,710
really does look like the flow of some

90
00:03:25,710 --> 00:03:28,289
kind of, you know, slightly granular fluid.

91
00:03:28,784 --> 00:03:30,944
So there too, if we see it if

92
00:03:30,944 --> 00:03:32,064
we look at it on the right sort

93
00:03:32,064 --> 00:03:34,004
of time scale and from the right perspective,

94
00:03:34,544 --> 00:03:36,625
you know, again, there's this sort of intuitive

95
00:03:36,625 --> 00:03:39,584
feeling that there's something there's some grand principle,

96
00:03:39,584 --> 00:03:42,539
there's some physics in this system. That's what

97
00:03:42,539 --> 00:03:44,620
this work that I was writing about is

98
00:03:44,620 --> 00:03:46,539
trying to explore. So I say I'm looking

99
00:03:46,539 --> 00:03:48,939
out the window, and there are a flock

100
00:03:48,939 --> 00:03:50,800
of sheep. You know, there's walls,

101
00:03:51,259 --> 00:03:51,759
there's

102
00:03:52,139 --> 00:03:52,639
hedges,

103
00:03:53,235 --> 00:03:56,594
There's occasionally the odd sheepdog. And at this

104
00:03:56,594 --> 00:03:59,235
time of year, when the tourists have gone,

105
00:03:59,235 --> 00:04:00,614
the sheep are sort of,

106
00:04:01,155 --> 00:04:03,235
yeah, just left to their own devices. When

107
00:04:03,235 --> 00:04:05,074
there are tourists here walking their dogs, the

108
00:04:05,074 --> 00:04:07,129
sheep are scared all over the place. But

109
00:04:07,129 --> 00:04:09,770
it's it's I've to be perfectly honest, as

110
00:04:09,770 --> 00:04:11,370
a bit of a nerd, I have enjoyed

111
00:04:11,370 --> 00:04:12,270
watching the sheep,

112
00:04:13,129 --> 00:04:14,810
more at this time of year where they're

113
00:04:14,810 --> 00:04:16,569
just left to their own devices, as I

114
00:04:16,569 --> 00:04:18,410
say, and they have this kind of well,

115
00:04:18,410 --> 00:04:20,175
yeah, they are a flock, aren't they?

116
00:04:20,654 --> 00:04:23,134
Yeah. It is. And, you know, what so

117
00:04:23,134 --> 00:04:24,814
so what does that mean? It, you know,

118
00:04:24,814 --> 00:04:26,754
it means there's a a load of them,

119
00:04:26,814 --> 00:04:28,654
but it means something more than that. It

120
00:04:28,654 --> 00:04:30,194
means that they are

121
00:04:30,495 --> 00:04:33,314
somehow interacting with an with each other

122
00:04:33,740 --> 00:04:35,199
to stay coherent.

123
00:04:36,139 --> 00:04:37,759
And that's something that

124
00:04:38,060 --> 00:04:39,199
pretty much all,

125
00:04:40,060 --> 00:04:43,100
groups of animals do or groups of living

126
00:04:43,100 --> 00:04:45,740
organisms. Actually, even cells do that. You can

127
00:04:45,740 --> 00:04:47,580
see it with bacteria. There are some people

128
00:04:47,580 --> 00:04:49,754
who develop models like this to try to

129
00:04:49,754 --> 00:04:52,235
understand bacteria and how they sort of flow

130
00:04:52,235 --> 00:04:54,314
around, and they too can kind of sense

131
00:04:54,314 --> 00:04:55,295
each other's presence.

132
00:04:56,634 --> 00:04:59,295
And, you know, I I I kind of

133
00:04:59,675 --> 00:05:01,435
we we if we think of how we

134
00:05:01,435 --> 00:05:03,660
move around in space, we're clearly doing that

135
00:05:03,660 --> 00:05:04,939
as well. We like to think, you know,

136
00:05:04,939 --> 00:05:07,259
we're all individuals. But if we're moving down

137
00:05:07,259 --> 00:05:09,500
a crowded pavement or something, then we are

138
00:05:09,500 --> 00:05:12,779
interacting with others. We're certainly, we're trying to

139
00:05:12,779 --> 00:05:14,939
avoid collisions if we're not staring at our

140
00:05:14,939 --> 00:05:16,879
phones. We're trying to avoid bumping into,

141
00:05:17,180 --> 00:05:17,840
each other.

142
00:05:18,375 --> 00:05:21,014
But also, occasionally, we might be moving with

143
00:05:21,014 --> 00:05:23,014
a group of friends or with family. And

144
00:05:23,014 --> 00:05:25,574
so there's a kind of attraction, a kind

145
00:05:25,574 --> 00:05:27,814
of cohesion there. We're trying to to stick

146
00:05:27,814 --> 00:05:28,314
together.

147
00:05:28,694 --> 00:05:31,275
So, you know, there are these these principles

148
00:05:31,975 --> 00:05:32,399
that

149
00:05:32,879 --> 00:05:33,860
aren't totally

150
00:05:34,160 --> 00:05:36,259
unlike the way inanimate

151
00:05:36,560 --> 00:05:37,060
particles

152
00:05:37,600 --> 00:05:38,419
might interact

153
00:05:38,720 --> 00:05:40,259
through forces of attraction

154
00:05:40,720 --> 00:05:43,839
and repulsion, repulsion being this tendency to avoid

155
00:05:43,839 --> 00:05:44,339
collisions.

156
00:05:45,544 --> 00:05:47,305
And and this was the kind of idea

157
00:05:47,305 --> 00:05:49,944
that, you know, lies behind all of these

158
00:05:49,944 --> 00:05:52,285
attempts to try to understand group motion.

159
00:05:53,225 --> 00:05:55,625
Is it is it something that we can

160
00:05:55,625 --> 00:05:56,125
describe

161
00:05:56,745 --> 00:05:59,460
just in terms of forces of attraction and

162
00:05:59,460 --> 00:06:01,160
repulsion between the individuals,

163
00:06:02,020 --> 00:06:02,520
who

164
00:06:03,060 --> 00:06:04,980
each of which, you know, have their own

165
00:06:04,980 --> 00:06:06,980
kind of agenda to some extent, but it

166
00:06:06,980 --> 00:06:08,900
may be a fairly simple agenda. I mean,

167
00:06:08,900 --> 00:06:10,500
even for us, if we're walking down a

168
00:06:10,500 --> 00:06:12,819
pavement, we're generally trying to get from one

169
00:06:12,819 --> 00:06:15,214
place to another, you know, probably as as

170
00:06:15,214 --> 00:06:16,595
as quickly as possible

171
00:06:17,294 --> 00:06:18,274
while navigating

172
00:06:18,574 --> 00:06:20,834
obstacles and while, you know, avoiding

173
00:06:21,694 --> 00:06:22,194
collisions.

174
00:06:23,134 --> 00:06:23,634
So

175
00:06:24,334 --> 00:06:24,834
that's

176
00:06:25,454 --> 00:06:28,819
that's a, in principle, that's a fairly simple

177
00:06:28,819 --> 00:06:29,319
situation

178
00:06:29,699 --> 00:06:31,540
to to try to model and, you know,

179
00:06:31,540 --> 00:06:33,540
that's kind of what you're probably seeing sheep

180
00:06:33,540 --> 00:06:35,620
do, except that there's, you know, there's also

181
00:06:35,620 --> 00:06:37,220
something else that I imagine you're kind of

182
00:06:37,220 --> 00:06:38,819
seeing them doing, which is, you know, they're

183
00:06:38,819 --> 00:06:40,259
not just trying to get from a to

184
00:06:40,259 --> 00:06:42,185
b. On the whole, what they tend to

185
00:06:42,185 --> 00:06:44,685
be doing is is wanting to to eat.

186
00:06:45,225 --> 00:06:47,384
They spend their most of their lives doing

187
00:06:47,384 --> 00:06:48,225
that, you know,

188
00:06:48,665 --> 00:06:50,205
chewing on on grass.

189
00:06:51,305 --> 00:06:53,464
So, you know, that's another aspect of the

190
00:06:53,464 --> 00:06:56,759
problem that's specific to grazing animals that they're

191
00:06:56,759 --> 00:06:58,759
not just trying to move. They will move

192
00:06:58,759 --> 00:07:00,759
occasionally, and in particular, they'll move to try

193
00:07:00,759 --> 00:07:03,259
to find more food, more grass.

194
00:07:04,279 --> 00:07:06,839
But there's that, there there's that sort of

195
00:07:06,839 --> 00:07:08,860
complex, you know, interplay between

196
00:07:09,444 --> 00:07:12,165
just trying to, you know, be left alone

197
00:07:12,165 --> 00:07:13,464
to graze, to eat,

198
00:07:13,845 --> 00:07:15,785
and moving around together.

199
00:07:16,165 --> 00:07:16,665
And

200
00:07:17,045 --> 00:07:18,345
with sheep in particular,

201
00:07:19,285 --> 00:07:20,745
as you say, they're nervous

202
00:07:21,045 --> 00:07:23,444
of other animals, particularly of dogs, but also

203
00:07:23,444 --> 00:07:26,189
of humans. So there's also that aspect. They're

204
00:07:26,189 --> 00:07:28,269
kind of looking out for what they might,

205
00:07:28,269 --> 00:07:29,810
you know, imagine are predators.

206
00:07:30,829 --> 00:07:32,849
So those are the kind of components

207
00:07:33,149 --> 00:07:35,310
of the behavior that need to go into

208
00:07:35,310 --> 00:07:36,849
trying to model this situation.

209
00:07:37,149 --> 00:07:38,370
As we've been talking,

210
00:07:38,735 --> 00:07:40,355
the sheep have gone from

211
00:07:41,055 --> 00:07:41,954
lying down

212
00:07:42,415 --> 00:07:42,915
behind

213
00:07:43,295 --> 00:07:45,454
a sort of hillock on the hillside, if

214
00:07:45,454 --> 00:07:46,814
you see what I mean. It's it's it's

215
00:07:46,814 --> 00:07:48,334
a a mound on the hillside, which is

216
00:07:48,334 --> 00:07:50,654
protecting them from the wind. That's why they're

217
00:07:50,654 --> 00:07:52,649
all there as far as I can see.

218
00:07:52,729 --> 00:07:54,329
Then the sun's come out, and they've all

219
00:07:54,329 --> 00:07:56,970
spread out across the field. Physics doesn't explain

220
00:07:56,970 --> 00:07:58,169
that, does it? I mean, apart from the

221
00:07:58,169 --> 00:07:59,149
sun coming out.

222
00:08:00,729 --> 00:08:02,509
Well, it it it

223
00:08:03,209 --> 00:08:04,910
I could see how it could

224
00:08:05,235 --> 00:08:07,394
in that, you know, if you were wanting

225
00:08:07,394 --> 00:08:07,894
to,

226
00:08:08,995 --> 00:08:11,254
to to include in a model like this,

227
00:08:11,714 --> 00:08:14,615
their wish to be in sunshine,

228
00:08:15,475 --> 00:08:15,975
then

229
00:08:16,354 --> 00:08:18,274
the you know, there's an attraction there. There's

230
00:08:18,274 --> 00:08:21,310
an attraction to a particular kind of, you

231
00:08:21,310 --> 00:08:22,990
know, you can model it as a as

232
00:08:22,990 --> 00:08:24,370
a force, as a field.

233
00:08:25,149 --> 00:08:25,889
I mean,

234
00:08:26,269 --> 00:08:27,970
there's literally a field as well.

235
00:08:28,350 --> 00:08:30,509
So you can see how you can, you

236
00:08:30,509 --> 00:08:33,070
know, build aspects like that into it. The

237
00:08:33,070 --> 00:08:35,549
other, interesting thing about sheep and about, you

238
00:08:35,549 --> 00:08:38,654
know, larger animals like that, which actually doesn't

239
00:08:38,654 --> 00:08:40,575
apply so much to us. You know, sheep

240
00:08:40,575 --> 00:08:43,534
are elongated. So from above, we're kind of,

241
00:08:43,534 --> 00:08:46,014
you know, circular blobs, and you can model

242
00:08:46,014 --> 00:08:47,554
us as circular particles.

243
00:08:47,934 --> 00:08:50,115
Sheep are kind of more like ellipsoidal

244
00:08:50,654 --> 00:08:51,154
particles.

245
00:08:52,629 --> 00:08:54,330
And so there's an orientational

246
00:08:54,710 --> 00:08:56,710
aspect to that. It's a bit like the

247
00:08:56,710 --> 00:08:57,850
contrast between,

248
00:08:58,470 --> 00:09:01,350
you know, simple atoms or or, you know,

249
00:09:01,350 --> 00:09:03,929
globular molecules and liquid crystals,

250
00:09:04,575 --> 00:09:07,134
which also have that tendency to kind of

251
00:09:07,134 --> 00:09:09,695
align, to respond to one another's presence by

252
00:09:09,695 --> 00:09:11,475
sort of aligning their their axis.

253
00:09:12,095 --> 00:09:14,335
And so, you know, that's a question we

254
00:09:14,335 --> 00:09:17,295
might ask and and the researchers have asked

255
00:09:17,295 --> 00:09:18,274
about sheep.

256
00:09:18,659 --> 00:09:21,379
Is there a a tendency to align? And

257
00:09:21,379 --> 00:09:23,940
if there is, then that looks a little

258
00:09:23,940 --> 00:09:24,919
bit like

259
00:09:25,299 --> 00:09:27,240
the way magnetic spins

260
00:09:27,700 --> 00:09:29,320
align in magnetic materials,

261
00:09:30,019 --> 00:09:32,100
and you can build that force into it

262
00:09:32,100 --> 00:09:33,764
as well. A force that,

263
00:09:34,965 --> 00:09:37,285
will make one sheep have a kind of

264
00:09:37,285 --> 00:09:39,545
tendency to align its body,

265
00:09:40,404 --> 00:09:42,884
with the direction of what its neighbors are

266
00:09:42,884 --> 00:09:43,945
doing. Why?

267
00:09:44,565 --> 00:09:46,745
Well, I I I guess I mean, certainly,

268
00:09:46,804 --> 00:09:49,009
if if you're if we're thinking about the

269
00:09:49,009 --> 00:09:50,389
way sheep move,

270
00:09:51,649 --> 00:09:52,850
you know, they seem to

271
00:09:53,490 --> 00:09:54,230
the the flocks,

272
00:09:54,769 --> 00:09:57,409
seem to occasionally to sort of wander as

273
00:09:57,409 --> 00:10:00,049
groups in the same direction. That's something that

274
00:10:00,049 --> 00:10:02,715
you just observe happening. So sheep, you know,

275
00:10:02,715 --> 00:10:04,735
have this tendency to follow one another.

276
00:10:05,595 --> 00:10:06,575
And if

277
00:10:07,514 --> 00:10:08,575
that's the case,

278
00:10:08,955 --> 00:10:10,955
they're clearly gonna have to be aligned in

279
00:10:10,955 --> 00:10:12,554
order to do that. So they're not, you

280
00:10:12,554 --> 00:10:14,475
know, literally, they're they're trying to align in

281
00:10:14,475 --> 00:10:15,535
the walking direction.

282
00:10:15,980 --> 00:10:18,539
So while they're moving, you know, they'll be

283
00:10:18,539 --> 00:10:20,779
trying to align. But also, you know, if

284
00:10:20,779 --> 00:10:22,000
they were in a dense

285
00:10:22,379 --> 00:10:24,299
flock and they were being, for example, if

286
00:10:24,299 --> 00:10:26,940
they were being shepherded by a dog, then

287
00:10:26,940 --> 00:10:29,259
again, they're they're one they're they're not gonna

288
00:10:29,259 --> 00:10:31,034
be wanting to bump into each other.

