1
00:00:00,690 --> 00:00:03,700
GUEST SPEAKER: It's a
great pleasure to present--

2
00:00:03,700 --> 00:00:04,200
thank you.

3
00:00:04,200 --> 00:00:05,955
Thank you, Steve--
at this conference,

4
00:00:05,955 --> 00:00:07,080
and it's a great gathering.

5
00:00:07,080 --> 00:00:12,420
And I thank Steve and John,
because I feel the fire again.

6
00:00:12,420 --> 00:00:17,430
The Auto-ID center went through
a period of similar ignition,

7
00:00:17,430 --> 00:00:18,870
and it was fascinating.

8
00:00:18,870 --> 00:00:21,540
It is like a Petri
dish with things,

9
00:00:21,540 --> 00:00:23,340
new life forms were being born.

10
00:00:23,340 --> 00:00:27,390
And then we moved into the
execution phase and EPCglobal

11
00:00:27,390 --> 00:00:31,020
of whom the many representatives
here started executing.

12
00:00:31,020 --> 00:00:36,390
And meanwhile, the Auto-ID
Lab refocused their attentions

13
00:00:36,390 --> 00:00:39,630
on new research, and we
have a family of Auto-ID Lab

14
00:00:39,630 --> 00:00:43,690
members, professors,
students, colleagues here, who

15
00:00:43,690 --> 00:00:46,350
have recreated this entire
environment and Elgar Fleisch

16
00:00:46,350 --> 00:00:49,052
and John [? Marai ?] and
John Williams and Peter Cole,

17
00:00:49,052 --> 00:00:51,510
of course, who's been involved
with us for a long time, Dan

18
00:00:51,510 --> 00:00:52,390
Engels.

19
00:00:52,390 --> 00:00:56,670
It's fantastic to see this
all come together again, Hal.

20
00:00:56,670 --> 00:00:57,732
So it's a great pleasure.

21
00:00:57,732 --> 00:00:59,940
It's almost like-- I keep
saying this every time I go

22
00:00:59,940 --> 00:01:01,530
to a Auto-ID--

23
00:01:01,530 --> 00:01:02,880
it's like a wedding.

24
00:01:02,880 --> 00:01:05,129
It's like going back
to a family wedding

25
00:01:05,129 --> 00:01:07,800
and recognizing
all your friends.

26
00:01:07,800 --> 00:01:09,030
So it's a great pleasure.

27
00:01:09,030 --> 00:01:11,730
What I thought I would do
is talk about the future,

28
00:01:11,730 --> 00:01:14,070
as I see it, and
actually funnily enough,

29
00:01:14,070 --> 00:01:17,070
I didn't have to change
my slides from about seven

30
00:01:17,070 --> 00:01:19,952
years ago and just paint
the landscape of what

31
00:01:19,952 --> 00:01:21,660
it is that academics
that to do research.

32
00:01:21,660 --> 00:01:23,202
And I should tell
you a little story.

33
00:01:23,202 --> 00:01:25,673
When I started this research
seven or eight years ago--

34
00:01:25,673 --> 00:01:26,840
David Brock and I, actually.

35
00:01:26,840 --> 00:01:29,110
Actually, David Brock gets
a lot of credit for this--

36
00:01:29,110 --> 00:01:30,690
I wasn't sure that--

37
00:01:30,690 --> 00:01:31,920
I was coming up for tenure.

38
00:01:31,920 --> 00:01:35,830
I was trying to write papers
and make an academic field out

39
00:01:35,830 --> 00:01:37,150
of this.

40
00:01:37,150 --> 00:01:40,595
And as many of my former
students sitting here

41
00:01:40,595 --> 00:01:41,970
will tell you, I
wasn't sure this

42
00:01:41,970 --> 00:01:43,710
would be a tenurable topic.

43
00:01:43,710 --> 00:01:44,672
Could I write papers?

44
00:01:44,672 --> 00:01:45,630
Could this be academic?

45
00:01:45,630 --> 00:01:47,280
Is there some real
researcher here,

46
00:01:47,280 --> 00:01:49,058
or am I just slapping
tags on things

47
00:01:49,058 --> 00:01:50,850
and reading them and
feeling good about it?

48
00:01:50,850 --> 00:01:51,480
Right?

49
00:01:51,480 --> 00:01:55,740
And has all the technology
development been done?

50
00:01:55,740 --> 00:01:59,190
And we as a group grabbed
the tiger by the tail,

51
00:01:59,190 --> 00:02:01,080
and the tiger took
us to places that we

52
00:02:01,080 --> 00:02:05,070
didn't know that even existed.

53
00:02:05,070 --> 00:02:06,900
And what I discovered
in that process

54
00:02:06,900 --> 00:02:08,900
and-- perhaps this applies
to other technologies

55
00:02:08,900 --> 00:02:11,070
as well-- is I think one
needs to understand that,

56
00:02:11,070 --> 00:02:13,200
in any new technology
and especially in RFID,

57
00:02:13,200 --> 00:02:15,930
there are embedded
layers of richness

58
00:02:15,930 --> 00:02:18,840
that are revealed as you
start making things practical,

59
00:02:18,840 --> 00:02:22,620
and the beauty is that they
come in different colors.

60
00:02:22,620 --> 00:02:24,212
It could be in security.

61
00:02:24,212 --> 00:02:25,170
It could mean protocol.

62
00:02:25,170 --> 00:02:27,450
There's so many
areas of richness

63
00:02:27,450 --> 00:02:28,862
that are waiting to be explored.

64
00:02:28,862 --> 00:02:31,320
That's the first thing, and
the second point I want to make

65
00:02:31,320 --> 00:02:33,810
was that we are very much at
the beginning of the journey,

66
00:02:33,810 --> 00:02:35,485
and that's kind of
where I'll start.

67
00:02:35,485 --> 00:02:37,110
So there is a lot of
self-satisfaction,

68
00:02:37,110 --> 00:02:39,630
which is I think in some ways
deserved, that oh, we're there.

69
00:02:39,630 --> 00:02:40,505
Let's start doing it.

70
00:02:40,505 --> 00:02:42,187
And that's great, we
should be doing it.

71
00:02:42,187 --> 00:02:44,520
But I have to say that we are
very much at the beginning

72
00:02:44,520 --> 00:02:45,160
of this journey.

73
00:02:45,160 --> 00:02:45,660
Thank you.

74
00:02:50,470 --> 00:02:52,870
We're very much at the
beginning of this journey.

75
00:02:52,870 --> 00:02:55,120
RFID been around for 50
years, and the Auto-ID center

76
00:02:55,120 --> 00:02:57,010
is only eight years old.

77
00:02:57,010 --> 00:02:59,205
But it would be very
arrogant for anyone

78
00:02:59,205 --> 00:03:01,330
to suggest that somehow
the research has been done.

79
00:03:01,330 --> 00:03:04,630
There's just so much
exciting stuff to be done.

80
00:03:04,630 --> 00:03:08,140
And if you look at the history,
the first papers in this--

81
00:03:08,140 --> 00:03:09,808
and I did a reference search.

82
00:03:09,808 --> 00:03:11,350
By the way, there's
a wonderful paper

83
00:03:11,350 --> 00:03:14,410
called the "Shrouds of Time:
The History of RFID" by Landt.

84
00:03:14,410 --> 00:03:17,860
It was published as
an AIM publication.

85
00:03:17,860 --> 00:03:20,280
If you do a search on
Google, you'll find it.

86
00:03:20,280 --> 00:03:22,305
It captures the
history very well.

87
00:03:22,305 --> 00:03:23,680
But if you do a
reference search,

88
00:03:23,680 --> 00:03:25,513
the first papers
came out in '48.

89
00:03:25,513 --> 00:03:27,430
Peter Cole, who's sitting
right up front here,

90
00:03:27,430 --> 00:03:31,050
was involved with many of the
early implementations of RFID

91
00:03:31,050 --> 00:03:34,150
in the '70s, automotive
license plates.

92
00:03:34,150 --> 00:03:35,380
Is that about right, Peter?

93
00:03:35,380 --> 00:03:38,440
Is that about the first
reference began to appear.

94
00:03:38,440 --> 00:03:44,860
And then RFID went into this
highly specialized research

95
00:03:44,860 --> 00:03:45,850
phase.

96
00:03:45,850 --> 00:03:49,000
We as a center started
in '98, and then we

97
00:03:49,000 --> 00:03:50,652
developed standards in 2001.

98
00:03:50,652 --> 00:03:53,110
We couldn't get the rest of
the industry to really join us,

99
00:03:53,110 --> 00:03:55,360
so we bucked the trend and
developed our own protocols

100
00:03:55,360 --> 00:03:57,370
with some innovative
startup companies,

101
00:03:57,370 --> 00:03:59,410
like Alien Technology
and Matrix,

102
00:03:59,410 --> 00:04:01,370
and those were the
Gen 1 protocols,

103
00:04:01,370 --> 00:04:04,450
and that's how we started
getting the industry going.

104
00:04:04,450 --> 00:04:08,950
But the reason I'm putting
that up there is that 2000's.

105
00:04:08,950 --> 00:04:10,510
This is 2006.

106
00:04:10,510 --> 00:04:13,120
It's not that long ago, and
it would be very arrogant

107
00:04:13,120 --> 00:04:15,760
for any of us to suggest that
somehow in five or six years

108
00:04:15,760 --> 00:04:20,290
we figured out everything that
one needs to know about RFID.

109
00:04:20,290 --> 00:04:23,440
And the first bulk tagging
is only a year old, folks.

110
00:04:23,440 --> 00:04:25,630
Really, it's only one year old.

111
00:04:25,630 --> 00:04:28,780
The first mandate really
hit in early 2005.

112
00:04:28,780 --> 00:04:31,180
Here we are, in January 2006.

113
00:04:31,180 --> 00:04:34,760
So there's a heck of a lot
of work waiting to be done.

114
00:04:34,760 --> 00:04:37,460
And what I did seven
or eight years ago is

115
00:04:37,460 --> 00:04:38,450
I created a landscape.

116
00:04:38,450 --> 00:04:41,810
I looked at every research topic
I could possibly do in RFID,

117
00:04:41,810 --> 00:04:44,480
because remember, I was trying
to get tenure, write papers,

118
00:04:44,480 --> 00:04:45,960
and I wanted to
create a landscape.

119
00:04:45,960 --> 00:04:47,960
And actually, that hasn't
fleshed out very much.

120
00:04:47,960 --> 00:04:49,585
I'll put that up now,
and I'll use that

121
00:04:49,585 --> 00:04:53,310
as a theme for the
rest of my talk.

122
00:04:53,310 --> 00:04:54,690
You can break it
up in many ways.

123
00:04:54,690 --> 00:04:56,800
You can look at technology,
and the beauty of RFID

124
00:04:56,800 --> 00:05:00,480
is it spans everything from
semiconductor manufacturing

125
00:05:00,480 --> 00:05:02,790
to systems research
and distributed

126
00:05:02,790 --> 00:05:04,540
discovery and things like that.

127
00:05:04,540 --> 00:05:05,520
So you have tags.

128
00:05:05,520 --> 00:05:06,540
You have semiconductors.

129
00:05:06,540 --> 00:05:07,860
You have antennae on tags.

130
00:05:07,860 --> 00:05:09,070
You have material constants.

131
00:05:09,070 --> 00:05:10,278
You have material properties.

132
00:05:10,278 --> 00:05:12,240
You have the interconnect
between the chip

133
00:05:12,240 --> 00:05:15,480
and an antenna by itself
is a significant PhD

134
00:05:15,480 --> 00:05:16,780
thesis waiting to happen.

135
00:05:16,780 --> 00:05:17,280
Right?

136
00:05:17,280 --> 00:05:19,170
Just that, right?

137
00:05:19,170 --> 00:05:23,160
You have packaging of the tag.

138
00:05:23,160 --> 00:05:26,370
You have how does the
tag react to packaging?

139
00:05:26,370 --> 00:05:30,150
We had a wonderful project
here that Dan Engels led on tag

140
00:05:30,150 --> 00:05:31,590
packaging and performance.

