WEBVTT
Language: en-US

1
00:00:01.110 --> 00:00:03.890
<v Speaker 1>Yeah, I, I think all of the same things
apply, right?

2
00:00:04.250 --> 00:00:08.930
<v Speaker 1>And, you know, there's a transition, I
think people make in how they think about the

3
00:00:08.930 --> 00:00:13.149
<v Speaker 1>interactions with the computer when
they're interacting with, um, You know, a

4
00:00:13.149 --> 00:00:16.989
<v Speaker 1>generative AI based system, whether that's
a traditional one or a, uh, a agentic one

5
00:00:16.989 --> 00:00:17.879
<v Speaker 1>or what have you.

6
00:00:18.239 --> 00:00:21.860
<v Speaker 1>Um, and I think that is thinking it is a
bit more like, um, again, like a colleague,

7
00:00:21.959 --> 00:00:22.479
<v Speaker 1>right?

8
00:00:22.840 --> 00:00:26.260
<v Speaker 1>It's just some system with all these kind
of powers and abilities and knowledge,

9
00:00:26.440 --> 00:00:26.790
<v Speaker 1>right?

10
00:00:27.040 --> 00:00:28.860
<v Speaker 1>But it's not a fool foolproof thing.

11
00:00:29.399 --> 00:00:29.879
<v Speaker 1>Yeah.

12
00:00:30.280 --> 00:00:36.040
<v Speaker 1>um, and yeah, just like, um, your, your
human colleagues or what have you, you have

13
00:00:36.040 --> 00:00:38.919
<v Speaker 1>some issue that you, you want to solve.

14
00:00:40.500 --> 00:00:45.700
<v Speaker 2>Welcome to the trend detection podcast
powered by Siemens, and now your host Nel

15
00:00:45.700 --> 00:00:46.220
<v Speaker 2>Sullivan.

16
00:00:49.639 --> 00:00:53.630
<v Niall Sullivan>Hello and welcome to the latest episode of
the Trend detection podcast.

17
00:00:54.330 --> 00:00:58.369
<v Niall Sullivan>And today I'm joined by my colleague James
Loach, who's the head of research for

18
00:00:58.369 --> 00:01:00.009
<v Niall Sullivan>Sensii Predicting Maintenance.

19
00:01:00.630 --> 00:01:04.690
<v Niall Sullivan>Um, and James has been on a few times
before and we've we've of course talked about

20
00:01:04.690 --> 00:01:09.809
<v Niall Sullivan>some really interesting but also technical
subjects, but this one is really, um,

21
00:01:10.089 --> 00:01:10.970
<v Niall Sullivan>really exciting.

22
00:01:11.720 --> 00:01:15.279
<v Niall Sullivan>And potentially a really exciting
development in the world of predictive maintenance

23
00:01:15.279 --> 00:01:15.760
<v Niall Sullivan>as well.

24
00:01:15.959 --> 00:01:19.480
<v Niall Sullivan>Um, I'm sure it's a term you've heard
around Agentic AI and we're going to talk

25
00:01:19.480 --> 00:01:22.709
<v Niall Sullivan>about how it relates to, to predictive
maintenance.

26
00:01:22.959 --> 00:01:27.639
<v Niall Sullivan>So I guess before we go into it, for those
who don't know James, maybe just a quick

27
00:01:27.639 --> 00:01:31.580
<v Niall Sullivan>line from you in terms of the work you and
the team do within the Sensai team.

28
00:01:31.720 --> 00:01:32.339
<v Niall Sullivan>It'd be great.

29
00:01:33.580 --> 00:01:34.150
<v Speaker 1>Yeah, hi, Niall.

30
00:01:34.250 --> 00:01:35.989
<v Speaker 1>um, yeah, very, very happy to be here.

31
00:01:36.069 --> 00:01:40.910
<v Speaker 1>And yeah, in terms of a, a, a brief
introduction, so I'm head of research at Sensi

32
00:01:41.150 --> 00:01:44.129
<v Speaker 1>and have been since the, the early days
of, of the company.