289
00:10:32,154 --> 00:10:34,475
So, you know, there's probably there too going

290
00:10:34,475 --> 00:10:36,475
to be a tendency to kind of line

291
00:10:36,475 --> 00:10:37,534
themselves up.

292
00:10:38,394 --> 00:10:40,394
So, you know, that it's really,

293
00:10:40,954 --> 00:10:43,754
just either a matter of, doing that in

294
00:10:43,754 --> 00:10:47,370
order to, coordinate their motions or doing it

295
00:10:47,370 --> 00:10:50,090
so that they can pack more efficiently in

296
00:10:50,090 --> 00:10:50,990
a small space.

297
00:10:51,690 --> 00:10:53,769
I do there's a a behavior that I

298
00:10:53,769 --> 00:10:55,370
observe here. I don't know if you've seen

299
00:10:55,370 --> 00:10:56,509
an answer to this, but,

300
00:10:57,210 --> 00:10:59,550
very regularly in the summer months, there's,

301
00:11:00,674 --> 00:11:02,674
you know, a tourist dog who should be

302
00:11:02,674 --> 00:11:04,035
on a lead isn't on a lead, but

303
00:11:04,035 --> 00:11:06,294
we'll we'll sketch over that particular issue.

304
00:11:07,554 --> 00:11:09,394
But one of the sheep gets isolated from

305
00:11:09,394 --> 00:11:10,375
the rest of the flock.

306
00:11:10,914 --> 00:11:13,750
Then the dog goes away and that sheep

307
00:11:13,830 --> 00:11:17,190
the isolated sheep then calls out and tries

308
00:11:17,190 --> 00:11:18,710
to find its way back to the to

309
00:11:18,710 --> 00:11:20,710
the flock. It well, it seems that they

310
00:11:20,710 --> 00:11:22,230
do. I mean, a lot of animals do,

311
00:11:22,470 --> 00:11:25,110
particularly if there's a predator around because it's

312
00:11:25,110 --> 00:11:26,870
the kind of safety in numbers thing, you

313
00:11:26,870 --> 00:11:29,324
know, that that actually a sheep on its

314
00:11:29,324 --> 00:11:31,964
own is gonna feel much more exposed than

315
00:11:31,964 --> 00:11:33,725
a sheep in a flock and particularly a

316
00:11:33,725 --> 00:11:35,884
sheep in the center of the flock. So,

317
00:11:35,884 --> 00:11:38,605
you know, there's clearly a, there's a sort

318
00:11:38,605 --> 00:11:40,524
of attraction if you like there for the

319
00:11:40,524 --> 00:11:42,444
sheep to a sort of cohesive force that

320
00:11:42,444 --> 00:11:44,464
will keep them together for safety.

321
00:11:44,980 --> 00:11:46,899
And, you know, more than that, that what

322
00:11:46,899 --> 00:11:48,340
you actually see in,

323
00:11:48,740 --> 00:11:51,220
sheep flocks and in other sometimes other groups

324
00:11:51,220 --> 00:11:53,559
of animals that are exposed to predators

325
00:11:54,100 --> 00:11:55,539
is they they all wanna be in the

326
00:11:55,539 --> 00:11:57,620
middle, which kind of makes sense, you know,

327
00:11:57,620 --> 00:11:58,360
that that's,

328
00:11:59,139 --> 00:12:00,519
that's the safest place.

329
00:12:00,875 --> 00:12:03,294
And so there's this kind of constant churn,

330
00:12:03,355 --> 00:12:04,495
you know, going on.

331
00:12:05,355 --> 00:12:06,475
And in fact, it's,

332
00:12:07,914 --> 00:12:11,034
some, biologists have, you know, talked about this

333
00:12:11,034 --> 00:12:13,375
in terms of this kind of selfish flock

334
00:12:13,620 --> 00:12:14,120
theory,

335
00:12:14,899 --> 00:12:16,340
where, you know, everyone wants to be in

336
00:12:16,340 --> 00:12:18,259
the middle. And so what what kind of

337
00:12:18,259 --> 00:12:18,759
dynamics,

338
00:12:19,299 --> 00:12:20,679
result from that?

339
00:12:21,379 --> 00:12:23,379
So, yeah, that that that, you know, if

340
00:12:23,379 --> 00:12:25,779
if if, an individual or small groups get

341
00:12:25,779 --> 00:12:28,504
isolated, then there is this sort of attraction

342
00:12:28,565 --> 00:12:30,184
to bring them back into the flock

343
00:12:30,805 --> 00:12:33,125
for the the the purposes of safety. In

344
00:12:33,125 --> 00:12:35,845
the selfish flock, my understanding is that at

345
00:12:35,845 --> 00:12:37,365
least it's the way it's talked about, I

346
00:12:37,365 --> 00:12:38,985
think, in the march of the penguins.

347
00:12:39,605 --> 00:12:42,004
They they take it in turns, and that's

348
00:12:42,004 --> 00:12:43,384
less selfish and more,

349
00:12:44,080 --> 00:12:46,240
altruistic, more part of the group kind of

350
00:12:46,240 --> 00:12:47,840
looking after each other. Is that just a

351
00:12:47,840 --> 00:12:49,279
kind of a nice way of looking at

352
00:12:49,279 --> 00:12:51,279
it? Well, that you know, that's a good

353
00:12:51,279 --> 00:12:53,299
question because it's quite possible,

354
00:12:53,919 --> 00:12:55,120
and I don't know if this is the

355
00:12:55,120 --> 00:12:58,245
case for penguins. I think it's more probably

356
00:12:58,304 --> 00:13:00,004
the case for for sheep, certainly.

357
00:13:00,384 --> 00:13:02,225
It's quite possible that if you have a

358
00:13:02,225 --> 00:13:05,684
model where there's the the the the individuals

359
00:13:05,824 --> 00:13:08,784
are moving around, they have this tendency to

360
00:13:08,784 --> 00:13:11,379
try to get into the center. That actually

361
00:13:11,379 --> 00:13:11,879
spontaneously

362
00:13:12,500 --> 00:13:14,419
out of that will come this kind of

363
00:13:14,419 --> 00:13:16,340
churn, this kind of turnover of who's on

364
00:13:16,340 --> 00:13:18,340
the edge. So it's not as though the

365
00:13:18,340 --> 00:13:20,100
individuals on the edge are kind of at

366
00:13:20,100 --> 00:13:21,480
any time consciously

367
00:13:21,779 --> 00:13:24,659
thinking, okay. I've had my turn now. You

368
00:13:24,659 --> 00:13:26,634
know, I've done my my my part to

369
00:13:26,634 --> 00:13:28,154
be on the edge now. Let me back

370
00:13:28,154 --> 00:13:29,995
in the middle. I don't think, it seem

371
00:13:30,074 --> 00:13:32,074
it seems unlikely that anything like that is

372
00:13:32,074 --> 00:13:34,334
going on. It seems more probable

373
00:13:34,794 --> 00:13:37,595
that something like this kind of behavior would

374
00:13:37,595 --> 00:13:38,654
emerge spontaneously

375
00:13:39,269 --> 00:13:42,070
simply from those movement and from those kind

376
00:13:42,070 --> 00:13:42,730
of psychological

377
00:13:43,430 --> 00:13:43,930
rules.

378
00:13:44,389 --> 00:13:45,990
And the other thing that seems to be

379
00:13:45,990 --> 00:13:48,330
spontaneous from what I can see is the

380
00:13:48,790 --> 00:13:50,870
who the leader is at any moment with

381
00:13:50,870 --> 00:13:52,470
the sheep in terms of the the leader

382
00:13:52,470 --> 00:13:53,930
of the movement, who takes

383
00:13:54,274 --> 00:13:56,355
this the flock in a particular direction or

384
00:13:56,355 --> 00:13:58,035
parts of the flock in a particular direction?

385
00:13:58,035 --> 00:14:00,295
Yeah. Sure. I mean, that that is something,

386
00:14:00,754 --> 00:14:03,075
that that is observed, and it's something that

387
00:14:03,075 --> 00:14:04,134
was looked at,

388
00:14:05,394 --> 00:14:07,139
in particular in one of the models that

389
00:14:07,139 --> 00:14:09,460
I discussed in the piece, work done by

390
00:14:09,460 --> 00:14:11,559
Fernando Peruana and colleagues,

391
00:14:12,019 --> 00:14:13,960
at one of the universities in Paris,

392
00:14:14,740 --> 00:14:17,620
where they they were watching. So they started

393
00:14:17,620 --> 00:14:20,019
off empirically. They were just watching small groups,

394
00:14:20,019 --> 00:14:23,044
sometimes really tiny groups of of sheep, you

395
00:14:23,044 --> 00:14:25,605
know, move around in a field grazing. And

396
00:14:25,605 --> 00:14:27,365
they found that there was this kind of

397
00:14:27,365 --> 00:14:27,865
tendency

398
00:14:28,804 --> 00:14:29,384
to occasionally

399
00:14:30,245 --> 00:14:32,404
develop to adopt to kind of, you know,

400
00:14:32,404 --> 00:14:34,404
moving mode where they all walk in some

401
00:14:34,404 --> 00:14:36,404
single file to a different part of the

402
00:14:36,404 --> 00:14:36,904
field.

403
00:14:37,730 --> 00:14:40,210
But what they observed was that there so

404
00:14:40,210 --> 00:14:41,429
there's a leader there,

405
00:14:41,970 --> 00:14:44,529
by Natesti if there's a line, but they

406
00:14:44,529 --> 00:14:46,850
they found that the sheep seem to take

407
00:14:46,850 --> 00:14:48,950
it in turns. So there's no,

408
00:14:49,649 --> 00:14:52,355
single leader who decides that in the flock

409
00:14:52,355 --> 00:14:53,095
each time.

410
00:14:53,475 --> 00:14:55,394
It seems to happen more or less at

411
00:14:55,394 --> 00:14:57,894
random. So pretty much each of the sheep

412
00:14:58,195 --> 00:14:59,095
gets a turn,

413
00:14:59,634 --> 00:15:00,934
at being the leader.

414
00:15:01,315 --> 00:15:03,394
And, again, you know, that sounds like that's

415
00:15:03,394 --> 00:15:04,774
all very kind of egalitarian

416
00:15:05,235 --> 00:15:07,620
and, you know, it's you could imagine the

417
00:15:07,620 --> 00:15:09,620
sheep saying, okay. Now it's my turn to

418
00:15:09,620 --> 00:15:11,700
be leader. But actually, you know, it seems

419
00:15:11,700 --> 00:15:13,220
like it doesn't happen or at least we

420
00:15:13,220 --> 00:15:15,059
don't need to assume that. If we simply

421
00:15:15,059 --> 00:15:15,559
assume

422
00:15:15,940 --> 00:15:17,079
that whoever,

423
00:15:18,579 --> 00:15:20,740
becomes the leader, that that is, you know,

424
00:15:20,740 --> 00:15:22,120
something that just happens

425
00:15:22,424 --> 00:15:24,684
randomly, that there is a particular chance

426
00:15:25,225 --> 00:15:28,184
at any moment that any individual will become

427
00:15:28,184 --> 00:15:30,184
the leader of one of those moving groups,

428
00:15:30,184 --> 00:15:31,325
that that's enough

429
00:15:31,705 --> 00:15:33,965
to account for for what we see.

430
00:15:34,759 --> 00:15:36,919
So, yeah, there there absolutely is. And the

431
00:15:37,080 --> 00:15:39,480
you you see the the advantage there's an

432
00:15:39,480 --> 00:15:42,200
advantage in in having that kind of, if

433
00:15:42,200 --> 00:15:43,980
you like, that kind of cheap algorithm,

434
00:15:44,360 --> 00:15:44,840
you know,

435
00:15:45,480 --> 00:15:47,340
inherent in in their behaviors,

436
00:15:48,404 --> 00:15:51,365
relative to their always being the same leader

437
00:15:51,365 --> 00:15:51,865
because,

438
00:15:52,725 --> 00:15:55,524
it means that there's more pooling of the

439
00:15:55,524 --> 00:15:56,024
information

440
00:15:56,485 --> 00:15:58,424
that all of the individuals have.

441
00:15:59,205 --> 00:16:00,965
So, you know, it may be that one

442
00:16:00,965 --> 00:16:02,664
of them knows or has has

443
00:16:02,965 --> 00:16:05,210
realized that one part of the field seems

444
00:16:05,210 --> 00:16:07,790
to be more attractive or less, you know,

445
00:16:08,169 --> 00:16:10,190
chewed on than another. And

446
00:16:11,210 --> 00:16:13,610
the ideal is that the flock has access

447
00:16:13,610 --> 00:16:15,629
to the information of all the individuals.

448
00:16:16,764 --> 00:16:18,605
You know, this is something that we might

449
00:16:18,605 --> 00:16:21,325
also see in in in flocking birds or

450
00:16:21,325 --> 00:16:23,245
in the other sort of flocking animals where

451
00:16:23,245 --> 00:16:24,764
it might be that 1 or 2 individuals

452
00:16:24,764 --> 00:16:27,165
spot a predator where the rest of the

453
00:16:27,165 --> 00:16:29,600
the group hasn't. And so you want to

454
00:16:29,600 --> 00:16:31,519
make use of that. You want to pool

455
00:16:31,519 --> 00:16:33,539
that information that everyone has.

456
00:16:33,919 --> 00:16:36,720
So this idea with sheep, this idea of

457
00:16:36,720 --> 00:16:39,679
constantly changing the the leader in these sort

458
00:16:39,679 --> 00:16:41,539
of episodes of walking,

459
00:16:44,325 --> 00:16:46,485
It's it's an it's an efficient way of

460
00:16:46,485 --> 00:16:46,985
working,

461
00:16:47,524 --> 00:16:49,784
in the sense that it's efficient in information

462
00:16:49,845 --> 00:16:52,644
sharing. I I can't help wondering if it's

463
00:16:52,644 --> 00:16:54,884
not that, efficient if you're trying to run

464
00:16:54,884 --> 00:16:56,565
a political party or a government to keep

465
00:16:56,565 --> 00:16:57,544
changing your computer.

466
00:16:59,199 --> 00:17:01,399
Well, you know, I mean, even there, that's

467
00:17:01,600 --> 00:17:04,160
although that's true, you know, there is a

468
00:17:04,160 --> 00:17:06,559
discussion about whether there are more sort of

469
00:17:06,559 --> 00:17:08,579
egalitarian, you know, whether it would be

470
00:17:08,880 --> 00:17:11,440
more sensible, for example, instead of having a

471
00:17:11,440 --> 00:17:14,275
House of Lords to have, you know, people

472
00:17:14,414 --> 00:17:14,914
randomly,

473
00:17:16,575 --> 00:17:18,515
you know, chosen from the population

474
00:17:18,974 --> 00:17:21,134
to lead a body like that and that

475
00:17:21,134 --> 00:17:23,855
there's a constant turnover there. And maybe that

476
00:17:23,855 --> 00:17:25,694
would be more efficient in terms of being

477
00:17:25,694 --> 00:17:28,029
able to make better use of the information

478
00:17:28,029 --> 00:17:30,349
that everyone has rather than that a few

479
00:17:30,349 --> 00:17:33,549
privileged individuals have for life. But wouldn't the

480
00:17:33,549 --> 00:17:35,710
concern there be that you could get someone

481
00:17:35,710 --> 00:17:37,950
that was sort of, you know, a a

482
00:17:37,950 --> 00:17:39,410
conspiracy theorist running

483
00:17:39,974 --> 00:17:42,055
an important part of the world. I mean,

484
00:17:42,055 --> 00:17:44,775
absolutely. You know, questions like that, they're they're

485
00:17:44,775 --> 00:17:47,575
they're complex ones, but I suppose the the

486
00:17:47,575 --> 00:17:49,734
the principle there is, you know, you could

487
00:17:49,734 --> 00:17:51,734
equally get them and you could equally get

488
00:17:51,734 --> 00:17:53,174
that in the house. And when you do

489
00:17:53,174 --> 00:17:54,750
get that in the house of lords now,

490
00:17:54,829 --> 00:17:56,210
and they're there for life.