141
00:05:31,590 --> 00:05:32,580
You have protocols.

142
00:05:32,580 --> 00:05:34,440
Of course, the
Auto-ID center did

143
00:05:34,440 --> 00:05:37,890
a lot of, and a lot of experts
here, on tag protocols.

144
00:05:37,890 --> 00:05:40,590
We're barely scratching
the surface of protocols.

145
00:05:40,590 --> 00:05:43,530
We have Gen 1
which is Gen 1 UHF.

146
00:05:43,530 --> 00:05:44,400
We have some HF.

147
00:05:44,400 --> 00:05:47,460
Now we have Gen 2, but then
there's ultra wideband.

148
00:05:47,460 --> 00:05:50,480
There's all sorts of interesting
stuff just waiting to be done.

149
00:05:50,480 --> 00:05:53,978
I'll touch upon it in
the next few slides.

150
00:05:53,978 --> 00:05:55,020
You have reader antennae.

151
00:05:55,020 --> 00:05:56,070
Think about this.

152
00:05:56,070 --> 00:05:57,810
Everyone thinks of that--

153
00:05:57,810 --> 00:06:01,195
I should say, there's a lot of
thought that perhaps antenna,

154
00:06:01,195 --> 00:06:02,070
not that interesting.

155
00:06:02,070 --> 00:06:03,360
Maybe there's not so
much research there,

156
00:06:03,360 --> 00:06:04,152
but think about it.

157
00:06:04,152 --> 00:06:07,080
You have miles and miles
and miles of store shelves.

158
00:06:07,080 --> 00:06:09,300
How on Earth are you going
to read all the stuff

159
00:06:09,300 --> 00:06:10,230
on those shelves?

160
00:06:10,230 --> 00:06:11,370
Now, initially,
people are looking

161
00:06:11,370 --> 00:06:12,453
at portal-based solutions.

162
00:06:12,453 --> 00:06:14,540
You put readers and doors,
but eventually, you're

163
00:06:14,540 --> 00:06:17,123
going to want to get real-time
inventory, not based on transit

164
00:06:17,123 --> 00:06:19,680
but on a real integrated
measure that is

165
00:06:19,680 --> 00:06:21,268
a measure of actual inventory.

166
00:06:21,268 --> 00:06:23,310
How are you going to put
readers on these things?

167
00:06:23,310 --> 00:06:26,310
Is it going to be
cheap antennae which

168
00:06:26,310 --> 00:06:29,190
multiplex using some
sort of printed ink

169
00:06:29,190 --> 00:06:32,760
multiplexing over a chip
wireless to a reader?

170
00:06:32,760 --> 00:06:35,307
Is that what's going to happen?

171
00:06:35,307 --> 00:06:36,390
Then, how do you power it?

172
00:06:36,390 --> 00:06:38,598
If you put readers everywhere,
how do you power them?

173
00:06:38,598 --> 00:06:39,793
Right?

174
00:06:39,793 --> 00:06:41,460
If you walk into a
retail store, and you

175
00:06:41,460 --> 00:06:43,002
see a display of
clothes, you will

176
00:06:43,002 --> 00:06:44,460
find that the
clothes sometimes are

177
00:06:44,460 --> 00:06:45,450
right in the middle
of the floor.

178
00:06:45,450 --> 00:06:46,367
How do you power them?

179
00:06:46,367 --> 00:06:49,410
Is it a battery, or do you
need to run wiring underneath?

180
00:06:49,410 --> 00:06:52,140
Such mundane problems,
power over Ethernet

181
00:06:52,140 --> 00:06:53,520
is a hot topic these days.

182
00:06:53,520 --> 00:06:54,582
Is that the solution?

183
00:06:54,582 --> 00:06:56,790
I'm sure there are 10 other
solutions we haven't even

184
00:06:56,790 --> 00:07:01,200
explored, and there's a tendency
to take the cliched solution,

185
00:07:01,200 --> 00:07:04,810
where something very innovative
and probably game-changing

186
00:07:04,810 --> 00:07:06,510
is just lurking
around the corner.

187
00:07:06,510 --> 00:07:08,385
So there's some very
interesting technologies

188
00:07:08,385 --> 00:07:10,650
just in these four topics
waiting to be researched,

189
00:07:10,650 --> 00:07:11,692
and then there's readers.

190
00:07:14,448 --> 00:07:18,270
We predicted some
10, 8 years ago

191
00:07:18,270 --> 00:07:20,700
that readers would eventually
be very inexpensive,

192
00:07:20,700 --> 00:07:21,540
hundreds of bucks.

193
00:07:21,540 --> 00:07:24,160
They're down to $1,000 now,
but there's a long way to go.

194
00:07:24,160 --> 00:07:26,493
And the fact of the matter
is the electronics in readers

195
00:07:26,493 --> 00:07:28,000
is not that complicated.

196
00:07:28,000 --> 00:07:30,068
So we have a lot of advancement.

197
00:07:30,068 --> 00:07:31,110
Then, there's middleware.

198
00:07:31,110 --> 00:07:33,040
There's software.

199
00:07:33,040 --> 00:07:34,290
There's the actual middleware.

200
00:07:34,290 --> 00:07:37,110
One of the problems with RFID
is that there's a tendency

201
00:07:37,110 --> 00:07:38,840
to believe that
RFID tags will be

202
00:07:38,840 --> 00:07:40,590
read in the same way
the barcodes will be.

203
00:07:40,590 --> 00:07:43,290
That is, you read the barcode,
you know exactly what's there,

204
00:07:43,290 --> 00:07:44,350
and you're done.

205
00:07:44,350 --> 00:07:46,575
The fact of the matter is
the whole point of RFID

206
00:07:46,575 --> 00:07:49,680
is reading all the time, and
you're going to miss a few.

207
00:07:49,680 --> 00:07:52,680
And missing reads results
in phantom inventory,

208
00:07:52,680 --> 00:07:54,750
especially when you
have a throughput model,

209
00:07:54,750 --> 00:07:57,060
where you look at
things going in and out.

210
00:07:57,060 --> 00:07:58,440
Right?

211
00:07:58,440 --> 00:08:01,260
It results in negative and
positive phantom inventory,

212
00:08:01,260 --> 00:08:04,590
and phantom inventory is
going to completely jam up

213
00:08:04,590 --> 00:08:07,230
any inventory control system,
any application that you

214
00:08:07,230 --> 00:08:08,220
develop.

215
00:08:08,220 --> 00:08:11,790
So for example, let's say, if
we all had tags, and god forbid,

216
00:08:11,790 --> 00:08:12,840
there was a fire.

217
00:08:12,840 --> 00:08:15,480
And we all went out, and there
were readers of the doors,

218
00:08:15,480 --> 00:08:18,840
and the readers at the
doors missed a few of us.

219
00:08:18,840 --> 00:08:20,820
This room would
legitimately claim

220
00:08:20,820 --> 00:08:23,910
that those of us that didn't
get read are still in this room,

221
00:08:23,910 --> 00:08:26,610
and it might send a fireman
into the room looking

222
00:08:26,610 --> 00:08:30,200
for those people that
the reader missed.

223
00:08:30,200 --> 00:08:31,700
Now, you can see right away--

224
00:08:31,700 --> 00:08:33,188
now, that's a dramatic example--

225
00:08:33,188 --> 00:08:34,730
but you can see
right away that, when

226
00:08:34,730 --> 00:08:38,720
you're doing inventory
control, how terribly that

227
00:08:38,720 --> 00:08:41,309
could send an inventory
control system out of control.

228
00:08:41,309 --> 00:08:41,809
Right?

229
00:08:41,809 --> 00:08:43,280
How do you manage
things like that?

230
00:08:43,280 --> 00:08:44,697
Well, you know
what, maybe there's

231
00:08:44,697 --> 00:08:47,105
a reader on the way out,
and that picked me up.

232
00:08:47,105 --> 00:08:49,730
So you can say, listen, although
this the reader didn't pick me

233
00:08:49,730 --> 00:08:51,440
up, a later reader saw me.

234
00:08:51,440 --> 00:08:52,790
So let's correct inventory here.

235
00:08:52,790 --> 00:08:54,710
That's a very deep
database concept.

236
00:08:54,710 --> 00:08:55,850
All right?

237
00:08:55,850 --> 00:08:58,700
It's a generalized
predicate of consistency

238
00:08:58,700 --> 00:09:00,980
that you need to superimpose
on top of a database

239
00:09:00,980 --> 00:09:02,990
to make sure that your
data is consistent.

240
00:09:02,990 --> 00:09:06,890
Right there is a research
topic just waiting for someone

241
00:09:06,890 --> 00:09:09,510
to look into.

242
00:09:09,510 --> 00:09:10,820
And then there's databases.

243
00:09:10,820 --> 00:09:13,650
There's streaming
data, fast and furious.

244
00:09:13,650 --> 00:09:16,530
How do you handle two
joins on streaming data?

245
00:09:16,530 --> 00:09:19,333
There's actually a lot of
research in various computer

246
00:09:19,333 --> 00:09:20,500
science departments on that.

247
00:09:20,500 --> 00:09:22,100
There's enterprise architecture.

248
00:09:22,100 --> 00:09:23,720
Is this going to
fundamentally change

249
00:09:23,720 --> 00:09:26,030
the way in which we do
enterprise architecture?

250
00:09:26,030 --> 00:09:29,840
The world is built
around databases today.

251
00:09:29,840 --> 00:09:33,650
Will we go to fundamentally a
new enterprise architecture?

252
00:09:33,650 --> 00:09:35,450
My own feeling is
about 10 years we'll

253
00:09:35,450 --> 00:09:39,200
be in a completely different
way of thinking about enterprise

254
00:09:39,200 --> 00:09:40,640
software systems
based completely

255
00:09:40,640 --> 00:09:43,040
on events, as opposed
to based on databases.

256
00:09:43,040 --> 00:09:45,620
That's my view,
but I may be wrong.

257
00:09:45,620 --> 00:09:48,410
And then there are distributed
systems, discovery.

258
00:09:48,410 --> 00:09:50,900
We developed something called
the object name system.

259
00:09:50,900 --> 00:09:52,442
If you're familiar
with it, it's just

260
00:09:52,442 --> 00:09:55,370
a way to find the company
that owns a particular EPC.

261
00:09:55,370 --> 00:09:57,200
That was a drop in the bucket.

262
00:09:57,200 --> 00:10:00,290
Already, we have
seen that is useful,

263
00:10:00,290 --> 00:10:02,600
but it's 1% of the
way to everything

264
00:10:02,600 --> 00:10:05,070
you need to do to really
take advantage of RFID.

265
00:10:05,070 --> 00:10:05,570
Right?

266
00:10:05,570 --> 00:10:07,320
I can say that,
because I was one

267
00:10:07,320 --> 00:10:08,570
of the people who wrote on it.

268
00:10:08,570 --> 00:10:10,170
So I can criticize it.

269
00:10:10,170 --> 00:10:12,770
But I think it's very important
for an academic community

270
00:10:12,770 --> 00:10:15,770
to point, focus a microscope,
and critique these things

271
00:10:15,770 --> 00:10:16,940
and start improving on them.

272
00:10:16,940 --> 00:10:18,273
And it's just waiting to happen.

273
00:10:18,273 --> 00:10:20,390
Is it going to be, I don't
know, the semantic web?

274
00:10:20,390 --> 00:10:21,080
What it going to be?

275
00:10:21,080 --> 00:10:21,580
Right?

276
00:10:21,580 --> 00:10:22,940
But you need to look at that.

277
00:10:22,940 --> 00:10:24,320
Identity management,
what's going

278
00:10:24,320 --> 00:10:26,330
to happen when you have hundreds
and hundreds of companies

279
00:10:26,330 --> 00:10:27,247
trying to access data?

280
00:10:27,247 --> 00:10:28,372
How do you manage identity?

281
00:10:28,372 --> 00:10:30,164
Is it going to be using
Microsoft Passport?

282
00:10:30,164 --> 00:10:31,790
Is it going to be
using something else?