33
00:01:44.629 --> 00:01:48.849
<v Speaker 1>And, and during that time, myself and my
group have been, um, basically responsible

34
00:01:48.849 --> 00:01:53.430
<v Speaker 1>for, for, for the algorithms that, that
underpin the product, um, and then

35
00:01:53.430 --> 00:01:57.470
<v Speaker 1>algorithm, you know, um, related
functionality in the, in the user interface and,

36
00:01:57.589 --> 00:01:58.569
<v Speaker 1>and, and so forth.

37
00:01:59.180 --> 00:01:59.550
<v Speaker 1>OK.

38
00:02:01.250 --> 00:02:02.459
<v Niall Sullivan>Fantastic, fantastic.

39
00:02:02.580 --> 00:02:06.220
<v Niall Sullivan>So yeah, I mean, let's, let's dive
straight in, and I guess we, we should set the

40
00:02:06.220 --> 00:02:08.850
<v Niall Sullivan>scene first of all for and, and start with
the basics.

41
00:02:08.910 --> 00:02:13.529
<v Niall Sullivan>So, um, a lot of people won't be aware of
this term agentic AI or they might have

42
00:02:13.529 --> 00:02:15.539
<v Niall Sullivan>heard about it, but don't know how it
relates to them.

43
00:02:15.619 --> 00:02:19.839
<v Niall Sullivan>So maybe you could just start by giving us
sort of a, a brief explanation what

44
00:02:19.940 --> 00:02:25.500
<v Niall Sullivan>agentic AI is and how does it differ from
traditional AI in an industrial context.

45
00:02:26.559 --> 00:02:26.979
<v Speaker 1>Yeah.

46
00:02:27.240 --> 00:02:27.440
<v Speaker 1>OK.

47
00:02:27.520 --> 00:02:31.800
<v Speaker 1>I mean, I, I, I think the best way to
think about it is in terms of the sort of the,

48
00:02:31.880 --> 00:02:34.279
<v Speaker 1>the user experience of, of this.

49
00:02:34.720 --> 00:02:41.199
<v Speaker 1>So an Eergenic AI system, um, relates to
you some way, um, in, in the, in the way

50
00:02:41.199 --> 00:02:44.270
<v Speaker 1>that a colleague or a, a worker would, OK?

51
00:02:44.639 --> 00:02:49.479
<v Speaker 1>So it's, um, it's an entity, um, that, you
know, has some general knowledge of the

52
00:02:49.479 --> 00:02:55.520
<v Speaker 1>world, has some knowledge of your
particular use case or your particular setup.

53
00:02:55.979 --> 00:03:00.860
<v Speaker 1>And importantly, has the ability to, um,
to do actions of different types.

54
00:03:00.960 --> 00:03:04.559
<v Speaker 1>Everything from looking up information to
perhaps, you know, interacting with the,

55
00:03:04.619 --> 00:03:11.440
<v Speaker 1>with the physical world, um, and the
ability to go off and, and do sort of delegated

56
00:03:11.440 --> 00:03:12.369
<v Speaker 1>tasks for you.

57
00:03:12.550 --> 00:03:14.490
<v Speaker 1>So this is a sense in which it's like a
worker.

58
00:03:14.740 --> 00:03:18.100
<v Speaker 1>It's a system that you can specify the
roughly what you want it to do.

59
00:03:18.399 --> 00:03:23.050
<v Speaker 1>You can expect it to go away and approach
that task with, with common sense, um, you

60
00:03:23.050 --> 00:03:25.080
<v Speaker 1>know, doing, um, a series of actions.

61
00:03:25.529 --> 00:03:29.449
<v Speaker 1>Figuring out what we wanted it to do and,
and coming back to you when it's done or

62
00:03:29.449 --> 00:03:30.750
<v Speaker 1>it, or it needs your help.

63
00:03:31.100 --> 00:03:33.440
<v Speaker 1>So it's, it's that sort, sort of
experience.