491
00:17:57,150 --> 00:17:58,910
So, you know, as long as you have

492
00:17:58,910 --> 00:18:00,690
that constant sort of turnover,

493
00:18:01,630 --> 00:18:03,630
the the the idea would be that, you

494
00:18:03,630 --> 00:18:05,730
know, on average, you're going to,

495
00:18:06,190 --> 00:18:08,430
get a better sort of representation and a

496
00:18:08,430 --> 00:18:09,410
fairer representation

497
00:18:09,934 --> 00:18:11,694
of all the information and all the views

498
00:18:11,694 --> 00:18:14,174
in the population. There's that kind of angle

499
00:18:14,174 --> 00:18:16,115
to it, but I wonder, is this actually

500
00:18:16,174 --> 00:18:18,734
useful for the farmer who's herding the sheep

501
00:18:18,734 --> 00:18:20,255
out there? Well, I I think I mean,

502
00:18:20,255 --> 00:18:22,194
first of all, I guess the question is,

503
00:18:22,390 --> 00:18:25,269
can a model like this even reproduce what

504
00:18:25,269 --> 00:18:27,829
we see in terms of herding? And, you

505
00:18:27,829 --> 00:18:30,169
know, that's the thing I found quite remarkable

506
00:18:30,230 --> 00:18:31,130
that it can.

507
00:18:31,509 --> 00:18:33,130
That without any

508
00:18:33,990 --> 00:18:36,410
assumptions and here I'm talking about a model,

509
00:18:36,789 --> 00:18:38,294
that's being developed by,

510
00:18:38,855 --> 00:18:40,474
peep researchers at Harvard,

511
00:18:41,014 --> 00:18:43,434
and others to try to look at shepherding

512
00:18:43,494 --> 00:18:45,894
behavior. And, you know, here, as as I

513
00:18:45,894 --> 00:18:47,974
said initially, it just seems like, how on

514
00:18:47,974 --> 00:18:49,894
earth do you make a physics based model

515
00:18:49,894 --> 00:18:52,134
of a sheepdog actually trying to herd, you

516
00:18:52,134 --> 00:18:53,359
know, a

517
00:18:53,740 --> 00:18:55,819
flock of sheep. But it seems that you

518
00:18:55,819 --> 00:18:58,220
can, that it's an optimization problem is how

519
00:18:58,220 --> 00:19:00,059
they present it. So you're you know, the

520
00:19:00,059 --> 00:19:02,460
the the challenge is to try to get

521
00:19:02,460 --> 00:19:03,200
the flock

522
00:19:03,660 --> 00:19:06,835
from a to b as quickly as possible

523
00:19:06,894 --> 00:19:09,234
without losing any sheep on the way.

524
00:19:09,615 --> 00:19:10,115
So

525
00:19:10,494 --> 00:19:13,055
what what is the optimum strategy for a

526
00:19:13,055 --> 00:19:14,914
dog to, to undertake

527
00:19:15,455 --> 00:19:16,994
to encourage that motion?

528
00:19:17,295 --> 00:19:19,934
Assuming that there's this repulsion between the dog

529
00:19:19,934 --> 00:19:22,279
and the sheep so that that's what's going

530
00:19:22,279 --> 00:19:24,279
to kind of move the the flock along.

531
00:19:24,279 --> 00:19:26,140
And there's there's this attraction

532
00:19:26,519 --> 00:19:28,359
between the sheep themselves and that's what's going

533
00:19:28,359 --> 00:19:30,460
to get to keep the flock coherent. So

534
00:19:30,519 --> 00:19:32,680
given those constraints, you know, what's the most

535
00:19:32,680 --> 00:19:34,599
efficient way of doing it? And they come

536
00:19:34,599 --> 00:19:36,214
up with three possibilities

537
00:19:36,674 --> 00:19:37,734
that do the job.

538
00:19:38,434 --> 00:19:40,115
One is that the dog just sort of

539
00:19:40,115 --> 00:19:42,214
runs from side to side of the flock,

540
00:19:42,275 --> 00:19:44,775
keeping it, you know, in order but gradually

541
00:19:44,994 --> 00:19:46,994
kind of getting it to move, in a

542
00:19:46,994 --> 00:19:47,815
single direction.

543
00:19:48,299 --> 00:19:50,779
Another is that the dog runs in circles

544
00:19:50,779 --> 00:19:52,720
around the flock, but they're kind of corkscrewing

545
00:19:52,860 --> 00:19:55,200
circles so that it's gradually getting somewhere.

546
00:19:55,660 --> 00:19:58,059
Both of those strategies are ones that are

547
00:19:58,059 --> 00:20:00,000
seen, literally in the field.

548
00:20:00,634 --> 00:20:03,214
And then there's a third strategy, which isn't

549
00:20:03,274 --> 00:20:05,434
seen as far as the researchers are aware,

550
00:20:05,434 --> 00:20:07,274
which is that the dog actually it almost

551
00:20:07,274 --> 00:20:09,134
kind of burrows into the flock.

552
00:20:10,154 --> 00:20:12,634
But, because the flock is coherent, it doesn't

553
00:20:12,634 --> 00:20:15,079
just scatter. It sort of stays together, but,

554
00:20:15,220 --> 00:20:17,779
you know, it it sort of draws back

555
00:20:17,779 --> 00:20:19,700
from the dog within it. And so by

556
00:20:19,700 --> 00:20:21,859
the by the dog moving, the flock, it

557
00:20:21,859 --> 00:20:23,380
kind of moves with it as if the

558
00:20:23,380 --> 00:20:26,039
dog is like a driver in a car,

559
00:20:26,099 --> 00:20:28,724
and so they call this strategy driving. So

560
00:20:28,724 --> 00:20:30,484
it seems to be one way that the

561
00:20:30,484 --> 00:20:32,884
the sheepdog could work, but but doesn't. And

562
00:20:32,884 --> 00:20:33,944
it may be that

563
00:20:34,244 --> 00:20:36,964
the the flocks just aren't really sort of

564
00:20:36,964 --> 00:20:39,365
cohesive enough to make that a viable strategy.

565
00:20:39,365 --> 00:20:40,984
We we, you know, we don't know.

566
00:20:41,285 --> 00:20:43,144
But, you know, it does seem to capture

567
00:20:43,669 --> 00:20:46,789
what actual sheepdog do. And so, you know,

568
00:20:46,789 --> 00:20:48,549
that's a kind of it seems like that's

569
00:20:48,549 --> 00:20:50,630
a kind of validation of the model itself.

570
00:20:50,630 --> 00:20:52,470
And so if you've got something like that

571
00:20:52,470 --> 00:20:54,710
that works, then you can start to think,

572
00:20:54,710 --> 00:20:56,710
okay. What are there ways that,

573
00:20:57,190 --> 00:20:58,569
you know, are there better ways

574
00:20:58,954 --> 00:21:00,474
of of of doing this or are there

575
00:21:00,474 --> 00:21:02,875
ways that we could improve the efficiency with

576
00:21:02,875 --> 00:21:05,615
which the the dogs work? Or in particular,

577
00:21:06,634 --> 00:21:09,195
are there ways that we might automate this

578
00:21:09,195 --> 00:21:12,474
process using, for example, drones or using, you

579
00:21:12,474 --> 00:21:13,295
know, robotic,

580
00:21:14,769 --> 00:21:17,490
you know, entities to to do the herding

581
00:21:17,490 --> 00:21:17,990
instead.

582
00:21:18,289 --> 00:21:19,650
And I, you know, I I loved it

583
00:21:19,650 --> 00:21:19,970
that,

584
00:21:20,609 --> 00:21:22,609
what some of the researchers involved in this

585
00:21:22,609 --> 00:21:24,369
work said, well, if you if you speak

586
00:21:24,369 --> 00:21:27,009
to farmers about that possibility of using drones

587
00:21:27,009 --> 00:21:28,275
or whatever, they just laugh,

588
00:21:29,315 --> 00:21:32,115
because, you know, the technology just isn't ready.

589
00:21:32,115 --> 00:21:33,095
But in particular,

590
00:21:33,474 --> 00:21:35,714
it seems that sheepdog you know, sheepdog are

591
00:21:35,714 --> 00:21:36,775
really smart dogs.

592
00:21:37,315 --> 00:21:38,994
That's why they hear that they're they're so

593
00:21:38,994 --> 00:21:40,295
good at doing this job.

594
00:21:41,309 --> 00:21:43,490
So they have a kind of an instinct

595
00:21:43,549 --> 00:21:44,289
it seems

596
00:21:44,589 --> 00:21:47,250
to be able to respond on the fly

597
00:21:47,549 --> 00:21:49,069
to what the sheep are doing. If you've

598
00:21:49,069 --> 00:21:50,669
got one that wanders off, you know, the

599
00:21:50,669 --> 00:21:53,714
sheepdog can respond to it in ways that

600
00:21:54,034 --> 00:21:56,355
it's probably gonna be very hard to program

601
00:21:56,355 --> 00:21:57,894
into robots or drones.

602
00:21:58,355 --> 00:22:00,115
But, you know, if you're ever going to

603
00:22:00,115 --> 00:22:03,494
do that, then models like this are precisely

604
00:22:03,634 --> 00:22:05,234
the kind of thing you'd need to figure

605
00:22:05,234 --> 00:22:07,494
out what are the basic rules of behavior

606
00:22:07,950 --> 00:22:09,809
that a an automated

607
00:22:10,589 --> 00:22:11,809
shepherd or sheepdog,

608
00:22:12,589 --> 00:22:14,509
would have to follow in order to do

609
00:22:14,509 --> 00:22:17,390
the job efficiently and reliably. I can't help

610
00:22:17,390 --> 00:22:19,069
but wonder why on earth you'd want to

611
00:22:19,069 --> 00:22:21,595
replace sheepdogs, though. I mean, what would be

612
00:22:21,595 --> 00:22:23,275
the reason? Yeah. Well, they require a lot

613
00:22:23,275 --> 00:22:23,934
of training,

614
00:22:24,554 --> 00:22:27,275
and, you know, they if if and they

615
00:22:27,275 --> 00:22:28,255
have their limitations.

616
00:22:28,634 --> 00:22:30,154
They can only go at a certain you

617
00:22:30,154 --> 00:22:31,855
know, they can't go too fast.

618
00:22:34,230 --> 00:22:35,130
And, you know,

619
00:22:35,910 --> 00:22:37,609
it it could be that particularly

620
00:22:37,910 --> 00:22:38,490
if you're

621
00:22:38,789 --> 00:22:40,410
working with bigger flocks

622
00:22:40,789 --> 00:22:43,910
that, you know, sheepdogs would a single sheepdog

623
00:22:43,910 --> 00:22:46,170
would struggle to, to to contain,

624
00:22:46,505 --> 00:22:48,265
you know, might it be then that you'd

625
00:22:48,265 --> 00:22:50,845
have to use something more systematic,

626
00:22:51,945 --> 00:22:54,684
and with great a greater range of capabilities

627
00:22:55,065 --> 00:22:56,605
in order to do the job.

628
00:22:56,985 --> 00:22:59,005
So, you know, I suppose you could imagine

629
00:22:59,065 --> 00:23:00,869
situations where where that's,

630
00:23:01,910 --> 00:23:03,289
where where that's the case.

631
00:23:03,990 --> 00:23:07,029
You know, drones never tire, drones never need

632
00:23:07,029 --> 00:23:07,529
feeding.

633
00:23:08,869 --> 00:23:11,029
But, I mean, you know, I I suspect

634
00:23:11,029 --> 00:23:13,690
that behind your question is the same feeling

635
00:23:13,750 --> 00:23:16,309
for me that actually it's amazing and lovely

636
00:23:16,309 --> 00:23:18,654
that sheepdog are able to do this. And,

637
00:23:18,654 --> 00:23:20,335
you know, what a shame it would be

638
00:23:20,335 --> 00:23:22,835
if they were replaced by something automatic.

639
00:23:23,214 --> 00:23:24,674
I know. I think the only

640
00:23:24,974 --> 00:23:26,894
possible reason I can think for it is

641
00:23:26,894 --> 00:23:29,294
if you had a sort of robotic dog

642
00:23:29,294 --> 00:23:30,115
that had

643
00:23:30,690 --> 00:23:32,929
some form of artificial intelligence, and it worked

644
00:23:32,929 --> 00:23:33,669
with the sheep

645
00:23:34,130 --> 00:23:36,529
so much that eventually you could ask it

646
00:23:36,529 --> 00:23:38,869
whether androids dream of electric sheep.

647
00:23:39,730 --> 00:23:42,049
Brilliant. Of course, you would need to get

648
00:23:42,049 --> 00:23:43,894
that in there. I guess, you know, I

649
00:23:43,894 --> 00:23:45,494
I mean, I I think the other,

650
00:23:45,894 --> 00:23:47,654
answer to that might be, well, you know,

651
00:23:47,654 --> 00:23:49,095
it may be that you don't wanna do

652
00:23:49,095 --> 00:23:50,855
this with sheep, but it may be that

653
00:23:50,855 --> 00:23:53,194
you wanna do it with other systems, including

654
00:23:53,414 --> 00:23:53,914
maybe,

655
00:23:55,109 --> 00:23:57,990
you know, sort of nanoscopic inorganic systems where

656
00:23:57,990 --> 00:23:58,970
you've got a load

657
00:23:59,349 --> 00:24:02,009
of what are are called active active particles

658
00:24:02,069 --> 00:24:03,750
that can move under their own steam, and

659
00:24:03,750 --> 00:24:05,910
you want to get them somewhere and you

660
00:24:05,910 --> 00:24:08,774
want some other, you know, active particle or

661
00:24:08,774 --> 00:24:11,174
nanorobot or something to, you know, be able

662
00:24:11,174 --> 00:24:13,894
to move them or other animal systems or

663
00:24:13,894 --> 00:24:16,054
possibly even cells. You know, if you want

664
00:24:16,054 --> 00:24:18,694
to marshal cells in a particular direction, can

665
00:24:18,694 --> 00:24:20,794
you do that with some kind of artificial

666
00:24:20,934 --> 00:24:23,220
entity that is programmed in certain way?