283
00:10:31,790 --> 00:10:32,360
Right?

284
00:10:32,360 --> 00:10:33,568
What are the standards there?

285
00:10:33,568 --> 00:10:35,570
Is it going to be Federated?

286
00:10:35,570 --> 00:10:39,560
Fascinating, tens and
hundreds of research topics

287
00:10:39,560 --> 00:10:41,300
waiting here.

288
00:10:41,300 --> 00:10:44,000
The problem will
lie in making sure

289
00:10:44,000 --> 00:10:45,870
that these research
projects solve

290
00:10:45,870 --> 00:10:49,010
the question with a
clear understanding

291
00:10:49,010 --> 00:10:50,390
of the application.

292
00:10:50,390 --> 00:10:55,250
And that is why I think this
conference is perfectly timed,

293
00:10:55,250 --> 00:10:59,330
because the industry is finally
looking at this very seriously.

294
00:10:59,330 --> 00:11:02,508
Mike Rose of J&J, we were just
having breakfast this morning,

295
00:11:02,508 --> 00:11:04,550
and we were talking about
how the industry really

296
00:11:04,550 --> 00:11:06,860
is struggling with
some serious problems.

297
00:11:06,860 --> 00:11:09,300
And they have some experts,
and they're academic experts,

298
00:11:09,300 --> 00:11:11,510
and finally now, there can
be a real conversation.

299
00:11:11,510 --> 00:11:13,585
When we were doing
this a few years ago,

300
00:11:13,585 --> 00:11:15,710
we were making stuff up
and projecting our industry

301
00:11:15,710 --> 00:11:16,970
and hoping we were right.

302
00:11:16,970 --> 00:11:20,300
And by luck, dumb luck in many
ways, we were kind of right,

303
00:11:20,300 --> 00:11:22,640
and we had great guidance
from the early sponsors

304
00:11:22,640 --> 00:11:25,640
the Gillettes, the J&Js, the
Walmarts, et cetera, Targets.

305
00:11:25,640 --> 00:11:26,300
Right?

306
00:11:26,300 --> 00:11:30,030
But now is a time to have a very
high bandwidth conversation.

307
00:11:30,030 --> 00:11:32,300
And finally, by the
way, I think that RFID

308
00:11:32,300 --> 00:11:35,030
is going to change the way
our business processes work.

309
00:11:35,030 --> 00:11:37,460
It's going to change the way
that you go into a store,

310
00:11:37,460 --> 00:11:38,870
and you interact with the store.

311
00:11:38,870 --> 00:11:41,390
It's going to change the
way a store is operated.

312
00:11:41,390 --> 00:11:44,180
The machinery is going
to change, and that to me

313
00:11:44,180 --> 00:11:44,960
is a technology.

314
00:11:44,960 --> 00:11:46,382
It's not business.

315
00:11:46,382 --> 00:11:47,840
It's a technology,
and that's going

316
00:11:47,840 --> 00:11:51,720
to be the most fascinating
part of all this.

317
00:11:51,720 --> 00:11:53,362
You can look at applications.

318
00:11:53,362 --> 00:11:55,320
Most of the attention
has been on supply chain,

319
00:11:55,320 --> 00:11:57,840
but supply chain
is barely a start.

320
00:11:57,840 --> 00:12:01,860
We talk about retail, health
care, business to business,

321
00:12:01,860 --> 00:12:03,210
critical goods.

322
00:12:03,210 --> 00:12:06,900
We just went through Katrina
in the US and the tsunami

323
00:12:06,900 --> 00:12:10,350
and the earthquake in Pakistan.

324
00:12:10,350 --> 00:12:13,380
What happens when the
next disaster occurs?

325
00:12:13,380 --> 00:12:16,380
How do we make sure that
we route blood correctly?

326
00:12:16,380 --> 00:12:19,800
How do we make sure that if
there's a bird flu episode

327
00:12:19,800 --> 00:12:24,030
somewhere in some town
somewhere that we route flu

328
00:12:24,030 --> 00:12:26,250
vaccines or Tamiflu correctly?

329
00:12:26,250 --> 00:12:29,070
These are all, again, very
fundamental opportunities

330
00:12:29,070 --> 00:12:31,380
for RFID, and it
behooves us to start

331
00:12:31,380 --> 00:12:34,060
dealing with these things
in a scientific way.

332
00:12:34,060 --> 00:12:38,130
And so the very really
ancient approaches

333
00:12:38,130 --> 00:12:40,620
we're taking to disaster,
recovery with RFID,

334
00:12:40,620 --> 00:12:43,020
we should be able to of
run-- we should have a map.

335
00:12:43,020 --> 00:12:45,390
It's like a weather
map of the disaster

336
00:12:45,390 --> 00:12:49,070
and where your
various resources are.

337
00:12:49,070 --> 00:12:51,570
And how do you map them,
and what you need to wear?

338
00:12:51,570 --> 00:12:53,515
Well, here's where
the flooding is.

339
00:12:53,515 --> 00:12:55,140
We need to send
certain equipment here.

340
00:12:55,140 --> 00:12:56,460
Here's where there's
a disease outbreak.

341
00:12:56,460 --> 00:12:57,570
We need to send something here.

342
00:12:57,570 --> 00:13:00,028
We should be able to do this
in a much more scientific way,

343
00:13:00,028 --> 00:13:03,340
and visibility is key.

344
00:13:03,340 --> 00:13:10,123
Steve talked about the
airline, the aircraft industry,

345
00:13:10,123 --> 00:13:12,790
and where what's happening there
as the aircraft industry, which

346
00:13:12,790 --> 00:13:15,050
has historically been
very inefficient--

347
00:13:15,050 --> 00:13:17,590
I happen to have studied
them in a past life--

348
00:13:17,590 --> 00:13:20,170
is again waiting for a
fundamentally new, efficient

349
00:13:20,170 --> 00:13:20,950
model.

350
00:13:20,950 --> 00:13:24,220
Boeing and Airbus cannot survive
the way they have survived with

351
00:13:24,220 --> 00:13:27,410
gigantic government subsidies
at some level of the other.

352
00:13:27,410 --> 00:13:28,000
Right?

353
00:13:28,000 --> 00:13:29,600
And they have done
the right thing,

354
00:13:29,600 --> 00:13:31,600
and they're trying to
reengineered their supply

355
00:13:31,600 --> 00:13:32,260
chains.

356
00:13:32,260 --> 00:13:34,390
It's very different than
even the automotive supply

357
00:13:34,390 --> 00:13:36,640
chain, because their
lots are much smaller.

358
00:13:36,640 --> 00:13:38,740
And the products are
much fewer, and they

359
00:13:38,740 --> 00:13:40,890
can get a lot more flexibility.

360
00:13:40,890 --> 00:13:42,640
And I think that's a
great new opportunity

361
00:13:42,640 --> 00:13:45,670
to use RFID to get visibility
through just-in-time

362
00:13:45,670 --> 00:13:47,720
manufacturing,
delivery, et cetera.

363
00:13:47,720 --> 00:13:50,470
So again, for people who
do operations research--

364
00:13:50,470 --> 00:13:53,410
Tully, right-- fantastic
opportunity there.

365
00:13:53,410 --> 00:13:58,930
Now, if you look at just
operations, factory operations,

366
00:13:58,930 --> 00:14:02,260
Airbus and Boeing
are going to do it.

367
00:14:02,260 --> 00:14:05,650
DC operations, Brian
Subirana, who's right there,

368
00:14:05,650 --> 00:14:07,420
spent a lot of
time with Gillette

369
00:14:07,420 --> 00:14:10,570
looking at their DC operations.

370
00:14:10,570 --> 00:14:13,612
When we started that research,
we thought, is this boring?

371
00:14:13,612 --> 00:14:15,070
Are we going to
find something new?

372
00:14:15,070 --> 00:14:16,900
Actually get into
it, it's fascinating.

373
00:14:16,900 --> 00:14:18,760
You start peeling
the onion, and you

374
00:14:18,760 --> 00:14:21,610
find that there's so many
opportunities to make DCs.

375
00:14:21,610 --> 00:14:27,290
This boring, dusty box of a
building, exciting, faster,

376
00:14:27,290 --> 00:14:29,380
more efficient,
less error prone,

377
00:14:29,380 --> 00:14:32,346
and fundamentally, you can
start reducing inventories.

378
00:14:32,346 --> 00:14:34,200
Right?

379
00:14:34,200 --> 00:14:37,680
And of course, you can look
at institutions, hospitals,

380
00:14:37,680 --> 00:14:40,180
hospital operations,
huge opportunity.

381
00:14:40,180 --> 00:14:40,950
Right?

382
00:14:40,950 --> 00:14:43,590
You can do several
[? PCDCs ?] probably looking

383
00:14:43,590 --> 00:14:49,225
at the logistics of
hospitals and drug

384
00:14:49,225 --> 00:14:51,100
delivery, missed
prescriptions in hospitals--

385
00:14:51,100 --> 00:14:53,380
I'll talk about that in a
second-- personal systems.

386
00:14:53,380 --> 00:14:56,830
So I barely scratched the
surface of applications.

387
00:14:56,830 --> 00:14:59,750
And then finally, the million
question for an academic

388
00:14:59,750 --> 00:15:01,212
is so what's the analysis?

389
00:15:01,212 --> 00:15:02,170
Where are my creations?

390
00:15:02,170 --> 00:15:04,840
This is all fine, but
where my creations?

391
00:15:04,840 --> 00:15:07,210
Can I write in an IEEE
journal or an ACM paper

392
00:15:07,210 --> 00:15:12,730
on this and vie for the
next Turing or what?

393
00:15:12,730 --> 00:15:14,980
I actually fundamentally
think, yes, there

394
00:15:14,980 --> 00:15:17,470
are some fundamental
intellectual opportunities

395
00:15:17,470 --> 00:15:21,820
at every level, everything from
RF systems and communications.

396
00:15:21,820 --> 00:15:23,812
Many years ago, I was
sitting with Peter,

397
00:15:23,812 --> 00:15:26,020
and I said, Peter, what is
the fundamental difference

398
00:15:26,020 --> 00:15:28,315
between RFID and other
communication systems,

399
00:15:28,315 --> 00:15:29,440
and he summarized it to me.

400
00:15:29,440 --> 00:15:33,280
He said, in RFID, unlike in
other communication systems,

401
00:15:33,280 --> 00:15:35,440
the bandwidths and
the power are just

402
00:15:35,440 --> 00:15:37,420
very different in
two directions,

403
00:15:37,420 --> 00:15:39,670
and you want to design
protocols that work

404
00:15:39,670 --> 00:15:41,620
in these asymmetric channels.

405
00:15:41,620 --> 00:15:43,270
It's just very different.

406
00:15:43,270 --> 00:15:46,810
Computing cycles, I'll talk
about that, are expensive.

407
00:15:46,810 --> 00:15:49,690
Communication theory,
very interesting, right?

408
00:15:49,690 --> 00:15:52,300
All the communication theory,
error correcting codes,

409
00:15:52,300 --> 00:15:55,480
all the stuff doesn't
necessarily map to RFID,

410
00:15:55,480 --> 00:15:57,730
because RFID is so different.

411
00:15:57,730 --> 00:16:00,910
System dynamics, security,
I won't belabor it.

412
00:16:00,910 --> 00:16:05,080
Beautiful, challenging,
frustrating, exciting problems

413
00:16:05,080 --> 00:16:08,110
wait to be attacked,
and it's great to have

414
00:16:08,110 --> 00:16:10,570
this community come together
and try and think about them.

415
00:16:10,570 --> 00:16:12,550
The thing that makes RFID
special is everything

416
00:16:12,550 --> 00:16:13,490
is different.

417
00:16:13,490 --> 00:16:15,200
Everything about
RFID is different.

418
00:16:15,200 --> 00:16:15,700
Right?

419
00:16:15,700 --> 00:16:17,410
First of all, power is limited.

420
00:16:17,410 --> 00:16:19,000
So range is limited.

421
00:16:19,000 --> 00:16:19,780
Right?