64
00:03:33.729 --> 00:03:36.410
<v Speaker 1>And, and you asked how it differs from
traditional AI.

65
00:03:36.729 --> 00:03:42.210
<v Speaker 1>So, I mean, uh, traditional AI is, I guess
is sort of a, an interesting term in

66
00:03:42.210 --> 00:03:42.660
<v Speaker 1>itself, right?

67
00:03:42.729 --> 00:03:47.600
<v Speaker 1>What, what we should take that, that to
mean, but perhaps the best thing to do is to

68
00:03:47.600 --> 00:03:51.809
<v Speaker 1>contrast it with, um, a simple chatbot,
um, you know, which obviously people have

69
00:03:51.809 --> 00:03:54.190
<v Speaker 1>plenty of experience with now.

70
00:03:54.639 --> 00:03:59.880
<v Speaker 1>So, um, you know, a simple chatbot can be
ergentic, you know, in a very sort of

71
00:03:59.880 --> 00:04:00.880
<v Speaker 1>simple way, right?

72
00:04:01.039 --> 00:04:05.119
<v Speaker 1>So, I mean, typically when you, you talk
to a chatbot, it's just a thing that knows

73
00:04:05.119 --> 00:04:09.160
<v Speaker 1>some stuff and you, you ask it some
questions, you, you get some answers, something

74
00:04:09.160 --> 00:04:09.679
<v Speaker 1>like that.

75
00:04:09.990 --> 00:04:14.039
<v Speaker 1>And it may have the ability inside it to
do some simple things like do some

76
00:04:14.039 --> 00:04:15.520
<v Speaker 1>information lookup for you, right?

77
00:04:15.600 --> 00:04:19.720
<v Speaker 1>Or maybe delegate you to another chatbot
or, move you to some bit of the app,

78
00:04:19.920 --> 00:04:20.940
<v Speaker 1>something like that.

79
00:04:21.450 --> 00:04:21.880
<v Speaker 1>OK?

80
00:04:22.480 --> 00:04:27.059
<v Speaker 1>So I, we, we might call that traditional,
like a sort of an example of traditional

81
00:04:27.320 --> 00:04:29.179
<v Speaker 1>AI, you know, in an interactive way.

82
00:04:29.239 --> 00:04:32.160
<v Speaker 1>Now, the agentic version of that is just
something much deeper, right?

83
00:04:32.239 --> 00:04:36.149
<v Speaker 1>So it's not something that you're just
doing a little Q&amp;A with to get some info.

84
00:04:36.510 --> 00:04:39.480
<v Speaker 1>It's something, you know, you're, you're
specifying a task to it and you're, you're

85
00:04:39.480 --> 00:04:40.670
<v Speaker 1>letting it get on with it, right?

86
00:04:40.880 --> 00:04:43.359
<v Speaker 1>You know, maybe it come back, come back,
it comes back to you.

87
00:04:43.959 --> 00:04:47.489
<v Speaker 1>In an hour with its reports or, or, or
what have you, right?

88
00:04:47.839 --> 00:04:52.480
<v Speaker 1>Yeah, so it's a similar kind of thing, I
guess, but just, just deeper, right, and

89
00:04:52.480 --> 00:04:53.220
<v Speaker 1>more sophisticated.

90
00:04:54.790 --> 00:04:58.149
<v Niall Sullivan>Yeah, and so, and when we, so when we
think about sort of like generative AI as

91
00:04:58.149 --> 00:05:00.709
<v Niall Sullivan>well, what, what is sort of the genetic AI
in relation to that?

92
00:05:00.750 --> 00:05:05.269
<v Niall Sullivan>Is it sort of the, the next sort of step
along, you know, in terms of technology

93
00:05:05.269 --> 00:05:05.790
<v Niall Sullivan>breakthrough?

94
00:05:05.910 --> 00:05:10.170
<v Niall Sullivan>Is that the best way to describe it, or
are they sort of sat alongside each other?