667
00:24:23,700 --> 00:24:24,919
So, you know, these

668
00:24:25,380 --> 00:24:28,339
these, questions of how to shepherd a load

669
00:24:28,339 --> 00:24:28,839
of,

670
00:24:29,220 --> 00:24:30,200
you know, moving

671
00:24:30,579 --> 00:24:32,679
possibly quite idiosyncratic particles,

672
00:24:33,140 --> 00:24:34,599
there's a broader applicability

673
00:24:34,980 --> 00:24:37,115
to it than just she. Can we apply

674
00:24:37,115 --> 00:24:39,595
this to to people? Can we is there

675
00:24:39,595 --> 00:24:42,315
some way we can help with crowding at,

676
00:24:42,315 --> 00:24:44,634
I don't know, concerts or football matches or

677
00:24:44,634 --> 00:24:46,714
something by looking at this? It's it's already

678
00:24:46,714 --> 00:24:49,035
being done. Models like this to try to

679
00:24:49,035 --> 00:24:50,494
understand crowd behavior

680
00:24:51,089 --> 00:24:53,890
have been used to try to plan things

681
00:24:53,890 --> 00:24:56,069
like events like Notting Hill Carnival

682
00:24:56,450 --> 00:24:58,450
where you can, you know, literally map out

683
00:24:58,450 --> 00:25:00,929
the real route that the carnival takes and,

684
00:25:00,929 --> 00:25:02,609
you know, get some sense of what the

685
00:25:02,609 --> 00:25:05,275
crowd motions might be like, where you might

686
00:25:05,275 --> 00:25:07,595
have pressure points and, you know, difficulties that

687
00:25:07,595 --> 00:25:10,234
might need extra policing or extra guidance or

688
00:25:10,234 --> 00:25:10,894
so on.

689
00:25:11,275 --> 00:25:12,575
It's also been

690
00:25:13,115 --> 00:25:15,355
applied models like this have been applied to

691
00:25:15,355 --> 00:25:15,855
the,

692
00:25:17,115 --> 00:25:19,535
the Muslim pilgrimage of the the Hajj,

693
00:25:20,609 --> 00:25:21,509
which has

694
00:25:23,330 --> 00:25:24,150
had problems,

695
00:25:24,769 --> 00:25:27,009
you know, some years because so many people

696
00:25:27,009 --> 00:25:29,490
come to it. It's this massive crowd event

697
00:25:29,490 --> 00:25:31,330
that has, you know, the real danger of

698
00:25:31,330 --> 00:25:33,330
there being sort of buildups of pressure in

699
00:25:33,330 --> 00:25:35,095
the crowd. And there have been,

700
00:25:35,494 --> 00:25:37,575
disasters really that have happened in the past

701
00:25:37,575 --> 00:25:39,654
because of this. And so this sort of

702
00:25:39,654 --> 00:25:42,615
modeling has been applied also to to events

703
00:25:42,615 --> 00:25:44,714
like that to see if you can create

704
00:25:44,934 --> 00:25:46,394
better safety measures.

705
00:25:47,480 --> 00:25:49,799
And just sort of crowd panic in general,

706
00:25:49,799 --> 00:25:52,539
you know, is is a really difficult problem,

707
00:25:53,159 --> 00:25:53,659
that,

708
00:25:55,319 --> 00:25:55,980
you know,

709
00:25:56,440 --> 00:25:59,259
models like this can can can inform

710
00:25:59,559 --> 00:26:00,134
at least.

711
00:26:00,535 --> 00:26:03,195
So, you know, it's not necessarily that you'd,

712
00:26:04,134 --> 00:26:05,734
you'd be thinking of a a sort of

713
00:26:05,734 --> 00:26:07,994
shepherding problem, but that you'd be anticipating

714
00:26:09,174 --> 00:26:11,575
problems that might arise in these sort of

715
00:26:11,575 --> 00:26:12,795
mass crowd movements

716
00:26:13,335 --> 00:26:16,669
that are possibly quite predictable ones just as

717
00:26:16,669 --> 00:26:18,990
they are for road traffic, and then think

718
00:26:18,990 --> 00:26:20,990
about the kind of safety measures that might

719
00:26:20,990 --> 00:26:23,470
alleviate that. You mentioned your book that you

720
00:26:23,470 --> 00:26:25,309
wrote a a few years ago, but haven't

721
00:26:25,309 --> 00:26:26,509
you got a new one that's sort of

722
00:26:26,509 --> 00:26:28,704
looking in this area? Oh, yeah. How life

723
00:26:28,704 --> 00:26:31,024
works. That's looking more at the kind of

724
00:26:31,024 --> 00:26:34,784
molecular details of, of of that question. But

725
00:26:34,784 --> 00:26:36,544
it does, you know, it it it does

726
00:26:36,544 --> 00:26:39,044
then sort of, open out into

727
00:26:39,424 --> 00:26:41,684
looking at higher levels of the system.

728
00:26:42,410 --> 00:26:43,470
So it's actually

729
00:26:44,009 --> 00:26:45,889
yeah. I mean, in the sense that the

730
00:26:46,089 --> 00:26:47,710
that book is looking at

731
00:26:48,250 --> 00:26:51,130
questions of emergence in biological systems, you know,

732
00:26:51,130 --> 00:26:53,130
that's what we're talking about. It's absolutely what

733
00:26:53,130 --> 00:26:55,929
we're talking about with these collective motions of

734
00:26:55,929 --> 00:26:58,054
animals. I mean, you know, the murmurations

735
00:26:58,434 --> 00:27:00,694
of of birds are one of the classic

736
00:27:00,755 --> 00:27:04,194
examples of of emergence but we absolutely see

737
00:27:04,194 --> 00:27:06,835
in biological systems in in all sorts of

738
00:27:06,835 --> 00:27:09,474
ways how you get these higher levels of

739
00:27:09,474 --> 00:27:09,974
organization

740
00:27:10,869 --> 00:27:11,609
that are

741
00:27:12,069 --> 00:27:14,329
somehow somehow seem to be autonomous

742
00:27:15,029 --> 00:27:18,390
from the really lower level details of what

743
00:27:18,390 --> 00:27:20,549
individual molecules are doing and what genes are

744
00:27:20,549 --> 00:27:23,690
switched on. So that broader question

745
00:27:24,224 --> 00:27:25,525
of how you get

746
00:27:25,984 --> 00:27:28,724
emergent sort of coherence and robustness

747
00:27:29,105 --> 00:27:30,404
in the face of

748
00:27:31,265 --> 00:27:31,765
microscopic

749
00:27:32,144 --> 00:27:32,644
details

750
00:27:33,025 --> 00:27:35,025
that might be quite sort of random and

751
00:27:35,025 --> 00:27:35,525
idiosyncratic.

752
00:27:36,839 --> 00:27:39,259
That's absolutely the same kind of question,

753
00:27:40,119 --> 00:27:42,200
that we we see in, you know, in

754
00:27:42,200 --> 00:27:44,759
these cases. And in fact, it's actually you

755
00:27:44,759 --> 00:27:46,460
could say the same kind of problem

756
00:27:46,759 --> 00:27:48,299
that, you know, in a sense,

757
00:27:49,335 --> 00:27:52,775
that emergence in biology continues beyond the level

758
00:27:52,775 --> 00:27:54,634
of the organism to the ecosystem

759
00:27:54,934 --> 00:27:56,555
and for us to the society.

760
00:27:57,575 --> 00:27:58,075
And

761
00:27:58,375 --> 00:28:00,315
there are similar kind of principles

762
00:28:01,009 --> 00:28:03,409
involved in the sense of how do you

763
00:28:03,409 --> 00:28:05,109
get a high level system

764
00:28:05,569 --> 00:28:07,589
that is reliable and robust

765
00:28:08,049 --> 00:28:09,970
and isn't going to break down as soon

766
00:28:09,970 --> 00:28:11,349
as something microscopic

767
00:28:11,809 --> 00:28:12,629
goes awry.

768
00:28:13,715 --> 00:28:15,394
A walk among the sheep. They they don't

769
00:28:15,394 --> 00:28:17,075
worry about me walking amongst them. They just

770
00:28:17,075 --> 00:28:19,555
stay where they are. But it the movements

771
00:28:19,555 --> 00:28:21,555
that they have, as you say, you know,

772
00:28:21,555 --> 00:28:23,335
sort of that bit of field looks

773
00:28:23,715 --> 00:28:25,494
tastier or whatever it is,

774
00:28:25,799 --> 00:28:28,120
It it it doesn't seem to make sense.

775
00:28:28,120 --> 00:28:29,500
It seems to be

776
00:28:29,799 --> 00:28:30,940
just sort of

777
00:28:31,240 --> 00:28:33,480
I'm going to move now. It doesn't seem

778
00:28:33,480 --> 00:28:35,480
to be any more to it than that.

779
00:28:35,480 --> 00:28:38,200
Right? It just seems like, yeah. I'm moving

780
00:28:38,200 --> 00:28:39,720
moving over there and then the others follow

781
00:28:39,720 --> 00:28:42,595
because they, oh, maybe they found something. It

782
00:28:42,595 --> 00:28:45,174
doesn't seem to be kind of, you know,

783
00:28:45,394 --> 00:28:46,295
in any way,

784
00:28:47,634 --> 00:28:49,335
decisive based on evidence

785
00:28:49,795 --> 00:28:52,134
about just sort of going somewhere. The information

786
00:28:52,275 --> 00:28:54,179
bit of it, that's really lovely. Right? That

787
00:28:54,179 --> 00:28:56,259
that they're sharing that information, but maybe the

788
00:28:56,259 --> 00:28:58,279
information is dodgy. It isn't better.

789
00:28:58,899 --> 00:29:00,740
Mhmm. Well, I mean, okay. So maybe the

790
00:29:00,740 --> 00:29:02,500
first thing to say about that is that,

791
00:29:03,139 --> 00:29:05,059
and this is something that I heard in

792
00:29:05,059 --> 00:29:08,255
in writing this article that, you know, farmers,

793
00:29:08,955 --> 00:29:11,055
will say, actually, sheep,

794
00:29:11,595 --> 00:29:14,015
they're more intelligent than you think they are.

795
00:29:14,955 --> 00:29:16,575
They're and they're more individual

796
00:29:17,115 --> 00:29:19,615
than you think they are. So for example,

797
00:29:20,075 --> 00:29:21,455
it seems that sometimes,

798
00:29:23,359 --> 00:29:24,339
sheep flocks

799
00:29:24,960 --> 00:29:25,460
respond

800
00:29:25,919 --> 00:29:26,419
to

801
00:29:26,960 --> 00:29:29,839
what the farmer has told the sheepdog to

802
00:29:29,839 --> 00:29:31,519
do. So the farmer can, you know, has

803
00:29:31,519 --> 00:29:34,000
whistles or whatever to instruct the sheepdog to

804
00:29:34,000 --> 00:29:36,799
do particular things. The sheep get to to

805
00:29:36,799 --> 00:29:37,299
recognize

806
00:29:38,105 --> 00:29:38,765
those calls

807
00:29:39,144 --> 00:29:41,705
and start behaving, you know, as the dog

808
00:29:41,705 --> 00:29:43,465
would want them to behave before the dog

809
00:29:43,465 --> 00:29:45,305
has really done anything, you know. So then

810
00:29:45,305 --> 00:29:46,984
they are picking up on these things. But

811
00:29:46,984 --> 00:29:48,664
also it seems like there's quite a lot

812
00:29:48,664 --> 00:29:49,244
of variability

813
00:29:49,705 --> 00:29:52,109
in the way they behave. Some sheep just

814
00:29:52,109 --> 00:29:54,109
look at the sheepdog and think whatever and

815
00:29:54,109 --> 00:29:56,109
ignore it. Whereas others think, oh my god,

816
00:29:56,109 --> 00:29:57,789
and run like mad the moment they see

817
00:29:57,789 --> 00:30:00,769
it. So that, you know, that's, that variability

818
00:30:01,150 --> 00:30:02,369
you need to build into,

819
00:30:02,910 --> 00:30:05,730
the into into these models. But it also

820
00:30:06,575 --> 00:30:09,214
undermines the the, you know, the the kind

821
00:30:09,214 --> 00:30:11,615
of meme like stereotype really is, you know,

822
00:30:11,615 --> 00:30:13,535
sheep are just they just all do the

823
00:30:13,535 --> 00:30:15,075
same thing. Well, they really don't.

824
00:30:15,934 --> 00:30:17,775
But the question of, you know, why they're

825
00:30:17,775 --> 00:30:19,695
doing this in the first place, it's it's

826
00:30:19,695 --> 00:30:21,234
it's all about foraging

827
00:30:21,599 --> 00:30:24,559
strategies, and each animal has, you know, has

828
00:30:24,559 --> 00:30:25,460
some. And

829
00:30:26,240 --> 00:30:28,640
the this has been, a very sort of

830
00:30:28,640 --> 00:30:31,380
rich area of research of how do animals

831
00:30:31,519 --> 00:30:34,660
decide what strategy to follow, what's what's efficient.

832
00:30:35,039 --> 00:30:36,579
Some of them seem to,

833
00:30:37,184 --> 00:30:39,825
undergo what are called levy flights where, you

834
00:30:39,825 --> 00:30:42,464
know, the idea is you go to one

835
00:30:42,464 --> 00:30:44,625
location and you forage around a little bit

836
00:30:44,625 --> 00:30:46,384
around there with random walk, and then you

837
00:30:46,384 --> 00:30:48,065
make a big jump to somewhere else and

838
00:30:48,065 --> 00:30:49,980
then you forage a little bit around there.

839
00:30:50,460 --> 00:30:52,539
That's something that you could describe using a

840
00:30:52,539 --> 00:30:54,480
particular kind of statistics.

841
00:30:55,339 --> 00:30:57,259
You know, other animal foragers. I mean, if

842
00:30:57,259 --> 00:30:59,660
we think the classic example really is bees.

843
00:30:59,660 --> 00:31:02,380
You know, that's that's extraordinary, the amount of

844
00:31:02,380 --> 00:31:05,875
cognition that's involved there that they will you

845
00:31:05,875 --> 00:31:06,934
know, they'll go out,

846
00:31:07,394 --> 00:31:09,315
they'll, you know, they'll find some,

847
00:31:10,355 --> 00:31:11,894
some sort of source of food,

848
00:31:12,195 --> 00:31:13,795
come back to the hive, and tell the

849
00:31:13,795 --> 00:31:16,115
others where it is but with this incredible

850
00:31:16,115 --> 00:31:18,355
waggle dance and the others will understand and

851
00:31:18,355 --> 00:31:20,919
off they'll go. And sometimes they won't. Sometimes

852
00:31:20,919 --> 00:31:22,440
they'll think, you know what? I'm it doesn't

853
00:31:22,599 --> 00:31:24,119
I you know, I've been there and it's

854
00:31:24,119 --> 00:31:25,720
not I mean, I'm you know, this isn't

855
00:31:25,720 --> 00:31:27,480
what the bees are thinking, but it's how

856
00:31:27,480 --> 00:31:30,039
they seem to behave. They don't all just

857
00:31:30,039 --> 00:31:32,595
sort of respond in this programmed way. So

858
00:31:32,595 --> 00:31:34,194
even in the case of a bee with

859
00:31:34,194 --> 00:31:36,275
this kind of, you know, brain the size

860
00:31:36,275 --> 00:31:37,654
of a grain of salt,

861
00:31:38,035 --> 00:31:40,994
let alone a sheep, there's an amazing amount

862
00:31:40,994 --> 00:31:44,054
of cognition, of subtle cognition going on there.

863
00:31:44,275 --> 00:31:44,775
So,

864
00:31:45,234 --> 00:31:46,775
you know, I I I think

865
00:31:47,230 --> 00:31:47,730
although,

866
00:31:48,349 --> 00:31:50,109
I mean in a way that's what these

867
00:31:50,109 --> 00:31:52,589
models are kind of showing us that, you

868
00:31:52,589 --> 00:31:55,970
know, we can be described using these particle

869
00:31:56,029 --> 00:31:58,130
based models in certain situations

870
00:31:58,750 --> 00:32:00,509
and yet we like to think that we

871
00:32:00,509 --> 00:32:02,964
have this extraordinary amount of cognition that, you

872
00:32:02,964 --> 00:32:04,964
know, we're making use of in other contexts.