422
00:16:19,780 --> 00:16:24,520
You can't compute, because every
compute cycle sucks up power.

423
00:16:24,520 --> 00:16:27,760
You can't put a lot of memory,
because every piece of memory

424
00:16:27,760 --> 00:16:29,410
increases cost and power.

425
00:16:29,410 --> 00:16:30,650
Cost is an issue.

426
00:16:30,650 --> 00:16:32,170
So you're kind of boxed in.

427
00:16:32,170 --> 00:16:36,910
So it's like you have
a little wiggle room,

428
00:16:36,910 --> 00:16:38,590
but yet you're trying
to do everything.

429
00:16:38,590 --> 00:16:41,260
Now, you add to that
bandwidth is limited,

430
00:16:41,260 --> 00:16:42,940
because you're
always operating in

431
00:16:42,940 --> 00:16:46,390
industrial scientific medical
bands or some free band.

432
00:16:46,390 --> 00:16:48,363
At least the general
RFID movement's

433
00:16:48,363 --> 00:16:49,780
always going to
be in a free band,

434
00:16:49,780 --> 00:16:50,988
so bandwidth is very limited.

435
00:16:50,988 --> 00:16:52,900
You can't do much
with this stuff.

436
00:16:52,900 --> 00:16:54,850
Memory is at a premium,
as we discovered

437
00:16:54,850 --> 00:16:56,860
at the Auto-ID center.

438
00:16:56,860 --> 00:17:00,220
You get a lot of data from RFID
tags, but it's very fallible.

439
00:17:00,220 --> 00:17:02,470
In other words, it can
be a lot of wrong data.

440
00:17:02,470 --> 00:17:05,349
Like as I said, it's
not wrong in the sense

441
00:17:05,349 --> 00:17:08,300
that the reader is wrong,
but the false negatives

442
00:17:08,300 --> 00:17:10,750
that is a misread
can manifest itself

443
00:17:10,750 --> 00:17:15,250
in ways that can screw up
your typical algorithm.

444
00:17:15,250 --> 00:17:15,880
Right?

445
00:17:15,880 --> 00:17:18,369
So you've got to
correct for that.

446
00:17:18,369 --> 00:17:19,720
Here's a very interesting one.

447
00:17:19,720 --> 00:17:21,530
Tag connectivity is sporadic.

448
00:17:21,530 --> 00:17:23,260
That's a very subtle one.

449
00:17:23,260 --> 00:17:27,069
Even my cell phone, let us say
that Cingular Wireless decides

450
00:17:27,069 --> 00:17:29,710
that it wants to upgrade the
firmware on my cell phone.

451
00:17:29,710 --> 00:17:31,420
It can kind of do,
it because it's got--

452
00:17:31,420 --> 00:17:33,070
80% of the time it's on.

453
00:17:33,070 --> 00:17:35,943
With an RFID Tag,
you don't know where

454
00:17:35,943 --> 00:17:37,110
you're going to see it next.

455
00:17:37,110 --> 00:17:39,552
Let's say you wanted to
write some data to it.

456
00:17:39,552 --> 00:17:41,510
You don't know where it's
going to see it next.

457
00:17:41,510 --> 00:17:42,650
It could be in your company.

458
00:17:42,650 --> 00:17:43,317
It could be MIT.

459
00:17:43,317 --> 00:17:45,878
It could be at P&G. It
could be at Walmart.

460
00:17:45,878 --> 00:17:47,670
You don't know when
you're going to see it.

461
00:17:47,670 --> 00:17:49,250
So what are you going
to do, preposition

462
00:17:49,250 --> 00:17:51,380
all that information in
every reader on the planet

463
00:17:51,380 --> 00:17:53,600
that you want to
put on that tag?

464
00:17:53,600 --> 00:17:55,710
Well, you couldn't
possibly do that either.

465
00:17:55,710 --> 00:17:58,070
So there's some
compromise, where

466
00:17:58,070 --> 00:18:00,977
you kind of know where
stuff is, and you think

467
00:18:00,977 --> 00:18:02,060
you're going to sit there.

468
00:18:02,060 --> 00:18:03,620
And you push the information
there in the hope

469
00:18:03,620 --> 00:18:06,410
that, when it shows up, you
can push that information down.

470
00:18:06,410 --> 00:18:09,740
That's just a fundamental
distributed computing challenge

471
00:18:09,740 --> 00:18:16,300
of extremely irritating but
yet interesting magnitude.

472
00:18:16,300 --> 00:18:16,840
Right?

473
00:18:16,840 --> 00:18:22,270
So tag connectivity being
sporadic compounds everything.

474
00:18:22,270 --> 00:18:24,820
And to add to
everything, cell phones

475
00:18:24,820 --> 00:18:27,253
are targeting a human
being carrying this thing

476
00:18:27,253 --> 00:18:28,420
and holding it to their ear.

477
00:18:28,420 --> 00:18:29,410
That's it.

478
00:18:29,410 --> 00:18:31,330
Tags can show up on products.

479
00:18:31,330 --> 00:18:33,070
They can show up on animals.

480
00:18:33,070 --> 00:18:33,940
Right?

481
00:18:33,940 --> 00:18:36,280
They can show up on pallets.

482
00:18:36,280 --> 00:18:38,330
They can show up on
metal containers.

483
00:18:38,330 --> 00:18:40,223
They can show up inside
metal containers.

484
00:18:40,223 --> 00:18:40,910
Right?

485
00:18:40,910 --> 00:18:42,160
They can have sensors on them?

486
00:18:42,160 --> 00:18:43,035
They could be active.

487
00:18:43,035 --> 00:18:44,490
They could be passive.

488
00:18:44,490 --> 00:18:46,540
So you've got to
prepare for everything.

489
00:18:46,540 --> 00:18:48,660
So if you have a
reader here, it's

490
00:18:48,660 --> 00:18:51,720
got to be ready to read that
passive tag sitting right there

491
00:18:51,720 --> 00:18:55,810
and possibly an active tag
with a one kilometer radius

492
00:18:55,810 --> 00:18:57,780
which happens to be
in the next building

493
00:18:57,780 --> 00:19:00,780
and make sense out of it.

494
00:19:00,780 --> 00:19:03,250
It's fascinating
to think about it,

495
00:19:03,250 --> 00:19:05,520
and these are problems that
people are facing today.

496
00:19:05,520 --> 00:19:07,103
These are not problems
I just made up.

497
00:19:07,103 --> 00:19:10,020
This is just beginning
to happen in real life.

498
00:19:10,020 --> 00:19:11,370
Right?

499
00:19:11,370 --> 00:19:15,780
And finally, the range
of technologies is huge.

500
00:19:15,780 --> 00:19:17,700
You will have active
tags, passive tags.

501
00:19:17,700 --> 00:19:18,940
You're going to have GPS.

502
00:19:18,940 --> 00:19:19,440
Right?

503
00:19:19,440 --> 00:19:22,340
How do you in fact integrate
GPS with these systems?

504
00:19:22,340 --> 00:19:24,090
Actually, by the way,
that's not bad news.

505
00:19:24,090 --> 00:19:25,770
It's all good news, and
there's all these people that

506
00:19:25,770 --> 00:19:27,630
are doing these
things, but there

507
00:19:27,630 --> 00:19:30,540
is no cohesive overarching
thinking pulling it

508
00:19:30,540 --> 00:19:32,070
all together.

509
00:19:32,070 --> 00:19:36,460
You have cellular technology,
readers inside cell phones,

510
00:19:36,460 --> 00:19:38,580
fascinating.

511
00:19:38,580 --> 00:19:42,120
So what I thought it would
do is I'll take this map.

512
00:19:42,120 --> 00:19:44,370
It's kind of like
looking at a planet,

513
00:19:44,370 --> 00:19:46,628
and I don't have very much time.

514
00:19:46,628 --> 00:19:48,420
But I'd like to burrow
down in a few places

515
00:19:48,420 --> 00:19:50,820
and just give you a taste
of some of the research

516
00:19:50,820 --> 00:19:53,100
that the Auto-ID center
did, other people did.

517
00:19:53,100 --> 00:19:56,760
And I want to show, if you
haven't seen this, give you

518
00:19:56,760 --> 00:19:59,640
a sense of how rich it can get,
how quickly it can get rich,

519
00:19:59,640 --> 00:20:00,900
when you burrow down.

520
00:20:00,900 --> 00:20:03,837
So I picked, when I was
making these slides,

521
00:20:03,837 --> 00:20:06,420
I created this landscape, and I
said, let's look at protocols.

522
00:20:06,420 --> 00:20:07,740
Let's look at RFID protocols.

523
00:20:07,740 --> 00:20:09,220
How rich can that really be?

524
00:20:09,220 --> 00:20:09,720
Right?

525
00:20:09,720 --> 00:20:12,420
It was a question I asked
myself, in 1998, when

526
00:20:12,420 --> 00:20:15,100
I was deciding with David
Brock to get into this research

527
00:20:15,100 --> 00:20:15,600
topic.

528
00:20:15,600 --> 00:20:17,830
Could I write a paper here?

529
00:20:17,830 --> 00:20:18,330
Right?

530
00:20:18,330 --> 00:20:20,205
What I can tell you now
is there are probably

531
00:20:20,205 --> 00:20:21,210
a million papers there.

532
00:20:21,210 --> 00:20:23,220
At that time, I
thought there was zero.

533
00:20:23,220 --> 00:20:23,838
Right?

534
00:20:23,838 --> 00:20:25,630
So when you get into
it-- so let's zoom in.

535
00:20:25,630 --> 00:20:26,640
What's protocols?

536
00:20:26,640 --> 00:20:30,000
Well, it turns out, protocols
are not just protocols.

537
00:20:30,000 --> 00:20:32,520
There's a physical layer,
and there's a logical layer.

538
00:20:32,520 --> 00:20:34,770
In the physical layer, there
are different frequencies

539
00:20:34,770 --> 00:20:36,030
to consider.

540
00:20:36,030 --> 00:20:38,280
There are different schemes,
like ultrawide band, that

541
00:20:38,280 --> 00:20:40,350
are popping up to consider.

542
00:20:40,350 --> 00:20:40,850
OK.

543
00:20:40,850 --> 00:20:41,892
Well, that's pretty good.

544
00:20:41,892 --> 00:20:43,700
Right?

545
00:20:43,700 --> 00:20:45,290
And then there's
a logical layer.

546
00:20:45,290 --> 00:20:47,450
In the logical layer, you need
to look at readers interfering

547
00:20:47,450 --> 00:20:48,117
with each other.

548
00:20:48,117 --> 00:20:51,140
Dan Engels and I wrote
a paper, in '99 I think,

549
00:20:51,140 --> 00:20:52,723
on readers interfering,
and we thought

550
00:20:52,723 --> 00:20:55,182
it was just pie in the sky,
just an intellectual problem we

551
00:20:55,182 --> 00:20:55,980
invented.

552
00:20:55,980 --> 00:20:57,980
Turns out, it's actually
a very serious problem.

553
00:20:57,980 --> 00:20:59,180
Right?

554
00:20:59,180 --> 00:21:01,940
So you need to make sure
that readers don't interfere,

555
00:21:01,940 --> 00:21:04,140
and there are a couple
of ways to do that.

556
00:21:04,140 --> 00:21:06,860
You can do something called
tag sessions which actually

557
00:21:06,860 --> 00:21:08,750
is in the Gen 2 protocol.

558
00:21:08,750 --> 00:21:11,340
And you can do something called
in-band reader coordination,

559
00:21:11,340 --> 00:21:14,070
which we couldn't get
into the Gen 2 protocol,

560
00:21:14,070 --> 00:21:16,110
but there are other
reasons for that, as well.

561
00:21:16,110 --> 00:21:17,777
There are technical
challenges that need

562
00:21:17,777 --> 00:21:19,380
to be addressed to do that.

563
00:21:19,380 --> 00:21:21,030
There's security.

564
00:21:21,030 --> 00:21:21,530
Right?