95
00:05:11.149 --> 00:05:16.250
<v Speaker 1>Yeah, I mean, generative AI to me is, is,
is like, um, uh, underlying technology,

96
00:05:16.630 --> 00:05:17.019
<v Speaker 1>you know.

97
00:05:17.309 --> 00:05:22.309
<v Speaker 1>So in, in the, the, the context of the
things we're just talking about, so, you

98
00:05:22.309 --> 00:05:26.100
<v Speaker 1>know, the, The most important generative
AI technology that would be, would be a

99
00:05:26.100 --> 00:05:30.739
<v Speaker 1>language model, OK, which is something
that really just can can continue text for

100
00:05:30.739 --> 00:05:31.609
<v Speaker 1>you, OK?

101
00:05:31.750 --> 00:05:36.000
<v Speaker 1>It has some model of all the text on the
internet and so forth and given some

102
00:05:36.160 --> 00:05:38.529
<v Speaker 1>prompt, you know, it can continue that,
right?

103
00:05:38.890 --> 00:05:43.260
<v Speaker 1>So that's like sort of an underlying thing
and not just continuation of text is

104
00:05:43.260 --> 00:05:45.179
<v Speaker 1>something that can be turned into a
chatbot, right?

105
00:05:45.420 --> 00:05:49.260
<v Speaker 1>Because instead of continuing text, you
know, just like in paragraphs or what have

106
00:05:49.260 --> 00:05:49.579
<v Speaker 1>you.

107
00:05:50.010 --> 00:05:53.250
<v Speaker 1>Um, the thing can be formed into something
that's completing like a movie script,

108
00:05:53.459 --> 00:05:58.190
<v Speaker 1>you know, and you, the user is playing one
part and it's, you know, completing by

109
00:05:58.190 --> 00:06:01.630
<v Speaker 1>filling in the next part as a helpful
assistant, you know.

110
00:06:01.989 --> 00:06:07.109
<v Speaker 1>So you have this text model that can be
used to create this sort of Q&amp;A, um, back

111
00:06:07.109 --> 00:06:07.850
<v Speaker 1>and forth.

112
00:06:08.989 --> 00:06:09.359
<v Speaker 1>Right?

113
00:06:09.679 --> 00:06:11.529
<v Speaker 1>Um, and then you sort of build on top of
that, right?

114
00:06:11.630 --> 00:06:18.480
<v Speaker 1>So instead of just doing this Q&amp;A
prompted, um, you can allow the model to, to

115
00:06:18.480 --> 00:06:19.320
<v Speaker 1>use tools.

116
00:06:19.600 --> 00:06:22.799
<v Speaker 1>So instead of sending a message to you, it
can send a message to some computer

117
00:06:22.799 --> 00:06:26.040
<v Speaker 1>system to retrieve some information or, or
do something like that.

118
00:06:26.399 --> 00:06:26.790
<v Speaker 1>OK?

119
00:06:27.200 --> 00:06:29.600
<v Speaker 1>So the simple back and forth is a simple
chatbot.

120
00:06:30.040 --> 00:06:33.600
<v Speaker 1>You give it a little lookup tool that
becomes a slightly more sophisticated chatbot.

121
00:06:34.230 --> 00:06:36.950
<v Speaker 1>And yeah, you just make it more
sophisticated, right?

122
00:06:37.320 --> 00:06:43.440
<v Speaker 1>So you, you, uh, fine tune these models,
um, to do extended tasks, right?

123
00:06:43.549 --> 00:06:47.160
<v Speaker 1>And maintain coherence over extended
periods of time, right?

124
00:06:47.390 --> 00:06:50.350
<v Speaker 1>You have them reason in the background
about things.

125
00:06:50.480 --> 00:06:54.980
<v Speaker 1>You give them dozens of different tools to
use, you give them a sort of memory that

126
00:06:54.980 --> 00:06:58.200
<v Speaker 1>they can manage, and it, it all sort of
builds up, right?