873
00:32:04,964 --> 00:32:07,065
And so we should probably not underestimate

874
00:32:07,764 --> 00:32:10,565
the amount of cognition that other animals are

875
00:32:10,565 --> 00:32:11,544
capable of,

876
00:32:12,565 --> 00:32:14,825
and the the way that the,

877
00:32:15,444 --> 00:32:17,304
the the way that that is

878
00:32:18,000 --> 00:32:19,220
harvested collectively.

879
00:32:20,000 --> 00:32:21,680
You know, that's the key thing that the

880
00:32:22,000 --> 00:32:23,779
these creatures aren't making decisions

881
00:32:24,400 --> 00:32:25,539
just by themselves.

882
00:32:26,000 --> 00:32:28,640
They're actually that that these strategies have this

883
00:32:28,640 --> 00:32:30,339
sort of collective ability

884
00:32:31,195 --> 00:32:33,535
to pool information and to make effective

885
00:32:33,835 --> 00:32:35,055
use of the group.

886
00:32:35,674 --> 00:32:36,174
So

887
00:32:36,875 --> 00:32:38,815
it's it's a much richer

888
00:32:39,355 --> 00:32:39,855
environment

889
00:32:40,475 --> 00:32:42,414
than you might think understandably

890
00:32:42,795 --> 00:32:44,634
by going up to a sheep and thinking,

891
00:32:44,634 --> 00:32:46,440
there's not a lot going on in there.

892
00:32:46,519 --> 00:32:49,079
Actually, you know, there may or there may

893
00:32:49,079 --> 00:32:50,759
not be, but what comes out of it

894
00:32:50,759 --> 00:32:52,059
is really quite rich.

895
00:32:53,240 --> 00:32:56,460
With no offense to any sheep or farmers

896
00:32:56,519 --> 00:32:57,019
listening,

897
00:32:57,319 --> 00:32:59,099
I wanted to know more about

898
00:32:59,575 --> 00:33:01,755
the human interaction in crowds.

899
00:33:03,174 --> 00:33:05,434
Someone who won the Ignat Nobel Prize

900
00:33:05,735 --> 00:33:06,634
for the research

901
00:33:07,095 --> 00:33:08,555
into crowd movement

902
00:33:09,095 --> 00:33:09,994
is Alessandro

903
00:33:10,535 --> 00:33:11,035
Corveto.

904
00:33:13,609 --> 00:33:15,690
I am an assistant professor at the Endoben

905
00:33:15,690 --> 00:33:16,670
University of Technology

906
00:33:17,369 --> 00:33:19,630
in the department of applied physics.

907
00:33:20,410 --> 00:33:22,170
I have here my own group that is

908
00:33:22,170 --> 00:33:24,830
called the AI for complex flows and traffic.

909
00:33:25,289 --> 00:33:28,029
And, what we do here is,

910
00:33:28,345 --> 00:33:30,684
well, essentially 2 things. So first,

911
00:33:31,305 --> 00:33:31,805
use

912
00:33:32,105 --> 00:33:32,605
tools

913
00:33:32,985 --> 00:33:34,605
of physics to explore,

914
00:33:34,985 --> 00:33:38,605
let's say, new physical systems, specifically traffic and,

915
00:33:39,225 --> 00:33:40,045
most importantly,

916
00:33:40,664 --> 00:33:43,485
crowd traffic. And the second thing is

917
00:33:44,000 --> 00:33:44,820
we use AI

918
00:33:45,200 --> 00:33:47,840
tools to try to crack problems in in

919
00:33:47,840 --> 00:33:50,820
complex fluids and complex flows that were that

920
00:33:51,840 --> 00:33:52,980
are open for long.

921
00:33:53,440 --> 00:33:56,500
And, and therefore AI is giving new opportunities

922
00:33:56,559 --> 00:33:57,380
in this direction.

923
00:33:57,784 --> 00:33:59,085
We won the 2021,

924
00:33:59,784 --> 00:34:03,085
Iqnoebel Prize for a research that we started

925
00:34:03,304 --> 00:34:04,684
somewhere around 2013

926
00:34:05,625 --> 00:34:07,944
and, that is, that is a very big

927
00:34:07,944 --> 00:34:09,065
part of my work,

928
00:34:09,864 --> 00:34:10,844
today as well.

929
00:34:11,385 --> 00:34:12,364
So we

930
00:34:12,869 --> 00:34:13,849
try to understand,

931
00:34:15,509 --> 00:34:19,049
whether we can establish some basic physical understanding

932
00:34:19,109 --> 00:34:19,609
of

933
00:34:20,309 --> 00:34:23,130
of the way pedestrian crowds moves. Okay?

934
00:34:23,829 --> 00:34:24,329
And

935
00:34:24,934 --> 00:34:26,795
and this means try to establish

936
00:34:27,414 --> 00:34:30,954
a physical model that reproduce, that explain,

937
00:34:31,414 --> 00:34:33,514
what we see as we look at crowds.

938
00:34:34,855 --> 00:34:35,355
And

939
00:34:36,454 --> 00:34:38,309
let's say crowds are

940
00:34:38,869 --> 00:34:40,869
very, let's say, as you can as you

941
00:34:40,869 --> 00:34:43,210
can imagine, a very random, a very stochastic

942
00:34:43,269 --> 00:34:45,750
system. Right? So imagine that you are yourself

943
00:34:45,750 --> 00:34:47,050
in a in a train station.

944
00:34:47,430 --> 00:34:49,349
What you do today is different from what

945
00:34:49,349 --> 00:34:51,764
you'll do tomorrow, and what another person does

946
00:34:51,764 --> 00:34:54,005
today is different from what they will do

947
00:34:54,005 --> 00:34:55,704
tomorrow and what you will do today.

948
00:34:56,324 --> 00:34:58,964
So it's an extremely stochastic system. And the

949
00:34:58,964 --> 00:35:00,264
the key idea

950
00:35:00,964 --> 00:35:03,444
of of our work was to explore this

951
00:35:03,444 --> 00:35:06,659
system, collecting as much data as possible, in

952
00:35:06,739 --> 00:35:08,920
this case, in a train station. Now imagine

953
00:35:09,299 --> 00:35:09,780
tracking,

954
00:35:10,179 --> 00:35:12,579
all the people passing in this train station

955
00:35:12,579 --> 00:35:13,719
for, like, 1 year.

956
00:35:14,179 --> 00:35:17,219
Now we we speak about, 100,000 people per

957
00:35:17,219 --> 00:35:17,719
day,

958
00:35:18,339 --> 00:35:20,819
roughly speaking. So you you multiply this by

959
00:35:20,819 --> 00:35:23,965
360 days. You are immediately in the

960
00:35:24,344 --> 00:35:24,925
the millions

961
00:35:26,664 --> 00:35:27,164
scale.

962
00:35:27,545 --> 00:35:29,085
Well, many millions of scale.

963
00:35:29,945 --> 00:35:30,425
And then,

964
00:35:30,985 --> 00:35:33,804
the endeavor was or the endeavor still is

965
00:35:33,945 --> 00:35:34,925
to find

966
00:35:36,119 --> 00:35:39,739
universal pattern or universal physical features that emerge

967
00:35:39,960 --> 00:35:41,159
as we look at this,

968
00:35:41,480 --> 00:35:42,839
dynamical system at this,

969
00:35:43,319 --> 00:35:44,460
this large scale.

970
00:35:45,239 --> 00:35:45,739
And

971
00:35:46,199 --> 00:35:47,019
let's say

972
00:35:47,319 --> 00:35:49,159
you need to start simple. Right? So this

973
00:35:49,159 --> 00:35:51,295
is a this is a extremely complex system.

974
00:35:51,295 --> 00:35:52,815
So we start simple. And,

975
00:35:53,454 --> 00:35:55,875
so first first step is to look at

976
00:35:56,494 --> 00:35:59,775
how individual particles, let's say, individual pedestrian move,

977
00:35:59,775 --> 00:36:00,914
what is their statistical

978
00:36:01,215 --> 00:36:03,659
fingerprint. So how can you describe this with

979
00:36:03,659 --> 00:36:05,839
the with the stochastic differential equation?

980
00:36:06,460 --> 00:36:07,199
And, specifically,

981
00:36:07,739 --> 00:36:09,119
the ignoble was

982
00:36:09,500 --> 00:36:11,739
centered around the, let's say, the volume 2

983
00:36:11,739 --> 00:36:13,659
that is, when you move from a single

984
00:36:13,659 --> 00:36:15,679
particle to 2 particles interacting.

985
00:36:16,065 --> 00:36:16,964
So we did,

986
00:36:17,264 --> 00:36:17,764
something,

987
00:36:18,065 --> 00:36:20,864
let's say, in some sense inspired by by

988
00:36:20,864 --> 00:36:24,065
by impact parameter studies in in scattering physics.

989
00:36:24,065 --> 00:36:25,664
That is imagine that now you have 2

990
00:36:25,664 --> 00:36:27,045
particles that that

991
00:36:27,425 --> 00:36:27,824
are,

992
00:36:28,304 --> 00:36:31,320
walking or are moving in opposite direction. And

993
00:36:31,320 --> 00:36:32,619
then, I mean,

994
00:36:33,640 --> 00:36:34,140
obviously,

995
00:36:34,840 --> 00:36:37,239
you don't collide typically with with the particle

996
00:36:37,239 --> 00:36:39,239
coming in in in the opposite direction. And

997
00:36:39,239 --> 00:36:41,739
so there is, there is a whole avoidance

998
00:36:41,800 --> 00:36:44,059
mechanism that is that is triggered.

999
00:36:44,904 --> 00:36:47,244
And we characterize this avoidance mechanism

1000
00:36:47,545 --> 00:36:49,784
looking, again in the in the hundreds of

1001
00:36:49,784 --> 00:36:50,844
thousands of trajectories.

1002
00:36:51,304 --> 00:36:53,625
And, and we wrote the differential equation that

1003
00:36:53,625 --> 00:36:56,585
reproduce the same statistical fingerprint. So what happens

1004
00:36:57,590 --> 00:36:59,590
So what are the the the the common

1005
00:36:59,590 --> 00:37:00,829
way in which this process,

1006
00:37:01,269 --> 00:37:01,769
happens?

1007
00:37:02,869 --> 00:37:04,630
The fact that every now and then there

1008
00:37:04,630 --> 00:37:06,409
are indeed collisions and

1009
00:37:06,710 --> 00:37:09,269
how, let's say, the interaction decays in space

1010
00:37:09,269 --> 00:37:11,994
and and so on. Okay. Okay. So, like,

1011
00:37:11,994 --> 00:37:14,635
for example, one one time, I was walking

1012
00:37:14,635 --> 00:37:17,434
along Tottenham Court Road in London. A taxi

1013
00:37:17,434 --> 00:37:19,054
pulled up. I wasn't concentrating.

1014
00:37:19,434 --> 00:37:21,034
It was before I had a mobile phone,

1015
00:37:21,034 --> 00:37:22,235
so I wasn't looking at that. I don't

1016
00:37:22,235 --> 00:37:24,000
know. I was probably looking at a a

1017
00:37:24,000 --> 00:37:26,159
Lego shop or something. And, the,

1018
00:37:26,640 --> 00:37:28,559
the door opened of the of the taxi

1019
00:37:28,559 --> 00:37:30,800
without me noticing, and somebody stepped out, and

1020
00:37:30,800 --> 00:37:31,860
I bumped into,

1021
00:37:32,880 --> 00:37:35,599
that person. And that person was Billy Piper,

1022
00:37:35,599 --> 00:37:37,119
who at the time was in doctor who,

1023
00:37:37,119 --> 00:37:38,864
which was terribly exciting for me.

1024
00:37:39,425 --> 00:37:41,664
I imagine it wasn't terribly exciting for her.

1025
00:37:41,664 --> 00:37:43,264
So that kind of collision has had a

1026
00:37:43,264 --> 00:37:45,844
different reaction with the 2 different particles.

1027
00:37:46,144 --> 00:37:48,385
So one of those particles had a, you

1028
00:37:48,385 --> 00:37:48,885
know,

1029
00:37:49,344 --> 00:37:52,170
reaction of boredom and annoyance, and the other

1030
00:37:52,170 --> 00:37:53,469
one had a reaction

1031
00:37:53,769 --> 00:37:54,829
of excitement

1032
00:37:55,210 --> 00:37:57,309
and can't wait to tell its other particle

1033
00:37:57,369 --> 00:37:58,109
nerd friends

1034
00:37:58,489 --> 00:37:59,469
about the interaction.

1035
00:37:59,849 --> 00:38:01,469
Let's say this is, I think,

1036
00:38:03,210 --> 00:38:05,609
connected with with the other Nobel Prize that

1037
00:38:05,609 --> 00:38:07,744
has been that has been won the same

1038
00:38:07,744 --> 00:38:09,824
year by the group in Japan of of

1039
00:38:09,824 --> 00:38:11,525
Claudio Feliciani and the others.

1040
00:38:12,065 --> 00:38:14,864
And, yeah, they they started, let's say, the

1041
00:38:14,864 --> 00:38:15,925
the case in which

1042
00:38:16,864 --> 00:38:18,945
which people are not paying attention because they

1043
00:38:18,945 --> 00:38:21,900
are distracted by by mobile phones. And they

1044
00:38:21,900 --> 00:38:22,559
are examined

1045
00:38:23,179 --> 00:38:23,679
essentially

1046
00:38:24,059 --> 00:38:24,559
how

1047
00:38:24,859 --> 00:38:28,539
this interaction mechanism gets penalized as as people,

1048
00:38:28,780 --> 00:38:30,539
don't pay attention. What does it tell us

1049
00:38:30,539 --> 00:38:31,039
about

1050
00:38:31,659 --> 00:38:32,159
ourselves,

1051
00:38:32,940 --> 00:38:33,440
crowds,

1052
00:38:33,914 --> 00:38:37,215
or physics in more widely? Bumping or perhaps

1053
00:38:37,275 --> 00:38:38,175
not bumping

1054
00:38:38,474 --> 00:38:38,875
is,

1055
00:38:39,434 --> 00:38:39,934
is

1056
00:38:40,235 --> 00:38:42,954
the basic interaction mechanism that there is in

1057
00:38:42,954 --> 00:38:43,535
this type

1058
00:38:44,394 --> 00:38:46,289
of system. Let's say that that, you know,

1059
00:38:46,289 --> 00:38:48,309
as physicists or as applied mathematician,

1060
00:38:48,610 --> 00:38:50,690
what mathematicians, what we'd like to do is

1061
00:38:50,690 --> 00:38:53,250
start from something complex. Right? We have many

1062
00:38:53,250 --> 00:38:54,849
agents, many people, and so on and so

1063
00:38:54,849 --> 00:38:56,070
forth, and we try to

1064
00:38:56,929 --> 00:38:59,910
reduce it using, the language of of mathematical

1065
00:39:00,050 --> 00:39:01,590
physics. Right? And so immediately,

1066
00:39:02,984 --> 00:39:05,704
you you would reduce a system of interacting

1067
00:39:05,704 --> 00:39:08,264
pedestrians into a system of particles that, in

1068
00:39:08,264 --> 00:39:09,085
a way or another,

1069
00:39:09,545 --> 00:39:09,784
have,

1070
00:39:10,344 --> 00:39:13,385
interactions that are in some sense similar to

1071
00:39:13,385 --> 00:39:14,684
gravity, but probably

1072
00:39:15,429 --> 00:39:17,750
anti gravity. Right? So you repel each other

1073
00:39:17,750 --> 00:39:18,489
instead of,

1074
00:39:19,269 --> 00:39:19,769
attracting,

1075
00:39:20,309 --> 00:39:22,150
each other. And, and,

1076
00:39:23,590 --> 00:39:24,789
the interesting thing,

1077
00:39:25,190 --> 00:39:25,849
is that

1078
00:39:28,284 --> 00:39:30,144
considering a a particle system

1079
00:39:30,445 --> 00:39:32,364
with with this very basic,

1080
00:39:33,005 --> 00:39:34,144
repulsion mechanism,

1081
00:39:35,324 --> 00:39:36,704
is enough to recover

1082
00:39:37,005 --> 00:39:37,985
some of the

1083
00:39:38,684 --> 00:39:39,824
qualitative features

1084
00:39:40,284 --> 00:39:42,204
of this system. So you might be aware

1085
00:39:42,204 --> 00:39:43,505
of of of the following.