565
00:21:21,530 --> 00:21:26,682
How do you make sure that
privacy is protected,

566
00:21:26,682 --> 00:21:29,690
someone can't
eavesdrop on targets,

567
00:21:29,690 --> 00:21:33,740
reads and see what inventory
they've purchased so they can

568
00:21:33,740 --> 00:21:34,310
game them.

569
00:21:34,310 --> 00:21:35,440
Right?

570
00:21:35,440 --> 00:21:37,140
There's different data types.

571
00:21:37,140 --> 00:21:37,970
There's the EPC.

572
00:21:37,970 --> 00:21:39,740
There might be sensor data.

573
00:21:39,740 --> 00:21:42,350
Then, the sensors, how do you
make sure, when you need a tag,

574
00:21:42,350 --> 00:21:43,940
let's say if it has
a sensor, that you

575
00:21:43,940 --> 00:21:45,620
know it was a
temperature sensor,

576
00:21:45,620 --> 00:21:49,070
that the data follows the EPC?

577
00:21:49,070 --> 00:21:51,950
Well, the temperature that are
reported, is it in Celsius,

578
00:21:51,950 --> 00:21:54,110
or is it in Fahrenheit?

579
00:21:54,110 --> 00:21:56,090
Is it a time series, or
is it the temperature

580
00:21:56,090 --> 00:21:58,450
taken at that moment?

581
00:21:58,450 --> 00:22:01,870
If it's a time series, what
units are we using for time?

582
00:22:01,870 --> 00:22:02,840
Is it compressed?

583
00:22:02,840 --> 00:22:05,330
Then, what's the
compression algorithm?

584
00:22:05,330 --> 00:22:06,290
Right?

585
00:22:06,290 --> 00:22:07,220
Isn't that incredible?

586
00:22:07,220 --> 00:22:09,578
Just that question
you can go on and on.

587
00:22:09,578 --> 00:22:11,120
Well, if you're
going to compress it,

588
00:22:11,120 --> 00:22:12,828
what's the cheapest
compression algorithm

589
00:22:12,828 --> 00:22:14,420
you can put on a
low cost chip which

590
00:22:14,420 --> 00:22:17,900
has very little power or little
bandwidth and a little memory

591
00:22:17,900 --> 00:22:21,090
and little computing power?

592
00:22:21,090 --> 00:22:22,022
So OK.

593
00:22:22,022 --> 00:22:23,480
So there are a lot
of things there.

594
00:22:23,480 --> 00:22:24,580
So let's focus in a little more.

595
00:22:24,580 --> 00:22:26,570
So suddenly in our protocols
is a very rich topic.

596
00:22:26,570 --> 00:22:26,990
All right.

597
00:22:26,990 --> 00:22:28,198
Let's focus in a little more.

598
00:22:28,198 --> 00:22:31,460
Let's look at just
security, just security.

599
00:22:31,460 --> 00:22:34,310
So this happened to
me in 2001, 2002.

600
00:22:34,310 --> 00:22:37,740
Was it 2002 when I gave a
talk in Ron Rivest's group?

601
00:22:37,740 --> 00:22:38,690
He's a professor here.

602
00:22:38,690 --> 00:22:40,730
He is one of the
security gurus at MIT,

603
00:22:40,730 --> 00:22:43,370
and I went and gave a
talk there about RFID

604
00:22:43,370 --> 00:22:45,870
and some early
thinking on security.

605
00:22:45,870 --> 00:22:47,040
So we dived into it.

606
00:22:47,040 --> 00:22:47,550
All right.

607
00:22:47,550 --> 00:22:49,520
So let's look at
security a little deeper.

608
00:22:49,520 --> 00:22:53,430
Well, that's not good enough,
but if you get into security

609
00:22:53,430 --> 00:22:55,560
there are layers
upon layers there.

610
00:22:55,560 --> 00:22:57,720
The reader to tag
security making sure

611
00:22:57,720 --> 00:23:00,480
that the reader and tag
communication is made secure,

612
00:23:00,480 --> 00:23:02,370
and we'll expand upon that.

613
00:23:02,370 --> 00:23:04,475
There's making sure the
middleware is secure.

614
00:23:04,475 --> 00:23:05,850
That is that the
middleware talks

615
00:23:05,850 --> 00:23:08,040
to the reader in a way
that is secure, especially

616
00:23:08,040 --> 00:23:09,450
if it's wireless.

617
00:23:09,450 --> 00:23:13,920
There's security between the
middleware and applications--

618
00:23:13,920 --> 00:23:16,140
base AP, bound, whatever it is.

619
00:23:16,140 --> 00:23:18,420
There's inter-corporation
security.

620
00:23:18,420 --> 00:23:19,920
Right?

621
00:23:19,920 --> 00:23:21,950
Let us say that you have
a fridge at home that's

622
00:23:21,950 --> 00:23:24,840
reading tags inside
the fridge, and someone

623
00:23:24,840 --> 00:23:27,510
is sitting outside
looking at your ONS hits.

624
00:23:27,510 --> 00:23:32,590
They can figure out,
Shriram, he likes margarine.

625
00:23:32,590 --> 00:23:33,090
Right?

626
00:23:33,090 --> 00:23:34,440
Maybe you're worried about that.

627
00:23:34,440 --> 00:23:34,950
Right?

628
00:23:34,950 --> 00:23:36,280
So that is a very rich problem.

629
00:23:36,280 --> 00:23:36,780
OK.

630
00:23:36,780 --> 00:23:37,822
That's getting too broad.

631
00:23:37,822 --> 00:23:40,770
So now, we're two
levels down, folks.

632
00:23:40,770 --> 00:23:42,538
All right?

633
00:23:42,538 --> 00:23:43,330
So that's too much.

634
00:23:43,330 --> 00:23:45,690
Let's look at
reader tag security.

635
00:23:45,690 --> 00:23:47,850
Guess what, that gets
very interesting too.

636
00:23:47,850 --> 00:23:49,230
Now, we're three levels down.

637
00:23:49,230 --> 00:23:49,730
Right?

638
00:23:49,730 --> 00:23:54,600
I had thought there wasn't
a paper there in 2000, 2001.

639
00:23:54,600 --> 00:23:57,480
Well, there's
authenticating the tag.

640
00:23:57,480 --> 00:23:58,590
What's that?

641
00:23:58,590 --> 00:24:00,240
Authenticating the
tag is making sure

642
00:24:00,240 --> 00:24:02,430
that the tag you read
as not a spoof tag.

643
00:24:02,430 --> 00:24:06,420
It is that very tag that
J&J put on a J&J product,

644
00:24:06,420 --> 00:24:08,640
so that you can
prevent counterfeits.

645
00:24:08,640 --> 00:24:10,590
And that has fundamental
business benefits,

646
00:24:10,590 --> 00:24:13,200
health benefits,
social benefits.

647
00:24:13,200 --> 00:24:14,780
There's authenticating
the reader,

648
00:24:14,780 --> 00:24:16,530
which is a tag doesn't
respond to a reader

649
00:24:16,530 --> 00:24:18,000
if it's not the right reader.

650
00:24:18,000 --> 00:24:19,640
That's privacy.

651
00:24:19,640 --> 00:24:20,470
Right?

652
00:24:20,470 --> 00:24:23,800
The cheaper tags, passive tags,
the whole privacy discussion

653
00:24:23,800 --> 00:24:26,140
arose because the
tag is promiscuous.

654
00:24:26,140 --> 00:24:27,400
Any reader can read it.

655
00:24:27,400 --> 00:24:29,650
So if I have a bottle of
some medication in my pocket,

656
00:24:29,650 --> 00:24:30,825
Trevor come and read it.

657
00:24:30,825 --> 00:24:32,200
And I certainly
don't want Trevor

658
00:24:32,200 --> 00:24:34,103
to know what medication
I have in my pocket.

659
00:24:34,103 --> 00:24:34,625
Right?

660
00:24:34,625 --> 00:24:37,000
But if the tags say, I will
only respond to a reader that

661
00:24:37,000 --> 00:24:38,860
belongs to anyone but Trevor--

662
00:24:38,860 --> 00:24:42,490
I'm sorry, Trevor-- then
I am protecting my privacy

663
00:24:42,490 --> 00:24:43,210
from Trevor.

664
00:24:43,210 --> 00:24:43,710
Right?

665
00:24:43,710 --> 00:24:45,268
So there's authenticating
the reader.

666
00:24:45,268 --> 00:24:46,810
Then, I need to
prevent eavesdropping

667
00:24:46,810 --> 00:24:47,410
which is all right.

668
00:24:47,410 --> 00:24:49,190
The reason a tag will
authenticate each other,

669
00:24:49,190 --> 00:24:50,857
I want to make sure
a third parity can't

670
00:24:50,857 --> 00:24:52,930
listen in and figure
out what inventories

671
00:24:52,930 --> 00:24:54,910
in Target's warehouse.

672
00:24:54,910 --> 00:24:57,040
Oh, Target's running a
back to school promotion.

673
00:24:57,040 --> 00:24:58,780
I can see that.

674
00:24:58,780 --> 00:25:00,340
I'm at Staples.

675
00:25:00,340 --> 00:25:02,880
I'm going to compete with them.

676
00:25:02,880 --> 00:25:03,380
Right?

677
00:25:03,380 --> 00:25:05,653
They're going a week
early this year.

678
00:25:05,653 --> 00:25:07,070
So I'm going to
go a week earlier.

679
00:25:07,070 --> 00:25:08,840
That's actually a big
deal for these guys.

680
00:25:08,840 --> 00:25:09,710
Right?

681
00:25:09,710 --> 00:25:11,960
There's preventing tracking,
which is making sure

682
00:25:11,960 --> 00:25:17,090
that I don't know by put readers
at every T-station in Boston,

683
00:25:17,090 --> 00:25:24,440
where an ex's girlfriend
or boyfriend lives.

684
00:25:24,440 --> 00:25:25,108
Right?

685
00:25:25,108 --> 00:25:26,900
And then how do you
personalize everything?

686
00:25:26,900 --> 00:25:28,760
When I buy socks
and take them home,

687
00:25:28,760 --> 00:25:30,890
how do I make sure
that only my reader can

688
00:25:30,890 --> 00:25:32,630
read my socks and my medicines?

689
00:25:32,630 --> 00:25:34,340
And if someone else
tries to read them,

690
00:25:34,340 --> 00:25:35,360
these are personalized.

691
00:25:35,360 --> 00:25:38,390
Even the maker of the product
can't read them anymore.

692
00:25:38,390 --> 00:25:39,320
That's getting rich.

693
00:25:39,320 --> 00:25:39,950
Right?

694
00:25:39,950 --> 00:25:41,990
So now, we're three levels
down, and again, it's

695
00:25:41,990 --> 00:25:44,120
turned into a really
complex problem.

696
00:25:44,120 --> 00:25:45,440
Right?

697
00:25:45,440 --> 00:25:49,670
And just for the flavor
of it, let's brainstorm.

698
00:25:49,670 --> 00:25:51,350
It actually gets
very complicated.

699
00:25:51,350 --> 00:25:54,990
My initial guess, I had a
student called Joe Foley.

700
00:25:54,990 --> 00:25:57,320
We said, let's just encrypt
the number on the tag.

701
00:25:57,320 --> 00:25:59,420
Privacy will be taken care of.

702
00:25:59,420 --> 00:26:00,442
No.

703
00:26:00,442 --> 00:26:02,150
If you just encrypt
the number on the tag

704
00:26:02,150 --> 00:26:04,610
and put it on the
tag, yeah, I can

705
00:26:04,610 --> 00:26:08,840
tell that you have Gucci shoes
on, but I know the number.

706
00:26:08,840 --> 00:26:09,470
Are they Gucci?

707
00:26:09,470 --> 00:26:09,640
Really?

708
00:26:09,640 --> 00:26:10,220
No.

709
00:26:10,220 --> 00:26:12,505
I know the number--

710
00:26:12,505 --> 00:26:13,410
no.

711
00:26:13,410 --> 00:26:15,410
I know the number, and
I can still track you.

712
00:26:15,410 --> 00:26:17,810
So it doesn't actually
solve the privacy problem.