127
00:06:58.600 --> 00:07:02.679
<v Speaker 1>So generous VIA to me is a sort of, it's
either like um an umbrella term or it's a

128
00:07:02.679 --> 00:07:07.239
<v Speaker 1>description of the fundamental technology
and then these systems that we're talking

129
00:07:07.239 --> 00:07:10.760
<v Speaker 1>about today are just things that you sort
of build with that build with that

130
00:07:10.760 --> 00:07:11.790
<v Speaker 1>technology, if that makes sense.

131
00:07:11.880 --> 00:07:13.480
<v Niall Sullivan>Yeah, no, no, it does, it does.

132
00:07:13.839 --> 00:07:17.790
<v Niall Sullivan>And so if we're taking that, you know,
that wonderful description you just gave and

133
00:07:17.790 --> 00:07:22.359
<v Niall Sullivan>apply it to predicting maintenance, so
how, how does agentic AI enhance sort of

134
00:07:22.359 --> 00:07:25.739
<v Niall Sullivan>predictive maintenance solutions like
Senseii, for example?

135
00:07:27.149 --> 00:07:32.269
<v Speaker 1>Um, I, I think the, the best way to think
about this actually sort of harks back to

136
00:07:32.269 --> 00:07:36.179
<v Speaker 1>the, um, to Sensai before the, the
generative AI days, right?

137
00:07:36.390 --> 00:07:40.950
<v Speaker 1>Because it's sort of, um, an image that
we've had of the product in a way that we've

138
00:07:40.950 --> 00:07:41.690
<v Speaker 1>structured it.

139
00:07:42.429 --> 00:07:46.369
<v Speaker 1>That, that still applies and is sort of
made a lot, a lot more real and powerful by

140
00:07:46.369 --> 00:07:47.790
<v Speaker 1>agentic AI, OK?

141
00:07:48.000 --> 00:07:52.279
<v Speaker 1>And that thing is basically thinking about
the Sensai application, um, it's

142
00:07:52.279 --> 00:07:55.459
<v Speaker 1>something like, uh, like a colleague, OK?

143
00:07:55.799 --> 00:08:01.619
<v Speaker 1>It's, um, It's a system that sits there
and has this ability to look at all the data

144
00:08:01.619 --> 00:08:04.119
<v Speaker 1>at once that's coming in from all your
assets.

145
00:08:04.429 --> 00:08:06.119
<v Speaker 1>That makes it quite an unusual colleague,
right?

146
00:08:06.429 --> 00:08:10.429
<v Speaker 1>And then in its, in its traditional days,
it's got some ability to reason over that

147
00:08:10.429 --> 00:08:14.190
<v Speaker 1>using traditional machine learning, so
forth and tell you when it thinks there might

148
00:08:14.190 --> 00:08:18.459
<v Speaker 1>be a problem and explain what, what's,
what, what, what's, what's, what's happening,

149
00:08:18.570 --> 00:08:18.989
<v Speaker 1>right?

150
00:08:19.390 --> 00:08:22.250
<v Speaker 1>So, you know, it's a system that's got
access to all this data.

151
00:08:22.720 --> 00:08:23.910
<v Speaker 1>It can look at it.

152
00:08:24.149 --> 00:08:27.760
<v Speaker 1>It can tell you when there's something up
and, you know, help, help you decide what

153
00:08:27.760 --> 00:08:27.950
<v Speaker 1>to do.

154
00:08:27.959 --> 00:08:29.670
<v Speaker 1>It's sort of a decision support system.

155
00:08:29.839 --> 00:08:31.059
<v Speaker 1>So Sensi is always.

156
00:08:31.730 --> 00:08:34.559
<v Speaker 1>Um, always sort of been that, right?

157
00:08:35.169 --> 00:08:41.299
<v Speaker 1>Um, what generative AI in general and this
agenic AI, um, specifically does is it

158
00:08:41.299 --> 00:08:45.190
<v Speaker 1>sort of, in some way changes the, the
characteristics of that colleague that, that

159
00:08:45.190 --> 00:08:46.359
<v Speaker 1>Sensi is, right?