1086
00:39:44,269 --> 00:39:46,609
This is a very, very much used case

1087
00:39:46,750 --> 00:39:49,070
to explain some of this emergent behavior that

1088
00:39:49,070 --> 00:39:50,829
you have in this in this crowd system.

1089
00:39:50,829 --> 00:39:51,469
Right? So,

1090
00:39:51,869 --> 00:39:53,789
okay. Okay. Maybe let let me add something

1091
00:39:53,789 --> 00:39:54,289
that,

1092
00:39:55,150 --> 00:39:58,184
one of the the connection point between between

1093
00:39:58,425 --> 00:40:01,144
crowd dynamics and and physics is in the

1094
00:40:01,144 --> 00:40:03,704
direction of complex system. Right? So complex system

1095
00:40:03,704 --> 00:40:04,364
is about,

1096
00:40:05,305 --> 00:40:08,105
how, let's say, complex pattern or complex structure

1097
00:40:08,105 --> 00:40:09,085
emerge as

1098
00:40:09,385 --> 00:40:09,885
basic

1099
00:40:10,659 --> 00:40:12,900
entities interact with simple rules. So the fact

1100
00:40:12,900 --> 00:40:15,780
that you can get macroscopically nontrivial patterns out

1101
00:40:15,780 --> 00:40:16,760
of basic

1102
00:40:17,219 --> 00:40:19,880
rules between many, many agents. And in crowds,

1103
00:40:20,179 --> 00:40:21,059
the very basic,

1104
00:40:21,539 --> 00:40:24,360
emergent behavior that is considered is the following.

1105
00:40:24,500 --> 00:40:25,960
So imagine that you have

1106
00:40:27,394 --> 00:40:30,034
a crowd of people, let's say, with with

1107
00:40:30,034 --> 00:40:33,094
red shirts that are moving along 1 one

1108
00:40:34,675 --> 00:40:37,315
one street from one side to another and

1109
00:40:37,315 --> 00:40:39,474
another crowd of people with blue shirts that

1110
00:40:39,474 --> 00:40:41,235
are moving on the same street but in

1111
00:40:41,235 --> 00:40:43,369
the opposite direction. Okay? So now you have

1112
00:40:43,369 --> 00:40:45,930
the 2 big crowds, and they're facing each

1113
00:40:45,930 --> 00:40:46,829
other and walking

1114
00:40:47,289 --> 00:40:49,610
in in opposite direction. And then the very

1115
00:40:49,610 --> 00:40:50,110
basic,

1116
00:40:51,289 --> 00:40:53,869
emergent behavior that this system has is that

1117
00:40:55,144 --> 00:40:55,644
particles,

1118
00:40:56,344 --> 00:40:58,284
or or people that have

1119
00:40:58,824 --> 00:41:00,344
that are going in the same direction, which

1120
00:41:00,344 --> 00:41:02,344
means that they they come with a a

1121
00:41:02,344 --> 00:41:03,644
shirt of the same color,

1122
00:41:04,264 --> 00:41:07,224
would align in stripes. So now imagine yourself

1123
00:41:07,224 --> 00:41:09,085
looking up. Now these 2 crowds

1124
00:41:09,860 --> 00:41:12,019
walked, and now they are crossing each other.

1125
00:41:12,019 --> 00:41:14,179
But then you wouldn't observe that, the system

1126
00:41:14,179 --> 00:41:16,900
is completely disordered, so you wouldn't observe completely

1127
00:41:16,900 --> 00:41:19,619
random distribution of red, of red and blue

1128
00:41:19,619 --> 00:41:20,119
shirts.

1129
00:41:20,739 --> 00:41:22,119
You would rather observe

1130
00:41:22,494 --> 00:41:24,514
the fact that the system has ordered

1131
00:41:25,214 --> 00:41:26,594
in in in stripes

1132
00:41:26,894 --> 00:41:29,054
of shirts of the same color, meaning of

1133
00:41:29,054 --> 00:41:31,214
stripes of people going in the same same

1134
00:41:31,214 --> 00:41:34,195
direction. And the the fascinating thing is that

1135
00:41:35,789 --> 00:41:36,289
it's

1136
00:41:37,230 --> 00:41:37,730
extremely

1137
00:41:38,110 --> 00:41:42,030
well, it's surprisingly simple to regenerate or to

1138
00:41:42,030 --> 00:41:42,530
to,

1139
00:41:43,150 --> 00:41:44,210
explain, to reproduce

1140
00:41:44,590 --> 00:41:45,730
this emerging behavior

1141
00:41:46,030 --> 00:41:46,769
by just

1142
00:41:47,150 --> 00:41:50,050
having particle moving in opposite direction that

1143
00:41:50,765 --> 00:41:53,325
know only a single rule beside the the

1144
00:41:53,325 --> 00:41:54,684
the need of moving that is the need

1145
00:41:54,684 --> 00:41:57,164
of avoiding. So the need of not entering

1146
00:41:57,164 --> 00:42:00,045
into very string stronger contact. Right? So so

1147
00:42:00,045 --> 00:42:00,785
the mechanism

1148
00:42:01,164 --> 00:42:03,324
of not bumping into each other is already

1149
00:42:03,324 --> 00:42:05,184
enough to explain quite interesting

1150
00:42:06,250 --> 00:42:06,750
emergent,

1151
00:42:07,210 --> 00:42:09,130
emergent behavior. This is only one. Right? There

1152
00:42:09,130 --> 00:42:10,269
are there are others.

1153
00:42:10,730 --> 00:42:11,230
And,

1154
00:42:12,809 --> 00:42:15,050
Yeah. Yeah. It took a very long road

1155
00:42:15,050 --> 00:42:17,610
here. Brilliant. So when you are you're doing

1156
00:42:17,610 --> 00:42:20,025
all of this research into into crowds, what

1157
00:42:20,025 --> 00:42:22,684
what's your goal? There are multiple goals. So,

1158
00:42:23,144 --> 00:42:25,065
let's say, as a physicist, there is the

1159
00:42:25,065 --> 00:42:26,764
challenge of of explaining

1160
00:42:27,065 --> 00:42:28,204
something that is

1161
00:42:28,984 --> 00:42:31,464
extremely complex, extremely stochastic. So there is the

1162
00:42:31,464 --> 00:42:33,804
the the underlying question, can we put down

1163
00:42:34,329 --> 00:42:37,609
physical rules that that regulate this system? And

1164
00:42:37,609 --> 00:42:39,609
here, one needs to go deep enough because,

1165
00:42:39,609 --> 00:42:42,030
obviously, you cannot expect to have a mathematical

1166
00:42:42,089 --> 00:42:44,569
model that reproduces what one person does because

1167
00:42:44,569 --> 00:42:45,469
this is obviously,

1168
00:42:46,170 --> 00:42:48,109
I mean, meaningless. So this is completely

1169
00:42:48,914 --> 00:42:51,155
senseless because it's just unpredictable. But then,

1170
00:42:52,114 --> 00:42:54,775
in in our case, going deep enough means

1171
00:42:55,394 --> 00:42:56,375
having enough

1172
00:42:57,474 --> 00:42:57,974
data

1173
00:42:58,275 --> 00:42:59,255
of this system,

1174
00:43:00,195 --> 00:43:03,130
in such a way that that statistical fingerprint

1175
00:43:03,190 --> 00:43:05,750
emerged. Like, the fact that the average behavior

1176
00:43:05,750 --> 00:43:07,590
of the system is this. The fluctuation the

1177
00:43:07,590 --> 00:43:10,550
typical fluctuation is that. Events that happen 1

1178
00:43:10,550 --> 00:43:12,630
in 1,000 are these and these and these.

1179
00:43:12,630 --> 00:43:14,550
Events that happen 1 in 10000 are these

1180
00:43:14,550 --> 00:43:15,829
and these and these. And I think at

1181
00:43:15,829 --> 00:43:16,969
this point, it becomes

1182
00:43:17,385 --> 00:43:20,184
not so surprising the fact that about this

1183
00:43:20,184 --> 00:43:21,644
behavior, one can write,

1184
00:43:22,105 --> 00:43:23,085
can write physical,

1185
00:43:23,545 --> 00:43:24,045
equations.

1186
00:43:24,985 --> 00:43:28,425
And, and, and this aligns with, again, topics

1187
00:43:28,425 --> 00:43:31,244
such as, physics of active matter that is,

1188
00:43:32,059 --> 00:43:33,920
the physics that study all sort

1189
00:43:34,460 --> 00:43:36,059
of matter that in a way or another

1190
00:43:36,059 --> 00:43:38,619
is capable of of turning internal energy into

1191
00:43:38,619 --> 00:43:39,519
motion spontaneously.

1192
00:43:39,900 --> 00:43:41,360
Right? So we go from,

1193
00:43:41,820 --> 00:43:42,320
bacteria.

1194
00:43:42,940 --> 00:43:45,019
We go from, we consider, I don't know,

1195
00:43:45,019 --> 00:43:46,000
flocks of birds,

1196
00:43:47,994 --> 00:43:50,394
and so in general animal behavior and so

1197
00:43:50,394 --> 00:43:52,974
on. It aligns with the physics of complexity

1198
00:43:53,114 --> 00:43:54,335
that is, how,

1199
00:43:54,635 --> 00:43:55,775
as I said, how

1200
00:43:56,074 --> 00:43:59,034
nontrivial pattern emerged from simple inter from simple

1201
00:43:59,034 --> 00:43:59,534
interaction.

1202
00:43:59,914 --> 00:44:02,929
And this this case also, especially when we

1203
00:44:02,929 --> 00:44:04,769
go in a high density regime, it can

1204
00:44:04,769 --> 00:44:07,570
align with with, some condensed matter structure. So

1205
00:44:07,650 --> 00:44:09,650
but in general, there are there are very

1206
00:44:09,650 --> 00:44:12,630
many connections with different fields of physics.

1207
00:44:13,265 --> 00:44:15,045
And then, of course, there is also the

1208
00:44:15,105 --> 00:44:16,085
the the societal

1209
00:44:16,465 --> 00:44:18,785
slash engineering component that is,

1210
00:44:19,184 --> 00:44:21,265
now we get to have a model that

1211
00:44:21,265 --> 00:44:23,765
that is capable of reproducing the the stochastic

1212
00:44:23,824 --> 00:44:25,905
dynamics of the system. How about we use

1213
00:44:25,905 --> 00:44:26,724
it to

1214
00:44:27,880 --> 00:44:31,340
in design phase perhaps of of a facility?

1215
00:44:31,719 --> 00:44:33,400
Or we use this to do real time

1216
00:44:33,400 --> 00:44:35,719
control, which are things that actually are happening.

1217
00:44:35,719 --> 00:44:38,440
So let's say you observe your system at

1218
00:44:38,440 --> 00:44:40,554
a very large scale, and then you you

1219
00:44:40,554 --> 00:44:42,474
have a model that tries to predict what's

1220
00:44:42,474 --> 00:44:43,994
most likely going to happen in the next

1221
00:44:43,994 --> 00:44:46,155
5, 10 minutes, and then you use this

1222
00:44:46,155 --> 00:44:49,114
to to nudge. Right? So to to exert

1223
00:44:49,114 --> 00:44:50,715
some action in the system in such a

1224
00:44:50,715 --> 00:44:53,570
way that that it behaves as it, I

1225
00:44:53,570 --> 00:44:55,809
mean, to increase safety, for instance, or or

1226
00:44:55,809 --> 00:44:58,690
or comfort. And this, this is this might

1227
00:44:58,690 --> 00:45:00,389
sound science fiction, but this

1228
00:45:00,769 --> 00:45:01,829
is active research

1229
00:45:02,289 --> 00:45:04,289
on one hand, and already there are many

1230
00:45:04,289 --> 00:45:06,610
attempts in this direction already on the other.

1231
00:45:06,610 --> 00:45:08,594
So Yeah. Do you do you have any

1232
00:45:08,835 --> 00:45:10,454
examples of where it's been?

1233
00:45:11,315 --> 00:45:11,795
Well,

1234
00:45:12,195 --> 00:45:14,275
I can I can connect, with with my

1235
00:45:14,275 --> 00:45:14,775
research?

1236
00:45:15,795 --> 00:45:17,014
For instance, we're investigating,

1237
00:45:17,635 --> 00:45:19,175
the usage of of illumination

1238
00:45:20,114 --> 00:45:20,934
to nudge,

1239
00:45:22,275 --> 00:45:23,414
people into,

1240
00:45:24,829 --> 00:45:26,050
let's say, actually,

1241
00:45:26,989 --> 00:45:29,809
into distributing in space to maximize

1242
00:45:30,190 --> 00:45:31,250
some some property.

1243
00:45:33,150 --> 00:45:35,230
A typical application would be, for instance, in

1244
00:45:35,230 --> 00:45:37,090
museums. Right? So you would like

1245
00:45:37,715 --> 00:45:39,894
floor usage, to be uniform,

1246
00:45:40,755 --> 00:45:41,735
and not, that,

1247
00:45:42,114 --> 00:45:44,755
you know, your crowd is fully concentrated in

1248
00:45:44,755 --> 00:45:46,594
one room and then all the rest of

1249
00:45:46,594 --> 00:45:47,655
the museum is,

1250
00:45:48,034 --> 00:45:48,534
empty.

1251
00:45:49,795 --> 00:45:52,450
We've been working with, with with sound with

1252
00:45:52,450 --> 00:45:54,869
positional sound as well. Let's say you,

1253
00:45:56,050 --> 00:45:57,349
only in certain location,

1254
00:45:57,650 --> 00:45:59,030
you can hear some sound.

1255
00:45:59,570 --> 00:46:01,969
And, only for instance, if you have a

1256
00:46:01,969 --> 00:46:03,650
complete track this is something that we we

1257
00:46:03,650 --> 00:46:05,989
published recently that, let's say,

1258
00:46:06,974 --> 00:46:10,114
only if you are walking a specific trajectory,

1259
00:46:10,255 --> 00:46:12,835
you could hear a full, let's say, melody.

1260
00:46:13,214 --> 00:46:13,714
Right?

1261
00:46:14,094 --> 00:46:14,414
And,

1262
00:46:15,054 --> 00:46:17,135
and if you steer out from this, let's

1263
00:46:17,135 --> 00:46:20,034
say, target trajectory, you stop hearing the melody.