713
00:26:17,810 --> 00:26:19,350
It solved the
privacy of the shoes.

714
00:26:19,350 --> 00:26:21,500
It didn't solve Bernie's
privacy problem.

715
00:26:21,500 --> 00:26:22,070
Right?

716
00:26:22,070 --> 00:26:24,080
Not that easy, right?

717
00:26:24,080 --> 00:26:25,800
So that doesn't
solve the problem.

718
00:26:25,800 --> 00:26:26,300
OK.

719
00:26:26,300 --> 00:26:27,740
Then, we thought, ah, we got it.

720
00:26:27,740 --> 00:26:29,490
Let's just kill
the serial number.

721
00:26:29,490 --> 00:26:29,990
All right.

722
00:26:29,990 --> 00:26:32,383
So I don't know what that
particular pair of Gucci shoes.

723
00:26:32,383 --> 00:26:33,800
I know it's a pair
of Gucci shoes.

724
00:26:33,800 --> 00:26:34,880
Right?

725
00:26:34,880 --> 00:26:38,600
And a lot of people wear
Gucci shoes, so it's OK.

726
00:26:38,600 --> 00:26:42,650
But you know what, Bernie also
has a Rolex pen in his pocket,

727
00:26:42,650 --> 00:26:45,710
I believe, and he may
be the only person

728
00:26:45,710 --> 00:26:48,380
in Boston with high
probability who has both Gucci

729
00:26:48,380 --> 00:26:49,850
shoes and a Rolex pen.

730
00:26:49,850 --> 00:26:53,400
So by constellation tracking, I
know him, and I can track him.

731
00:26:53,400 --> 00:26:56,482
And maybe you add to that some
particular brand of jacket,

732
00:26:56,482 --> 00:26:57,690
that's it, I've triangulated.

733
00:26:57,690 --> 00:26:59,370
So again, I don't
solve the problem.

734
00:26:59,370 --> 00:26:59,870
Right?

735
00:26:59,870 --> 00:27:02,310
I'm just trying to
give you a sense of it.

736
00:27:02,310 --> 00:27:05,100
Tough problem, folks,
we thought we had it.

737
00:27:05,100 --> 00:27:05,610
Wrong.

738
00:27:05,610 --> 00:27:06,168
Right?

739
00:27:06,168 --> 00:27:07,210
You go deeper and deeper.

740
00:27:07,210 --> 00:27:08,070
Right?

741
00:27:08,070 --> 00:27:10,812
And then you insist that
tag needs to the reader.

742
00:27:10,812 --> 00:27:12,270
Well, that means
that the tag needs

743
00:27:12,270 --> 00:27:14,860
to have the password
of every reader?

744
00:27:14,860 --> 00:27:16,400
Well, that's not a cheap tag.

745
00:27:16,400 --> 00:27:16,900
Right?

746
00:27:19,510 --> 00:27:21,340
I won't belabor
everything, but it just

747
00:27:21,340 --> 00:27:23,320
turns into such an
interesting problem.

748
00:27:23,320 --> 00:27:24,760
Preventing counterfeits, right?

749
00:27:24,760 --> 00:27:26,590
This traditional
encryption is expensive.

750
00:27:26,590 --> 00:27:28,503
There's challenge response.

751
00:27:28,503 --> 00:27:29,920
That means you
need to put secrets

752
00:27:29,920 --> 00:27:34,177
on every tag, very rich, very
rich, just waiting for someone

753
00:27:34,177 --> 00:27:35,260
to go nail these problems.

754
00:27:35,260 --> 00:27:37,425
Now, these we've
looked at, and if you

755
00:27:37,425 --> 00:27:39,550
want to prevent eavesdropping,
many of these things

756
00:27:39,550 --> 00:27:42,160
will kind of solve,
but here's the secret.

757
00:27:42,160 --> 00:27:43,330
Right?

758
00:27:43,330 --> 00:27:47,920
Actually, we've solved some of
these problems individually.

759
00:27:47,920 --> 00:27:51,477
We have not solved the
bigger problem of addressing

760
00:27:51,477 --> 00:27:52,310
all these questions.

761
00:27:52,310 --> 00:27:55,090
We just haven't, and
right there, it's

762
00:27:55,090 --> 00:27:58,330
just a huge research topic
waiting to be addressed.

763
00:27:58,330 --> 00:28:01,690
And for what it's
worth, Steve Weis

764
00:28:01,690 --> 00:28:03,580
wrote a beautiful
master's thesis

765
00:28:03,580 --> 00:28:06,220
on this, in which he
provides some solutions.

766
00:28:06,220 --> 00:28:09,220
And he very explicitly lists
other things we couldn't solve,

767
00:28:09,220 --> 00:28:10,600
but that's a start.

768
00:28:10,600 --> 00:28:11,152
Right?

769
00:28:11,152 --> 00:28:13,360
So that's a master's thesis
that came out of one talk

770
00:28:13,360 --> 00:28:17,350
that I was pretty much saying
there's not much research on.

771
00:28:17,350 --> 00:28:17,850
Right?

772
00:28:17,850 --> 00:28:19,740
And all the master's
thesis, great thesis

773
00:28:19,740 --> 00:28:22,260
did was solve a
couple of subparts,

774
00:28:22,260 --> 00:28:26,310
and list about 10 other
problems that we didn't solve.

775
00:28:26,310 --> 00:28:26,810
OK?

776
00:28:26,810 --> 00:28:28,060
So let's zoom back up.

777
00:28:28,060 --> 00:28:34,438
That was protocols, 1 of
30 things I've listed here,

778
00:28:34,438 --> 00:28:35,730
and you can look at everything.

779
00:28:35,730 --> 00:28:38,150
Retail, what are the
opportunities in retail.

780
00:28:38,150 --> 00:28:39,560
What can I zoom in?

781
00:28:39,560 --> 00:28:40,670
Right?

782
00:28:40,670 --> 00:28:42,037
Retail, what is retail really?

783
00:28:42,037 --> 00:28:43,370
What is the retail supply chain?

784
00:28:43,370 --> 00:28:45,590
Retail supply chain is a
very interesting thing.

785
00:28:45,590 --> 00:28:48,710
Toothbrushes are
made in units of one.

786
00:28:48,710 --> 00:28:50,560
You make one
toothbrush at a time.

787
00:28:50,560 --> 00:28:51,440
OK?

788
00:28:51,440 --> 00:28:53,150
What happens in the
retail supply chain

789
00:28:53,150 --> 00:28:56,370
is you make one
toothbrush at a time.

790
00:28:56,370 --> 00:28:58,570
You assemble
toothbrushes into cases,

791
00:28:58,570 --> 00:29:00,450
so you have say six
toothbrushes in a case.

792
00:29:00,450 --> 00:29:02,910
You put cases together,
you make a pallet.

793
00:29:02,910 --> 00:29:04,950
You take pallets together,
you make a container.

794
00:29:04,950 --> 00:29:09,600
You assemble them up, so that
you can ship them together.

795
00:29:09,600 --> 00:29:10,320
All right?

796
00:29:10,320 --> 00:29:12,360
So this is what the
supply chain is.

797
00:29:12,360 --> 00:29:14,130
Then, you move
them, you sort them,

798
00:29:14,130 --> 00:29:17,480
and you repeat until the end.

799
00:29:17,480 --> 00:29:23,230
Aggregate, sort, transfer,
aggregate, de-aggregate, sort,

800
00:29:23,230 --> 00:29:25,530
transfer, and that's
basically a supply chain.

801
00:29:25,530 --> 00:29:28,390
It's a big mixing thing,
where you make things as one,

802
00:29:28,390 --> 00:29:31,030
but all of the same kind,
and that's a manufacturer.

803
00:29:31,030 --> 00:29:34,450
And people walk out with ones
but all of a different kind,

804
00:29:34,450 --> 00:29:35,925
and that's the consumer.

805
00:29:35,925 --> 00:29:37,300
And trying to do
that efficiently

806
00:29:37,300 --> 00:29:40,570
is the supply chain in the
retail world as we know it.

807
00:29:40,570 --> 00:29:41,440
Right?

808
00:29:41,440 --> 00:29:44,080
This is another
fascinating research area,

809
00:29:44,080 --> 00:29:46,450
and this is the one where
the industry has focused

810
00:29:46,450 --> 00:29:48,400
its attention most immediately.

811
00:29:48,400 --> 00:29:51,100
Because you're trying to do
this in a timely fashion,

812
00:29:51,100 --> 00:29:53,770
with very few errors, and
very little shrinkage.

813
00:29:53,770 --> 00:29:56,080
Well, if you actually
look at it in real life,

814
00:29:56,080 --> 00:30:00,970
it turns out that that process
is riddled with errors,

815
00:30:00,970 --> 00:30:06,370
too much, and that's excess,
inefficiencies, losses,

816
00:30:06,370 --> 00:30:11,860
dilution, counterfeit,
brand dilution.

817
00:30:11,860 --> 00:30:13,720
It's just incredible.

818
00:30:13,720 --> 00:30:16,990
And you can look at the
actual numbers that come out.

819
00:30:16,990 --> 00:30:19,960
We carry something like
30%, 40% of inventory

820
00:30:19,960 --> 00:30:21,160
we shouldn't be carrying.

821
00:30:21,160 --> 00:30:23,620
For all that, we are 10%
out of stock as an industry.

822
00:30:23,620 --> 00:30:24,560
Right?

823
00:30:24,560 --> 00:30:27,340
And there's some fundamental
research topics here waiting

824
00:30:27,340 --> 00:30:28,330
to be addressed.

825
00:30:28,330 --> 00:30:31,490
And whether they end up
being looked at by companies,

826
00:30:31,490 --> 00:30:34,390
whether they end up being
looked at by academics,

827
00:30:34,390 --> 00:30:36,640
the fact of the
matter is that these

828
00:30:36,640 --> 00:30:39,970
are rich research topics waiting
for academics to look at.

829
00:30:39,970 --> 00:30:41,620
Now, I'll say this.

830
00:30:41,620 --> 00:30:44,590
You can actually map the
space out in many ways.

831
00:30:44,590 --> 00:30:47,380
You can take, for example,
the supply chain--

832
00:30:47,380 --> 00:30:50,050
DC to back room, back room,
back room the sales floor,

833
00:30:50,050 --> 00:30:51,863
sales floor, sales
for the storage.

834
00:30:51,863 --> 00:30:53,530
And you can look at
different parameters

835
00:30:53,530 --> 00:30:56,240
you can get time,
quantities, configuration.

836
00:30:56,240 --> 00:30:58,618
And if you map it out that way--

837
00:30:58,618 --> 00:31:00,660
in the interest of speed,
I'll just write it out.

838
00:31:00,660 --> 00:31:02,160
I don't expect
everyone to see this.

839
00:31:02,160 --> 00:31:06,300
But what I'm saying is
right there in that matrix,

840
00:31:06,300 --> 00:31:11,580
I've captured probably
about 70% of the problems

841
00:31:11,580 --> 00:31:14,010
that most retailers
are looking at.

842
00:31:14,010 --> 00:31:17,520
They're looking at is
my fresh food arriving

843
00:31:17,520 --> 00:31:22,260
in a timely fashion at
the fresh food counter?

844
00:31:22,260 --> 00:31:26,460
So that when my consumer buys
that lettuce, it is fresh.

845
00:31:26,460 --> 00:31:28,740
If not, where is it
losing its freshness?

846
00:31:28,740 --> 00:31:30,873
Maybe, it's spending
too much time in a DC.

847
00:31:30,873 --> 00:31:32,790
Maybe, it's spending too
much time in a truck.

848
00:31:32,790 --> 00:31:34,350
Maybe, it's not cold
enough in the truck,

849
00:31:34,350 --> 00:31:35,670
and so it's losing freshness.

850
00:31:35,670 --> 00:31:37,410
And what RFID lets
you do is it's

851
00:31:37,410 --> 00:31:40,230
like injecting a dye
into your bloodstream

852
00:31:40,230 --> 00:31:43,660
and looking to see how
the body is operating.