160
00:08:46.760 --> 00:08:48.650
<v Speaker 1>You know, and that it makes it, um.

161
00:08:49.510 --> 00:08:51.200
<v Speaker 1>A lot smarter than it was.

162
00:08:51.270 --> 00:08:54.919
<v Speaker 1>It brings, you know, general knowledge and
common sense to bear.

163
00:08:55.440 --> 00:09:00.000
<v Speaker 1>And then via the tools that we were just
talking about, you know, um, the, the

164
00:09:00.000 --> 00:09:05.760
<v Speaker 1>ability to pull in information from
different places, um, to, to synthesize

165
00:09:05.760 --> 00:09:09.750
<v Speaker 1>information and really just to give you,
um, to do this tasks much better, right?

166
00:09:10.080 --> 00:09:14.869
<v Speaker 1>So to be more accurate and more precise in
telling you when that's something that

167
00:09:14.869 --> 00:09:16.500
<v Speaker 1>might be wrong.

168
00:09:17.270 --> 00:09:22.099
<v Speaker 1>Um, giving you much richer and more
detailed and again, more precise explanations of

169
00:09:22.099 --> 00:09:24.190
<v Speaker 1>what that, that could be, right?

170
00:09:24.429 --> 00:09:25.729
<v Speaker 1>Using information from the.

171
00:09:26.530 --> 00:09:29.080
<v Speaker 1>The, the, the sensor already, already has,
right?

172
00:09:29.330 --> 00:09:32.510
<v Speaker 1>OK, and coupling that in with
documentation and so forth.

173
00:09:33.049 --> 00:09:36.530
<v Speaker 1>Um, and then in the interactive part of
the application when you're trying to dig

174
00:09:36.530 --> 00:09:41.070
<v Speaker 1>into something to figure it out and, and
work out what to do, you just have a much

175
00:09:41.070 --> 00:09:45.229
<v Speaker 1>more powerful system, um, at, at your
disposal, right?

176
00:09:45.650 --> 00:09:49.979
<v Speaker 1>Um, so yeah, again, something that can
browse across the data, across the

177
00:09:49.979 --> 00:09:53.890
<v Speaker 1>documentation, you know, ultimately
something that can do web searches and what have

178
00:09:53.890 --> 00:09:54.260
<v Speaker 1>you.

179
00:09:54.630 --> 00:09:57.619
<v Speaker 1>And just bring all that stuff into one
place, OK?

180
00:09:57.869 --> 00:10:02.390
<v Speaker 1>So Site is like this kind of strange
maintenance colleague that you sort of work

181
00:10:02.390 --> 00:10:03.969
<v Speaker 1>with as part, part of your team.

182
00:10:04.590 --> 00:10:08.510
<v Speaker 1>All of these systems just make it a, a
much, you know, better colleague, like a much

183
00:10:08.510 --> 00:10:12.429
<v Speaker 1>more useful one, a much smarter one, you
know, yeah.

184
00:10:13.200 --> 00:10:16.320
<v Niall Sullivan>It's, it, it, you know, it's interesting,
James, what you say about, you know, sort

185
00:10:16.320 --> 00:10:17.789
<v Niall Sullivan>of before generative AI.

186
00:10:17.919 --> 00:10:22.400
<v Niall Sullivan>If there was a time, it feels like it's
been around forever so much talk around it,

187
00:10:22.479 --> 00:10:25.200
<v Niall Sullivan>but it is interesting that, cause it's a
really good point to make.

188
00:10:25.320 --> 00:10:29.900
<v Niall Sullivan>Sensi was, you know, even before Gen AI,
before Gene AI was already demonstrating,

189
00:10:30.239 --> 00:10:31.760
<v Niall Sullivan>you know, big success and stuff.

190
00:10:31.840 --> 00:10:35.359
<v Niall Sullivan>So I guess this is about taking it a step
further in terms of those like new

191
00:10:35.359 --> 00:10:36.440
<v Niall Sullivan>efficiencies and.