1264
00:46:20,190 --> 00:46:22,670
And so there we show that actually we

1265
00:46:22,670 --> 00:46:25,550
could have people following the the, let's say,

1266
00:46:25,550 --> 00:46:26,609
the target trajectory

1267
00:46:26,909 --> 00:46:28,909
due to the fact that there was, an

1268
00:46:28,909 --> 00:46:30,909
expectation by people of keep on hearing the

1269
00:46:30,909 --> 00:46:33,550
melody. You know? So it's like every step

1270
00:46:33,550 --> 00:46:34,530
you hear a different

1271
00:46:35,614 --> 00:46:38,194
tone. Right? And then people would be,

1272
00:46:39,214 --> 00:46:40,194
I mean, spontaneously,

1273
00:46:40,574 --> 00:46:42,494
you would like to hear what you expect

1274
00:46:42,494 --> 00:46:44,494
to be the next tone. Right? So imagine

1275
00:46:44,494 --> 00:46:45,694
that there is a there is a simple

1276
00:46:45,694 --> 00:46:48,014
melody, and and this enable us to to

1277
00:46:48,014 --> 00:46:49,610
push people to to follow,

1278
00:46:50,309 --> 00:46:52,470
certain trajectories. But but there is more. I

1279
00:46:52,470 --> 00:46:53,289
mean, there are

1280
00:46:53,670 --> 00:46:55,769
things that are very unexpected. And,

1281
00:46:56,630 --> 00:46:59,349
I mean, let's say at times, you are

1282
00:46:59,349 --> 00:47:01,590
maybe in a train station and and and

1283
00:47:01,590 --> 00:47:04,085
you are annoyed by the fact that your

1284
00:47:04,085 --> 00:47:04,984
mobile phone

1285
00:47:06,405 --> 00:47:08,585
is is is not connecting to the Internet.

1286
00:47:08,964 --> 00:47:10,885
And this is not because there was no

1287
00:47:10,885 --> 00:47:13,284
Internet there, but because the signal is locally

1288
00:47:13,284 --> 00:47:13,784
jammed

1289
00:47:14,405 --> 00:47:16,244
to make sure to make sure that people

1290
00:47:16,244 --> 00:47:17,864
don't clog in a

1291
00:47:18,200 --> 00:47:19,579
in a given, location.

1292
00:47:20,599 --> 00:47:22,360
Let's say we don't want people to stand

1293
00:47:22,360 --> 00:47:24,920
here because maybe the the the flow would

1294
00:47:24,920 --> 00:47:26,619
would would clog. Right? Imagine,

1295
00:47:27,320 --> 00:47:30,119
there is a staircase perhaps, and then people

1296
00:47:30,119 --> 00:47:32,200
would would climb the staircase and just wait

1297
00:47:32,200 --> 00:47:33,320
there. But obviously

1298
00:47:34,125 --> 00:47:36,364
and let's say that this this staircase leads

1299
00:47:36,364 --> 00:47:38,445
to to the platform of a train station.

1300
00:47:38,445 --> 00:47:40,525
Yeah? And, obviously, you don't want people just

1301
00:47:40,525 --> 00:47:42,204
to to climb the stairs and and wait

1302
00:47:42,204 --> 00:47:44,045
there. You would like them to distribute on

1303
00:47:44,045 --> 00:47:44,625
the platform.

1304
00:47:45,244 --> 00:47:47,744
And, so how do you ensure that?

1305
00:47:48,219 --> 00:47:50,380
Then you would locally jam the mobile phone

1306
00:47:50,380 --> 00:47:51,920
signal in front of the staircase.

1307
00:47:53,099 --> 00:47:55,820
And now and now, you know, the person

1308
00:47:55,820 --> 00:47:57,820
would would climb the stairs and then they

1309
00:47:57,820 --> 00:47:59,579
would just stand there and get out their

1310
00:47:59,579 --> 00:48:01,599
mobile phone and then they figure, oh,

1311
00:48:01,925 --> 00:48:02,824
it's not working.

1312
00:48:03,204 --> 00:48:05,045
Right? And then I move around to to

1313
00:48:05,045 --> 00:48:07,065
get it to work. But, yeah,

1314
00:48:07,525 --> 00:48:09,945
it was not it wasn't working not because

1315
00:48:10,324 --> 00:48:12,324
there was no signal. It was on purpose.

1316
00:48:12,324 --> 00:48:14,760
You know, you know and then it was

1317
00:48:14,760 --> 00:48:16,880
not the efficiency of the mobile phone company.

1318
00:48:16,880 --> 00:48:19,000
It was rather by design that you were

1319
00:48:19,000 --> 00:48:19,239
not,

1320
00:48:19,960 --> 00:48:21,960
That's amazing. Is that something that's going to

1321
00:48:21,960 --> 00:48:22,460
happen?

1322
00:48:23,000 --> 00:48:24,840
No. This is something that is happening already.

1323
00:48:24,840 --> 00:48:26,860
Really? Tell you this. Yes. Genuinely.

1324
00:48:27,160 --> 00:48:30,265
People are using people are blocking that. So,

1325
00:48:30,265 --> 00:48:32,344
basically, at halftime in the football, right, I

1326
00:48:32,344 --> 00:48:33,625
I sit there and try and check the

1327
00:48:33,625 --> 00:48:34,684
scores from elsewhere,

1328
00:48:35,065 --> 00:48:37,304
and it it you just can't because everybody's

1329
00:48:37,304 --> 00:48:39,784
doing exactly the same thing. Happily, in my

1330
00:48:39,784 --> 00:48:40,284
own,

1331
00:48:41,429 --> 00:48:43,190
football club that I support, they provide Wi

1332
00:48:43,190 --> 00:48:45,269
Fi, which does the job quite nicely. And

1333
00:48:45,269 --> 00:48:46,809
and they put the scores on the screen.

1334
00:48:46,949 --> 00:48:49,349
But there it does happen that it's blocked

1335
00:48:49,349 --> 00:48:51,030
because there's a number of people in it

1336
00:48:51,030 --> 00:48:52,469
as well. Right? Just so people are listening

1337
00:48:52,469 --> 00:48:54,150
and not thinking every time it's blocked, it's

1338
00:48:54,150 --> 00:48:56,150
some mobile phone company trying to move them

1339
00:48:56,150 --> 00:48:56,445
on.

1340
00:48:57,005 --> 00:48:58,364
Well, no. No. No. I mean, like, yeah.

1341
00:48:58,364 --> 00:48:59,644
Of course of course, there you have a

1342
00:48:59,644 --> 00:49:01,485
you have a network capacity issue. I I

1343
00:49:01,485 --> 00:49:03,885
think these these are different things. Here, I

1344
00:49:03,885 --> 00:49:05,565
mean, there is a possibility of doing very

1345
00:49:05,565 --> 00:49:07,644
local jamming. Right? The the the within a

1346
00:49:07,644 --> 00:49:11,339
within a very limited area, mobile phone is

1347
00:49:11,339 --> 00:49:11,579
is

1348
00:49:12,139 --> 00:49:14,460
reception is not is not good. But, again,

1349
00:49:14,460 --> 00:49:16,559
this is this is already in in place.

1350
00:49:16,779 --> 00:49:18,400
And, that's extraordinary.

1351
00:49:18,859 --> 00:49:20,940
That's extraordinary. Well, can you tell me just

1352
00:49:20,940 --> 00:49:22,799
going back to the light bit, the illumination

1353
00:49:22,940 --> 00:49:24,319
bit in the museum? Right.

1354
00:49:24,855 --> 00:49:26,375
How does that would we are you sort

1355
00:49:26,375 --> 00:49:28,775
of eliminating particular bits to draw people in?

1356
00:49:28,775 --> 00:49:30,454
Are you dimming bits? Yeah. We we we

1357
00:49:30,454 --> 00:49:32,554
ran a couple of of big experiments,

1358
00:49:33,815 --> 00:49:35,175
around here. Is,

1359
00:49:36,775 --> 00:49:37,594
is the question,

1360
00:49:38,960 --> 00:49:40,659
can we use light

1361
00:49:41,119 --> 00:49:41,619
to

1362
00:49:42,079 --> 00:49:43,059
to to provide

1363
00:49:43,679 --> 00:49:44,179
information

1364
00:49:44,480 --> 00:49:45,119
that is,

1365
00:49:45,759 --> 00:49:46,719
to convey, let's say

1366
00:49:47,440 --> 00:49:48,559
I I don't want to use the word

1367
00:49:48,559 --> 00:49:50,099
subtle, but to convey,

1368
00:49:50,559 --> 00:49:52,019
let's say, more and more implicit

1369
00:49:52,400 --> 00:49:53,920
information. And it works,

1370
00:49:54,465 --> 00:49:56,065
or at least the experiment that we did,

1371
00:49:56,305 --> 00:49:58,164
were based on on on dimming.

1372
00:49:58,785 --> 00:49:59,184
And,

1373
00:49:59,744 --> 00:50:01,525
we showed indeed the following

1374
00:50:01,985 --> 00:50:02,485
that,

1375
00:50:04,385 --> 00:50:06,485
imagine you have you have this big crowd.

1376
00:50:06,545 --> 00:50:09,179
We did this, in in this very experiment

1377
00:50:09,179 --> 00:50:10,940
that I'm referring to was that was run

1378
00:50:10,940 --> 00:50:12,780
during a big festival that we have here

1379
00:50:12,780 --> 00:50:13,440
in Andover.

1380
00:50:13,739 --> 00:50:14,559
So it's it's,

1381
00:50:15,579 --> 00:50:17,739
city walk throughout the entire city, let's say.

1382
00:50:17,739 --> 00:50:18,880
So, again,

1383
00:50:19,579 --> 00:50:21,039
tens of thousands of people.

1384
00:50:21,505 --> 00:50:24,385
And now imagine these these people all entering

1385
00:50:24,385 --> 00:50:26,085
into a corridor, then,

1386
00:50:26,625 --> 00:50:28,625
we wanted them to go, let's say, to

1387
00:50:28,625 --> 00:50:31,204
choose either the left or the right exit.

1388
00:50:32,224 --> 00:50:32,724
And,

1389
00:50:34,489 --> 00:50:37,369
in our experiment, we compared our effective it

1390
00:50:37,369 --> 00:50:39,609
is to show people, for instance, a normal

1391
00:50:39,609 --> 00:50:42,089
signage, so big arrow that says, okay. Go

1392
00:50:42,089 --> 00:50:43,789
left rather than go right.

1393
00:50:44,250 --> 00:50:44,750
Or

1394
00:50:45,609 --> 00:50:48,805
rather to change the illumination level between the

1395
00:50:48,805 --> 00:50:51,445
two ends or the two exits. Yeah. And

1396
00:50:51,445 --> 00:50:52,825
then we show that that,

1397
00:50:53,204 --> 00:50:54,184
within some,

1398
00:50:54,644 --> 00:50:56,405
let's say, crowd density limit,

1399
00:50:56,965 --> 00:50:59,204
you can have more or less the same

1400
00:50:59,204 --> 00:50:59,704
effectiveness

1401
00:51:00,164 --> 00:51:00,565
in,

1402
00:51:01,045 --> 00:51:01,545
in

1403
00:51:02,619 --> 00:51:04,139
light and signage have more or less the

1404
00:51:04,139 --> 00:51:04,800
same effectiveness.

1405
00:51:05,099 --> 00:51:07,359
And then it happens that as cloud density,

1406
00:51:08,539 --> 00:51:09,039
grows,

1407
00:51:09,579 --> 00:51:10,800
as as you can expect,

1408
00:51:11,260 --> 00:51:13,340
there is no indication that is really followed.

1409
00:51:13,340 --> 00:51:14,859
It doesn't matter whether it is based on

1410
00:51:14,859 --> 00:51:16,695
light or signage. It's it's like,

1411
00:51:17,175 --> 00:51:19,015
I mean, people at certain point when there

1412
00:51:19,015 --> 00:51:21,094
are these many in a confined space, they

1413
00:51:21,094 --> 00:51:22,535
just need to get out. So it doesn't

1414
00:51:22,535 --> 00:51:24,695
matter whether there is an indication that points

1415
00:51:24,695 --> 00:51:25,835
there them somewhere.

1416
00:51:26,535 --> 00:51:28,235
But but, let's say, within

1417
00:51:28,855 --> 00:51:29,355
operational,

1418
00:51:29,894 --> 00:51:31,914
density level that you can find

1419
00:51:32,300 --> 00:51:33,739
in museums and so on and so forth.

1420
00:51:33,739 --> 00:51:35,900
So in our experiment, we could show that

1421
00:51:35,900 --> 00:51:38,139
that you can convey the same type of,

1422
00:51:40,059 --> 00:51:41,980
signal or you can convey a signal that

1423
00:51:41,980 --> 00:51:43,840
that has the same effect on people,

1424
00:51:44,539 --> 00:51:45,440
either by using

1425
00:51:46,734 --> 00:51:48,894
variable light intensity or or,

1426
00:51:49,534 --> 00:51:51,054
signage. Could you tell me a bit more

1427
00:51:51,054 --> 00:51:53,135
as well about the the the audio version?

1428
00:51:53,135 --> 00:51:54,655
Do do you have the melody? In the

1429
00:51:54,655 --> 00:51:56,335
paper, you could find the 4 the 4

1430
00:51:56,335 --> 00:51:56,835
tunes.

1431
00:51:57,375 --> 00:51:58,335
It was it was,

1432
00:51:59,054 --> 00:51:59,554
really

1433
00:52:00,335 --> 00:52:03,269
piano chords. Yeah. Oh, okay. I'm literally sitting

1434
00:52:03,269 --> 00:52:05,030
at a piano right now. So Okay. So

1435
00:52:05,430 --> 00:52:06,789
we could see. Yeah. I I I yeah.

1436
00:52:06,789 --> 00:52:08,710
I I think yeah. If you want, I

1437
00:52:08,710 --> 00:52:10,550
can I can look for it even right

1438
00:52:10,550 --> 00:52:12,390
now? But it's, let's say in this paper,

1439
00:52:12,390 --> 00:52:13,849
then then you would see that

1440
00:52:14,175 --> 00:52:16,114
we divided the space into

1441
00:52:16,414 --> 00:52:19,695
areas, and then each area, each small area,

1442
00:52:19,695 --> 00:52:22,175
something like maybe half a meter times half

1443
00:52:22,175 --> 00:52:22,835
a meter,

1444
00:52:23,375 --> 00:52:24,335
came with different,

1445
00:52:24,735 --> 00:52:25,235
piano

1446
00:52:26,414 --> 00:52:27,855
tune. Right? With a different,

1447
00:52:28,630 --> 00:52:31,429
and then this was played one after the

1448
00:52:31,429 --> 00:52:32,650
other should you be,

1449
00:52:33,030 --> 00:52:33,530
walking

1450
00:52:34,869 --> 00:52:37,590
along the right path. Yes. Otherwise, you you

1451
00:52:37,590 --> 00:52:39,530
wouldn't hear anything. And,

1452
00:52:40,605 --> 00:52:42,204
okay. I'm not a music person. Right? So

1453
00:52:42,204 --> 00:52:43,804
now I would say something that I hope

1454
00:52:43,804 --> 00:52:45,405
makes sense to you because it doesn't make

1455
00:52:45,405 --> 00:52:46,625
much sense to me.

1456
00:52:47,244 --> 00:52:48,784
The sequence was c4c5

1457
00:52:50,125 --> 00:52:50,784
and thene4e5

1458
00:52:52,045 --> 00:52:53,440
and then g4g5andc5c6.

1459
00:52:55,579 --> 00:52:57,900
So should you be following the right path,

1460
00:52:57,900 --> 00:52:59,519
you would be hearing this.