853
00:31:43,660 --> 00:31:46,170
And it lets you go
through all that

854
00:31:46,170 --> 00:31:48,133
and figure out how
your body is doing.

855
00:31:48,133 --> 00:31:49,800
And if you look at
different retailers--

856
00:31:49,800 --> 00:31:53,160
look at, for example, Walmart.

857
00:31:53,160 --> 00:31:55,500
This is publicly what
they've said they're doing.

858
00:31:55,500 --> 00:31:59,010
If you look at Best Buy, this is
what they're going after first.

859
00:31:59,010 --> 00:32:02,340
If you look at Lowe's--

860
00:32:02,340 --> 00:32:06,093
Lowe's is the big do it
yourself store-- this

861
00:32:06,093 --> 00:32:07,260
is what they're going after.

862
00:32:07,260 --> 00:32:09,052
So everyone's looking
at a different thing,

863
00:32:09,052 --> 00:32:11,160
going after a different
piece of the puzzle.

864
00:32:11,160 --> 00:32:13,330
I have lots of examples.

865
00:32:13,330 --> 00:32:15,480
I zoom in and look
at various examples,

866
00:32:15,480 --> 00:32:17,490
and in this audience
are hundreds

867
00:32:17,490 --> 00:32:20,220
of other examples, where you
zoom into this landscape.

868
00:32:20,220 --> 00:32:21,960
It's kind of like
using Google Earth,

869
00:32:21,960 --> 00:32:24,090
seeing something
interesting and zooming in.

870
00:32:24,090 --> 00:32:27,160
Looking at a map, zooming
in, looking at a house,

871
00:32:27,160 --> 00:32:29,580
you can do that, and I
think fundamentally this

872
00:32:29,580 --> 00:32:30,960
is what this industry needs.

873
00:32:30,960 --> 00:32:32,700
And academics are
the microscopes,

874
00:32:32,700 --> 00:32:34,200
and they're going
to figure out what

875
00:32:34,200 --> 00:32:37,530
the opportunities are and figure
out what the solutions are.

876
00:32:37,530 --> 00:32:39,720
In the interest of time,
what I'm going to do

877
00:32:39,720 --> 00:32:44,400
is I'm going to skip over
most of what I wanted

878
00:32:44,400 --> 00:32:46,395
to say except this one slide.

879
00:32:51,320 --> 00:32:53,750
And I'll make two
statements and end

880
00:32:53,750 --> 00:32:57,020
in about one or two minutes.

881
00:32:57,020 --> 00:32:58,733
When you have
operations-- for example,

882
00:32:58,733 --> 00:33:00,900
actually, if you can ignore
the slide for a second--

883
00:33:00,900 --> 00:33:02,750
if you look at
factory operations,

884
00:33:02,750 --> 00:33:05,100
you expect performance
to be tight.

885
00:33:05,100 --> 00:33:08,540
When you say, listen,
Father's Day is coming around.

886
00:33:08,540 --> 00:33:11,073
I want to move inventory
out of the store, shaving

887
00:33:11,073 --> 00:33:12,740
blades for Father's
Day, because there's

888
00:33:12,740 --> 00:33:15,073
a big gift-buying event.

889
00:33:15,073 --> 00:33:17,240
I want to move it out two
weeks before Father's Day,

890
00:33:17,240 --> 00:33:18,680
and then I'll start
advertisement on TV.

891
00:33:18,680 --> 00:33:20,480
I'm going to spend
millions of dollars.

892
00:33:20,480 --> 00:33:24,880
You expect that things
will work out quite well.

893
00:33:24,880 --> 00:33:25,380
Right?

894
00:33:25,380 --> 00:33:27,770
People will move stuff out,
the retail supply chain

895
00:33:27,770 --> 00:33:30,200
is going to operate
as you expected it,

896
00:33:30,200 --> 00:33:31,920
and things will happen
as you expected.

897
00:33:31,920 --> 00:33:33,590
And you can run
your ads and you're

898
00:33:33,590 --> 00:33:35,390
going to get good efficiency.

899
00:33:35,390 --> 00:33:39,260
The fact of the matter is that
in many systems where lots

900
00:33:39,260 --> 00:33:41,660
of human beings are involved--
and retail is one place

901
00:33:41,660 --> 00:33:44,240
and hospitals are
another place--

902
00:33:44,240 --> 00:33:46,430
if you thought that the
performance you would get

903
00:33:46,430 --> 00:33:47,990
would be like that-- that
is the bulk of the product

904
00:33:47,990 --> 00:33:49,573
would be moved out
on the day that you

905
00:33:49,573 --> 00:33:53,000
expect it to be moved out,
you're in for a disappointment.

906
00:33:53,000 --> 00:33:58,460
Because really, in most
supply chains, performance is,

907
00:33:58,460 --> 00:34:00,260
unfortunately, despite
the best efforts--

908
00:34:00,260 --> 00:34:03,260
and we're talking about the top
corporations on the planet--

909
00:34:03,260 --> 00:34:05,690
they lose.

910
00:34:05,690 --> 00:34:06,410
All right?

911
00:34:06,410 --> 00:34:08,690
If you're familiar with
the statistical process

912
00:34:08,690 --> 00:34:10,889
control, the Six
Sigma variation,

913
00:34:10,889 --> 00:34:14,900
the coefficient of performance
of execution in most supply

914
00:34:14,900 --> 00:34:19,040
chains in many endeavors
where RFID would be useful

915
00:34:19,040 --> 00:34:20,610
is rather poor.

916
00:34:20,610 --> 00:34:22,610
In fact, the coefficient
of performance

917
00:34:22,610 --> 00:34:26,659
for a promotion that I looked at
recently is something like 0.1,

918
00:34:26,659 --> 00:34:29,940
where it should really be 1.3.

919
00:34:29,940 --> 00:34:30,449
OK?

920
00:34:30,449 --> 00:34:33,239
So if you understand what
statistical process control is,

921
00:34:33,239 --> 00:34:36,449
just a metric, I can tell
you in a very succinct way,

922
00:34:36,449 --> 00:34:39,795
there is a whole lot of room for
improvement in the supply chain

923
00:34:39,795 --> 00:34:41,670
to make it tighter, and
that's why people are

924
00:34:41,670 --> 00:34:43,420
looking at the supply chain.

925
00:34:43,420 --> 00:34:47,360
I'll end with this slide.

926
00:34:47,360 --> 00:34:47,860
OK?

927
00:34:50,540 --> 00:34:52,940
I just want to point
to the bottom right.

928
00:34:52,940 --> 00:34:55,340
There's also a lot of
exciting stuff on how

929
00:34:55,340 --> 00:34:58,170
RFID might fundamentally
change the supply chain,

930
00:34:58,170 --> 00:35:00,650
and that's in some ways the
most exciting part of all this.

931
00:35:00,650 --> 00:35:02,370
Right?

932
00:35:02,370 --> 00:35:04,910
Is it going to make some
macroeconomic changes

933
00:35:04,910 --> 00:35:06,720
to the way we operate
as an economy?

934
00:35:06,720 --> 00:35:07,220
Right?

935
00:35:07,220 --> 00:35:08,810
Now, that's getting
a little highfaluting

936
00:35:08,810 --> 00:35:10,643
and maybe getting a
little beyond ourselves,

937
00:35:10,643 --> 00:35:11,990
but this is academics.

938
00:35:11,990 --> 00:35:13,440
This is what we should be doing.

939
00:35:13,440 --> 00:35:16,490
So I just wanted to say,
let's fantasize a little bit.

940
00:35:16,490 --> 00:35:20,050
The opportunities are
vast, but perhaps--

941
00:35:20,050 --> 00:35:23,470
excuse me-- perhaps,
we need to start also

942
00:35:23,470 --> 00:35:26,470
thinking at an aggregate level.

943
00:35:26,470 --> 00:35:28,570
Will retailers finally
say that, listen,

944
00:35:28,570 --> 00:35:31,733
with RFID, we own our
own supply chains.

945
00:35:31,733 --> 00:35:33,400
We don't need them
anymore, because it's

946
00:35:33,400 --> 00:35:35,560
not a competitive
advantage anymore.

947
00:35:35,560 --> 00:35:39,085
Maybe, Walmart essentially
becomes, as many of you

948
00:35:39,085 --> 00:35:43,360
know this, essentially just real
estate with the Walmart name

949
00:35:43,360 --> 00:35:44,890
on it or Target name on it.

950
00:35:44,890 --> 00:35:47,290
And all they do is
they own the shelves,

951
00:35:47,290 --> 00:35:50,740
and the products there, it's
like Amazon, and you go buy it,

952
00:35:50,740 --> 00:35:52,420
but it's completely
up to the supplier

953
00:35:52,420 --> 00:35:53,470
to stock these shelves.

954
00:35:53,470 --> 00:35:54,860
That's called pay-per-scan
for those of you

955
00:35:54,860 --> 00:35:56,027
that are familiar with this.

956
00:35:56,027 --> 00:35:57,640
Maybe that's what'll happen.

957
00:35:57,640 --> 00:35:58,720
Right?

958
00:35:58,720 --> 00:36:01,690
Maybe what'll happen is
the supply chain goes away,

959
00:36:01,690 --> 00:36:05,760
and UPS takes over the
world, or Federal Express,

960
00:36:05,760 --> 00:36:10,657
and essentially,
manufacturers, FedEx, or UPS,

961
00:36:10,657 --> 00:36:12,240
they're stuffed
straight to the shelf.

962
00:36:12,240 --> 00:36:14,370
That's it.

963
00:36:14,370 --> 00:36:15,130
Who knows?

964
00:36:15,130 --> 00:36:15,630
Right?

965
00:36:15,630 --> 00:36:19,480
Certainly, we're gaming this,
and if this community doesn't,

966
00:36:19,480 --> 00:36:21,130
who's going to game it?

967
00:36:21,130 --> 00:36:23,770
In fact, Brian and I were
fantasizing about that,

968
00:36:23,770 --> 00:36:26,650
and we call it the ultimately
fragmented supply chain.

969
00:36:26,650 --> 00:36:28,180
Maybe that's where it'll go.

970
00:36:28,180 --> 00:36:29,380
Right?

971
00:36:29,380 --> 00:36:33,580
Maybe in the end, this might
change the way offshoring

972
00:36:33,580 --> 00:36:34,750
is working today.

973
00:36:34,750 --> 00:36:36,730
If you're familiar with
Zara supply chains--

974
00:36:36,730 --> 00:36:38,950
Zara is a Spanish retailer--

975
00:36:38,950 --> 00:36:40,750
what they do is they change--

976
00:36:40,750 --> 00:36:43,570
they're very tight in
terms of fashion cycles.

977
00:36:43,570 --> 00:36:46,210
So every time, I suppose,
Britney Spears has a concert,

978
00:36:46,210 --> 00:36:49,270
within a week, they
have those exact clothes

979
00:36:49,270 --> 00:36:52,430
on the market for two weeks,
and then they're gone.

980
00:36:52,430 --> 00:36:56,500
So if you go to Zara, you get
exactly what you want, right

981
00:36:56,500 --> 00:36:58,540
there, and because
it's so fashionable,

982
00:36:58,540 --> 00:37:01,330
and it is so agile, very
few other retailers are

983
00:37:01,330 --> 00:37:02,685
able to compete.

984
00:37:02,685 --> 00:37:04,060
But the only way
they can do that

985
00:37:04,060 --> 00:37:06,730
is not by manufacturing in
Asia, because the lead times

986
00:37:06,730 --> 00:37:09,530
are too long, but they're
manufacturing locally.

987
00:37:09,530 --> 00:37:10,190
Why?

988
00:37:10,190 --> 00:37:13,100
Maybe with the
visibility RFID enables,

989
00:37:13,100 --> 00:37:15,183
maybe a lot of retailers
can do this sort of thing

990
00:37:15,183 --> 00:37:17,642
more efficiently, and maybe
it's going to affect and impact

991
00:37:17,642 --> 00:37:18,710
the way we do offshoring.

992
00:37:18,710 --> 00:37:21,380
We worry in this country
about offshoring.