1461
00:53:09,844 --> 00:53:11,925
As when you're walking around in crowds, are

1462
00:53:11,925 --> 00:53:14,025
you sort of in a kind of Sherlock

1463
00:53:14,085 --> 00:53:17,285
mind palace with equations appearing in your peripheral

1464
00:53:17,285 --> 00:53:18,724
vision as you're walking, or are you just

1465
00:53:18,724 --> 00:53:20,470
kind of wondering how can we make this

1466
00:53:20,470 --> 00:53:22,890
better? Walking around is is mostly like,

1467
00:53:24,230 --> 00:53:26,309
about how to improve the models. So how

1468
00:53:26,309 --> 00:53:27,130
to include,

1469
00:53:28,390 --> 00:53:30,550
let's say, more factors in the models that,

1470
00:53:30,789 --> 00:53:32,390
that that we are using or that we

1471
00:53:32,390 --> 00:53:33,050
are making.

1472
00:53:34,295 --> 00:53:35,114
And I think,

1473
00:53:36,855 --> 00:53:37,755
yeah. I

1474
00:53:39,094 --> 00:53:40,614
mean so that there are very

1475
00:53:40,934 --> 00:53:43,255
there are a bunch of very interesting physical

1476
00:53:43,255 --> 00:53:45,734
directions that are let me make an example

1477
00:53:45,734 --> 00:53:47,734
that is more telling than others in in

1478
00:53:47,734 --> 00:53:50,039
in my view. That is, for instance, when

1479
00:53:50,039 --> 00:53:51,980
when a crowd walk, not necessarily

1480
00:53:52,360 --> 00:53:53,739
the interaction is cooperative.

1481
00:53:54,119 --> 00:53:56,200
Right? So imagine that that you would like

1482
00:53:56,200 --> 00:53:59,739
to to get, onto onto a train. Right?

1483
00:54:00,119 --> 00:54:02,235
And you would like to to be

1484
00:54:02,775 --> 00:54:04,394
inside the train first.

1485
00:54:06,375 --> 00:54:07,275
And this

1486
00:54:08,295 --> 00:54:10,775
eventually doesn't mean, for instance, that the people

1487
00:54:10,775 --> 00:54:12,775
that need to get into the train with

1488
00:54:12,775 --> 00:54:13,275
you,

1489
00:54:15,014 --> 00:54:15,514
altogether

1490
00:54:17,109 --> 00:54:19,510
will manage to get in in the most

1491
00:54:19,510 --> 00:54:20,409
efficient way.

1492
00:54:20,949 --> 00:54:22,010
Right? Because

1493
00:54:22,630 --> 00:54:23,130
typically,

1494
00:54:23,750 --> 00:54:26,550
the the behavior of of each particle here

1495
00:54:26,550 --> 00:54:27,690
is is competitive

1496
00:54:27,989 --> 00:54:31,074
instead of constructive. Right? So and then the

1497
00:54:31,074 --> 00:54:31,974
question arises,

1498
00:54:32,275 --> 00:54:33,954
so how to up deal with it, how

1499
00:54:33,954 --> 00:54:36,214
to model this? So what is the underlying

1500
00:54:36,275 --> 00:54:38,114
physics? How can we represent it? I think

1501
00:54:38,114 --> 00:54:39,554
this is, for instance, a very interesting,

1502
00:54:40,194 --> 00:54:40,994
a very interesting,

1503
00:54:41,315 --> 00:54:42,614
direction that is,

1504
00:54:43,210 --> 00:54:45,530
how to to consider this particle system in

1505
00:54:45,530 --> 00:54:46,430
which, which,

1506
00:54:46,969 --> 00:54:50,010
there is an interplay between between competition and

1507
00:54:50,010 --> 00:54:51,070
and and cooperation,

1508
00:54:52,010 --> 00:54:54,349
as the the context changes.

1509
00:54:54,650 --> 00:54:56,250
And so for instance, this is one of

1510
00:54:56,250 --> 00:54:58,784
the things that I'm asking myself lately. So

1511
00:54:59,264 --> 00:55:01,264
what are efficient way to deal with this

1512
00:55:01,264 --> 00:55:03,505
and then then the randomness in the crowds

1513
00:55:03,505 --> 00:55:05,284
and and and so on.

1514
00:55:05,585 --> 00:55:07,984
And about about the interest in general, I

1515
00:55:07,984 --> 00:55:08,484
think

1516
00:55:09,344 --> 00:55:11,045
I got into crowds because

1517
00:55:12,170 --> 00:55:14,329
I wanted to do I wanted to explore

1518
00:55:14,329 --> 00:55:16,829
a part of physics that had two characteristics.

1519
00:55:17,450 --> 00:55:18,590
And one that was,

1520
00:55:19,690 --> 00:55:20,190
close

1521
00:55:20,570 --> 00:55:22,890
to people, like, that was, that I could

1522
00:55:22,890 --> 00:55:23,390
communicate

1523
00:55:25,105 --> 00:55:25,925
easily about,

1524
00:55:27,025 --> 00:55:28,805
that was immediately impactful

1525
00:55:29,585 --> 00:55:30,244
in society.

1526
00:55:31,184 --> 00:55:32,724
That was new. Right?

1527
00:55:33,265 --> 00:55:35,364
Where new means that, that research,

1528
00:55:38,329 --> 00:55:39,150
let's say,

1529
00:55:39,449 --> 00:55:41,630
started not so long ago. I mean, for

1530
00:55:42,090 --> 00:55:44,090
crowds, I mean, obviously, I mean, crowds, you

1531
00:55:44,090 --> 00:55:45,929
know, have been I mean, people have been

1532
00:55:45,929 --> 00:55:48,170
trying to optimize, crowd flow from the times

1533
00:55:48,170 --> 00:55:50,684
of the Colosseum in Rome. Right? But nevertheless,

1534
00:55:51,065 --> 00:55:53,464
let's say physicists started to look very seriously

1535
00:55:53,464 --> 00:55:55,944
at crowds only since the beginning of nineties,

1536
00:55:55,944 --> 00:55:58,025
and the first experiments were since the beginning

1537
00:55:58,025 --> 00:55:59,864
of 2000. So I would say that this

1538
00:55:59,864 --> 00:56:01,244
is relatively new.

1539
00:56:01,625 --> 00:56:03,464
Wanted to do something something new and it's

1540
00:56:03,464 --> 00:56:04,605
also something that

1541
00:56:05,809 --> 00:56:08,530
being new enables to I mean, needs to

1542
00:56:08,530 --> 00:56:10,449
couple a lot of things. So how to

1543
00:56:10,449 --> 00:56:12,690
do experiment is still an open question. How

1544
00:56:12,690 --> 00:56:14,449
to model is an open question. How to

1545
00:56:14,449 --> 00:56:16,170
control is an open question. So it has

1546
00:56:16,170 --> 00:56:18,769
a very, very broad range of open questions.

1547
00:56:18,769 --> 00:56:20,230
And this gave me the

1548
00:56:20,684 --> 00:56:23,244
the opportunity to to to bridge many things

1549
00:56:23,244 --> 00:56:25,405
from the experimental part to the modeling part.

1550
00:56:25,405 --> 00:56:26,045
And this is,

1551
00:56:26,525 --> 00:56:28,765
yeah, also one of the reasons I enjoy

1552
00:56:28,765 --> 00:56:29,265
this.

1553
00:56:29,644 --> 00:56:31,724
I'm not sure whether I project the equations

1554
00:56:31,724 --> 00:56:32,945
in my mind, but

1555
00:56:33,324 --> 00:56:35,369
it goes in that direction for sure. I've

1556
00:56:35,530 --> 00:56:37,130
never spoken to anyone who's won the Ig

1557
00:56:37,130 --> 00:56:39,369
Nobel Prize before. How does that come about,

1558
00:56:39,369 --> 00:56:40,430
and how does it feel?

1559
00:56:41,530 --> 00:56:41,849
Well,

1560
00:56:42,570 --> 00:56:44,730
it works, in in the in the following

1561
00:56:44,730 --> 00:56:46,730
way that that you receive a mail from

1562
00:56:46,730 --> 00:56:47,550
some colleague

1563
00:56:48,010 --> 00:56:50,170
that you might or might not know that

1564
00:56:50,170 --> 00:56:50,670
tells.

1565
00:56:51,605 --> 00:56:52,105
Well,

1566
00:56:52,885 --> 00:56:54,885
well, typically, I think you you you don't

1567
00:56:54,885 --> 00:56:56,804
know the colleague and then, you receive this

1568
00:56:56,804 --> 00:56:57,304
email

1569
00:56:57,684 --> 00:57:00,005
this morning that says, hey. I read your

1570
00:57:00,005 --> 00:57:00,505
work.

1571
00:57:01,844 --> 00:57:03,764
It's very interesting. I'd like to ask you

1572
00:57:03,764 --> 00:57:05,764
some question. And then and then,

1573
00:57:07,160 --> 00:57:09,340
it goes more or less like, the famous

1574
00:57:09,400 --> 00:57:11,559
scene of the matrix, you know, the the

1575
00:57:11,559 --> 00:57:14,119
blue and red pill thing. So, okay, I'd

1576
00:57:14,119 --> 00:57:16,200
like to tell you that that, you got,

1577
00:57:16,920 --> 00:57:19,720
I mean, you have you won this, Nobel

1578
00:57:19,720 --> 00:57:20,940
Prize and then,

1579
00:57:23,445 --> 00:57:25,445
but you can talk with your colleagues and

1580
00:57:25,445 --> 00:57:27,465
decide whether or not to accept the price.

1581
00:57:27,684 --> 00:57:30,485
And then, you you, yeah, you contact again

1582
00:57:30,485 --> 00:57:32,325
this person that that wrote you in the

1583
00:57:32,325 --> 00:57:33,765
1st place and you go for the red

1584
00:57:33,765 --> 00:57:34,965
pill. And then it becomes

1585
00:57:35,420 --> 00:57:38,460
yeah, afterwards, it becomes obviously extremely interesting because

1586
00:57:38,460 --> 00:57:39,039
it's a

1587
00:57:39,660 --> 00:57:40,480
great opportunity

1588
00:57:40,780 --> 00:57:43,900
to to discuss with with very, very wide

1589
00:57:43,900 --> 00:57:45,280
audience from scientists

1590
00:57:45,579 --> 00:57:46,480
to to

1591
00:57:47,019 --> 00:57:49,840
popular science channel about about your work. Okay.

1592
00:57:49,905 --> 00:57:52,224
But so is that why you decided to

1593
00:57:52,224 --> 00:57:54,545
accept the prize because of that opportunity for

1594
00:57:54,545 --> 00:57:55,925
communicating about your work?

1595
00:57:56,465 --> 00:57:58,864
Well, this is definitely well, because it's a

1596
00:57:58,864 --> 00:58:00,644
prize. Right? So we're not. And

1597
00:58:01,265 --> 00:58:02,565
and, but also,

1598
00:58:03,025 --> 00:58:06,109
I mean, it's I think, it's an unprecedented

1599
00:58:06,409 --> 00:58:07,769
opportunity. I mean, I don't think,

1600
00:58:08,730 --> 00:58:10,329
yeah, I don't think, anyone,

1601
00:58:10,730 --> 00:58:12,969
could let it slip, honestly. Yeah. No. It's

1602
00:58:12,969 --> 00:58:14,250
a it's a it's a lovely thing. I

1603
00:58:14,250 --> 00:58:16,010
mean, my understanding of it, right, is is

1604
00:58:16,010 --> 00:58:18,909
that it's interesting research that also makes you

1605
00:58:18,969 --> 00:58:19,869
laugh. Right?

1606
00:58:20,224 --> 00:58:21,824
Yes. Yeah. I mean, the the I think

1607
00:58:21,824 --> 00:58:24,085
the line by by Mark Abrahams is,

1608
00:58:24,784 --> 00:58:27,125
research that makes you laugh and then think.

1609
00:58:27,505 --> 00:58:29,985
And, I think the beauty of this, global

1610
00:58:29,985 --> 00:58:30,704
prize is that,

1611
00:58:31,664 --> 00:58:33,605
every year there are there are discoveries

1612
00:58:33,985 --> 00:58:36,320
that, I mean, at first sight,

1613
00:58:37,180 --> 00:58:38,539
they they might look,

1614
00:58:39,260 --> 00:58:41,579
they might not look about science. Right? Because,

1615
00:58:42,140 --> 00:58:44,559
you know, maybe seen by the general public,

1616
00:58:44,619 --> 00:58:45,760
science needs to be

1617
00:58:46,140 --> 00:58:47,280
extremely complex,

1618
00:58:47,660 --> 00:58:49,375
very far from what we are doing in

1619
00:58:49,375 --> 00:58:51,375
our daily life, but, often it's not the

1620
00:58:51,375 --> 00:58:51,875
case.

1621
00:58:52,414 --> 00:58:52,815
And,

1622
00:58:54,974 --> 00:58:57,454
and, and, therefore, this is something that that

1623
00:58:57,454 --> 00:59:00,255
people would find surprising. Therefore, in many cases,

1624
00:59:00,255 --> 00:59:01,889
it can make you laugh. There are,

1625
00:59:02,849 --> 00:59:05,170
but then definitely, it's it's a channel for

1626
00:59:05,170 --> 00:59:07,429
having people thinking about surprising,

1627
00:59:09,409 --> 00:59:11,969
let's say, surprising emergence of science where they

1628
00:59:11,969 --> 00:59:12,789
wouldn't expect.

1629
00:59:13,089 --> 00:59:14,849
I think this is a growing field. I

1630
00:59:14,849 --> 00:59:15,829
mean, the cryodynamics

1631
00:59:16,130 --> 00:59:17,190
is a growing field

1632
00:59:18,224 --> 00:59:18,964
in physics,

1633
00:59:19,824 --> 00:59:21,045
engineering in general.

1634
00:59:22,464 --> 00:59:24,244
It's something that is definitely

1635
00:59:24,545 --> 00:59:25,605
challenging scientifically,

1636
00:59:26,065 --> 00:59:28,644
but also impactful in society as,

1637
00:59:29,505 --> 00:59:32,400
in in places that are very densely populated,

1638
00:59:32,699 --> 00:59:34,539
the Netherlands, UK, and so on and so

1639
00:59:34,539 --> 00:59:36,160
forth. It often happens that

1640
00:59:38,300 --> 00:59:40,940
you the the pressure on public infrastructure like

1641
00:59:40,940 --> 00:59:43,179
the train station grows, by the year, so

1642
00:59:43,179 --> 00:59:44,640
there are more people moving.

1643
00:59:44,974 --> 00:59:45,715
But often,

1644
00:59:46,015 --> 00:59:47,954
it's becoming impossible to extend

1645
00:59:48,894 --> 00:59:49,715
train stations

1646
00:59:50,335 --> 00:59:53,375
to to, cope with increased load. And so

1647
00:59:53,375 --> 00:59:54,594
now the only way

1648
00:59:55,135 --> 00:59:57,315
is is making the environment smarter,

1649
00:59:58,190 --> 01:00:00,030
And this is, one of the reason we

1650
01:00:00,030 --> 01:00:00,849
are doing this.

1651
01:00:01,869 --> 01:00:04,190
I'd like to thank Alessandro and Philip for

1652
01:00:04,190 --> 01:00:05,630
talking to me for this episode of the

1653
01:00:05,630 --> 01:00:07,010
Physics World Stories podcast.

1654
01:00:07,470 --> 01:00:09,470
And, of course, we'll post links to their

1655
01:00:09,470 --> 01:00:11,474
work on the Physics World website,

1656
01:00:12,035 --> 01:00:12,934
physics world.com.

1657
01:00:14,035 --> 01:00:17,175
Next month, we'll be looking at Pele's tears,

1658
01:00:17,315 --> 01:00:19,335
which have more to do with volcanoes

1659
01:00:19,954 --> 01:00:20,695
than football.

1660
01:00:20,994 --> 01:00:23,815
Until then, thank you very much for listening.