993
00:37:21,380 --> 00:37:23,150
I'm sure in other
countries as well.

994
00:37:23,150 --> 00:37:25,490
Maybe some tectonic
changes are awaiting us

995
00:37:25,490 --> 00:37:28,360
on the supply chain, and how
will RFID play into that?

996
00:37:28,360 --> 00:37:32,210
I think it's worth
discussing that.

997
00:37:32,210 --> 00:37:33,590
So I'll conclude
simply by saying

998
00:37:33,590 --> 00:37:36,380
that, obviously, the first
place that technology is going

999
00:37:36,380 --> 00:37:38,570
to make an impact is by
tweaking things the way

1000
00:37:38,570 --> 00:37:39,650
they're done today.

1001
00:37:39,650 --> 00:37:41,840
Eventually, there might
be some evolution,

1002
00:37:41,840 --> 00:37:44,090
but fundamentally, there's
going to be transformation.

1003
00:37:44,090 --> 00:37:45,715
And it's the right
end of the spectrum,

1004
00:37:45,715 --> 00:37:49,010
the stuff to the right,
that this community

1005
00:37:49,010 --> 00:37:51,920
and no other community will
be able to study and enable.

1006
00:37:51,920 --> 00:37:57,800
And it's a great pleasure to
address you and fantasize,

1007
00:37:57,800 --> 00:38:01,198
but also, I look forward to
talking to you about this.

1008
00:38:01,198 --> 00:38:02,990
And have a wonderful
conference, and thanks

1009
00:38:02,990 --> 00:38:06,870
for letting me subject
you to my crazy thoughts.

1010
00:38:06,870 --> 00:38:07,701
Thank you.

1011
00:38:07,701 --> 00:38:11,068
[APPLAUSE]

1012
00:38:13,480 --> 00:38:15,700
PROFESSOR: So maybe a
couple questions for Sanjay,

1013
00:38:15,700 --> 00:38:17,080
while we set up the next panel.

1014
00:38:22,000 --> 00:38:23,980
GUEST SPEAKER: Any questions?

1015
00:38:23,980 --> 00:38:24,940
Yes.

1016
00:38:24,940 --> 00:38:26,620
AUDIENCE: So there's a
million papers to be written.

1017
00:38:26,620 --> 00:38:26,890
GUEST SPEAKER: Yeah.

1018
00:38:26,890 --> 00:38:29,650
AUDIENCE: How many need to be
written before the diffusion

1019
00:38:29,650 --> 00:38:30,827
curve starts to take off?

1020
00:38:30,827 --> 00:38:31,660
GUEST SPEAKER: Yeah.

1021
00:38:31,660 --> 00:38:33,400
That's a great question.

1022
00:38:33,400 --> 00:38:35,860
I think that every--

1023
00:38:35,860 --> 00:38:36,980
I don't know.

1024
00:38:36,980 --> 00:38:39,640
I think that in every
topic, every new area,

1025
00:38:39,640 --> 00:38:42,700
there's a time of freshness,
where the papers are

1026
00:38:42,700 --> 00:38:43,540
fresh and different.

1027
00:38:43,540 --> 00:38:46,247
And pretty soon, they start
grinding into each other,

1028
00:38:46,247 --> 00:38:47,080
and it's repetitive.

1029
00:38:47,080 --> 00:38:47,620
Right?

1030
00:38:47,620 --> 00:38:49,090
I think we have a ways to go.

1031
00:38:49,090 --> 00:38:52,820
If you take wireless, you take
the internet, take Google.

1032
00:38:52,820 --> 00:38:53,320
Right?

1033
00:38:53,320 --> 00:38:55,630
Google is a research
project out of Stanford.

1034
00:38:55,630 --> 00:38:57,070
Right?

1035
00:38:57,070 --> 00:38:58,955
We're 15 years into it.

1036
00:38:58,955 --> 00:39:00,580
Take the internet
for 40 years into it.

1037
00:39:00,580 --> 00:39:01,090
Right?

1038
00:39:01,090 --> 00:39:02,860
I think we have a ways to
go, but there's no doubt

1039
00:39:02,860 --> 00:39:03,818
that's going to set in.

1040
00:39:03,818 --> 00:39:05,810
I think we need to be
watchful of that as well.

1041
00:39:05,810 --> 00:39:06,400
So I agree.

1042
00:39:06,400 --> 00:39:08,260
So we'll balloon, but
then eventually, we'll

1043
00:39:08,260 --> 00:39:10,950
kill ourselves and let it narrow
So that's a great question.

1044
00:39:13,397 --> 00:39:15,230
AUDIENCE: Part of the
goal of this gathering

1045
00:39:15,230 --> 00:39:18,290
was to try to bring together
end users and researchers.

1046
00:39:18,290 --> 00:39:20,755
So the end user and the
researcher, based on--

1047
00:39:20,755 --> 00:39:22,130
I'm sorry, the
researcher that is

1048
00:39:22,130 --> 00:39:24,060
doing the work based on that.

1049
00:39:24,060 --> 00:39:26,117
How do we facilitate
that to be the case so

1050
00:39:26,117 --> 00:39:28,200
that the research is being
driven in the direction

1051
00:39:28,200 --> 00:39:29,090
that business is going?

1052
00:39:29,090 --> 00:39:30,830
PROFESSOR: I think
that's a great question.

1053
00:39:30,830 --> 00:39:32,288
When we started
the Auto-ID center,

1054
00:39:32,288 --> 00:39:33,980
we basically made some stuff up.

1055
00:39:33,980 --> 00:39:35,340
Right?

1056
00:39:35,340 --> 00:39:39,380
And we went and just evangelized
it, and we were lucky.

1057
00:39:39,380 --> 00:39:42,330
We had some key end users
involved in the very beginning.

1058
00:39:42,330 --> 00:39:44,060
So we weren't completely
making stuff up.

1059
00:39:44,060 --> 00:39:47,510
We had some directed
things we were going after,

1060
00:39:47,510 --> 00:39:49,160
and we were right.

1061
00:39:49,160 --> 00:39:50,280
It's lucky.

1062
00:39:50,280 --> 00:39:50,780
Right?

1063
00:39:50,780 --> 00:39:54,350
But we can't afford to
wait on luck in the future.

1064
00:39:54,350 --> 00:39:56,535
The good news now is the
end users really get this.

1065
00:39:56,535 --> 00:39:58,160
In fact, in many
ways, they get it more

1066
00:39:58,160 --> 00:39:59,870
than the larger
academic community.

1067
00:39:59,870 --> 00:40:00,830
Right?

1068
00:40:00,830 --> 00:40:02,670
And we have a lot
of end users here.

1069
00:40:02,670 --> 00:40:06,080
So I think that the time
is now ripe for researchers

1070
00:40:06,080 --> 00:40:09,230
to pick up esoteric topics which
also happen to be practical.

1071
00:40:09,230 --> 00:40:10,550
How rarely does that happen?

1072
00:40:10,550 --> 00:40:12,140
Right?

1073
00:40:12,140 --> 00:40:14,780
Establishing that channel, there
are two things that happen.

1074
00:40:14,780 --> 00:40:16,260
One, the people
need to be there.

1075
00:40:16,260 --> 00:40:17,710
The good news is that
they're there now.

1076
00:40:17,710 --> 00:40:18,890
You have Mike Rose
sitting there.

1077
00:40:18,890 --> 00:40:19,973
You'll have Dick Campbell.

1078
00:40:19,973 --> 00:40:20,830
You have Tim Mill.

1079
00:40:20,830 --> 00:40:21,070
Right?

1080
00:40:21,070 --> 00:40:22,760
You have Richard
[? Esram ?] back there.

1081
00:40:22,760 --> 00:40:23,346
Right?

1082
00:40:23,346 --> 00:40:24,190
I believe Gillette, right?

1083
00:40:24,190 --> 00:40:25,773
So you have people
now willing to have

1084
00:40:25,773 --> 00:40:29,120
these deep conversations and
they're very well equipped,

1085
00:40:29,120 --> 00:40:30,680
but what's missing
is the medium.

1086
00:40:30,680 --> 00:40:31,180
Right?

1087
00:40:31,180 --> 00:40:33,020
I think this is a great medium.

1088
00:40:33,020 --> 00:40:35,390
I think we need to
present research papers

1089
00:40:35,390 --> 00:40:39,560
and have business guys
go, that's pointless,

1090
00:40:39,560 --> 00:40:40,790
but this I wish you would do.

1091
00:40:40,790 --> 00:40:42,710
We're not doing
that, and I think

1092
00:40:42,710 --> 00:40:45,720
that applies to all of academia.

1093
00:40:45,720 --> 00:40:46,793
Yeah, Chris.

1094
00:40:46,793 --> 00:40:48,710
AUDIENCE: I've got a
general question for you.

1095
00:40:48,710 --> 00:40:54,650
[INAUDIBLE] challenges is that
we all present different search

1096
00:40:54,650 --> 00:40:55,760
communities.

1097
00:40:55,760 --> 00:41:00,200
[INAUDIBLE] conferences, whereas
communication guys and RFID

1098
00:41:00,200 --> 00:41:00,800
go somewhere.

1099
00:41:00,800 --> 00:41:02,960
And every single time--
or at least at this stage,

1100
00:41:02,960 --> 00:41:05,252
we can confirm that we can
motivate any of the problems

1101
00:41:05,252 --> 00:41:06,440
that you mentioned.

1102
00:41:06,440 --> 00:41:09,950
[INAUDIBLE] sometimes
people in those communities

1103
00:41:09,950 --> 00:41:12,910
are not familiar with those
special challenges of RFID.

1104
00:41:12,910 --> 00:41:14,870
So do you think that in
a couple of years time,

1105
00:41:14,870 --> 00:41:19,190
we'll see a dedicated RFID
conference that covers a range,

1106
00:41:19,190 --> 00:41:21,442
maybe not all but a range
of the issues you mentioned?

1107
00:41:21,442 --> 00:41:23,150
GUEST SPEAKER: I think
that might happen.

1108
00:41:23,150 --> 00:41:24,440
I think you're
pointing at the deeper

1109
00:41:24,440 --> 00:41:25,550
problem of systems design.

1110
00:41:25,550 --> 00:41:25,730
Right?

1111
00:41:25,730 --> 00:41:27,470
And first of all, I think
we have too many conferences

1112
00:41:27,470 --> 00:41:28,610
in academic field.

1113
00:41:28,610 --> 00:41:31,730
They're too sub-selective,
and only those people

1114
00:41:31,730 --> 00:41:32,730
go to those conferences.

1115
00:41:32,730 --> 00:41:34,730
It's a very nice, cozy
shelter, where you don't really

1116
00:41:34,730 --> 00:41:36,022
have to talk to the other guys.

1117
00:41:36,022 --> 00:41:36,560
Right?

1118
00:41:36,560 --> 00:41:39,815
So I do think that an RFID
conference, an RFID systems

1119
00:41:39,815 --> 00:41:42,440
community will evolve, and maybe
this is the start of it, which

1120
00:41:42,440 --> 00:41:44,240
is why I congratulate everyone.

1121
00:41:44,240 --> 00:41:45,855
But I think we also
need to watch out

1122
00:41:45,855 --> 00:41:48,230
that that doesn't become an
ultra-academic conference not

1123
00:41:48,230 --> 00:41:50,900
connected to reality
which in itself becomes

1124
00:41:50,900 --> 00:41:53,325
a kind of mutant animal
that no one's interested in.

1125
00:41:53,325 --> 00:41:55,450
So we need to keep the end
user community involved.

1126
00:41:55,450 --> 00:41:57,500
That's my view anyway,
and that was the success,

1127
00:41:57,500 --> 00:41:59,870
as you remember, from
the Auto-ID center,

1128
00:41:59,870 --> 00:42:01,208
but I hope we can do that.

1129
00:42:01,208 --> 00:42:02,500
PROFESSOR: Thank you very much.

1130
00:42:02,500 --> 00:42:04,650
GUEST SPEAKER:
Thanks a lot, Steve.