192
00:10:37.190 --> 00:10:40.270
<v Niall Sullivan>Make, make people work more or helping
people to work more, more efficient.

193
00:10:40.390 --> 00:10:43.070
<v Niall Sullivan>I guess it's just taking that a step
further with new technology, right?

194
00:10:43.429 --> 00:10:43.770
<v Niall Sullivan>Um,

195
00:10:44.030 --> 00:10:45.669
<v Speaker 1>yeah, and it's, it's a very strange thing.

196
00:10:45.710 --> 00:10:49.390
<v Speaker 1>I mean, I, I've not really gone on, gone
into it on this call, but, you know,

197
00:10:50.140 --> 00:10:54.270
<v Speaker 1>discuss a lot about what, what agents
actually are and the technical definitions of

198
00:10:54.270 --> 00:10:55.090
<v Speaker 1>these things.

199
00:10:55.440 --> 00:11:00.590
<v Speaker 1>Um, but, you know, there there is a way of
looking at the sensor application before

200
00:11:00.590 --> 00:11:03.890
<v Speaker 1>generative AI long ago and saying it is an
aergentic application.

201
00:11:04.280 --> 00:11:09.109
<v Speaker 1>Because it's a system that you, you hand
off a task to the monitoring stuff and

202
00:11:09.109 --> 00:11:12.409
<v Speaker 1>getting back to you when it's, it's got
some, some, some problem, right?

203
00:11:12.909 --> 00:11:15.869
<v Speaker 1>And it can do that task and in a sense,
act as an agent.

204
00:11:16.599 --> 00:11:19.549
<v Speaker 1>Um, you know, without any generative AI in
it, right?

205
00:11:19.940 --> 00:11:22.650
<v Speaker 1>Um, and, and this is, this is a nice
thing.

206
00:11:22.729 --> 00:11:27.530
<v Speaker 1>It's just the way that we sort of formed
the product and, you know, it's, in some

207
00:11:27.530 --> 00:11:30.609
<v Speaker 1>sense, it's, it's by accident that these
technologies have just come along, right?

208
00:11:30.890 --> 00:11:33.200
<v Speaker 1>But they just slot in very, very nicely,
right?

209
00:11:33.530 --> 00:11:37.729
<v Speaker 1>You know, you've got the system doing a
task in a traditional, some, somewhat

210
00:11:37.729 --> 00:11:41.450
<v Speaker 1>traditional machine learning way, and you
just have the ability very naturally to

211
00:11:41.450 --> 00:11:46.059
<v Speaker 1>inject this kind of, um, yeah,
intelligence and fantastic power into it.

212
00:11:46.849 --> 00:11:48.799
<v Speaker 1>And just have it just do its job better,
right?

213
00:11:48.969 --> 00:11:50.950
<v Speaker 1>With the same general form, right?

214
00:11:52.919 --> 00:11:59.479
<v Niall Sullivan>Exactly, exactly, and I guess, er, cos
you've mentioned already a few, um, I guess

215
00:11:59.479 --> 00:12:02.400
<v Niall Sullivan>outside of industrial environment, you
know, examples of how agents might, you know,

216
00:12:02.599 --> 00:12:06.440
<v Niall Sullivan>help to make the sort of task more
efficient, but if we talk about in a maintenance

217
00:12:06.440 --> 00:12:09.799
<v Niall Sullivan>environment specifically, can we, could
you give an example of sort of where, where

218
00:12:09.799 --> 00:12:11.000
<v Niall Sullivan>this might apply?

219
00:12:12.080 --> 00:12:14.880
<v Niall Sullivan>Where you could apply the um AI agent.

220
00:12:16.630 --> 00:12:21.030
<v Speaker 1>Yeah, so, um, you, you can think of it in,
in diff different levels, right?

221
00:12:21.190 --> 00:12:26.099
<v Speaker 1>So to begin with, you might think about it
doing more, um, sophisticated, um,

222
00:12:26.750 --> 00:12:31.299
<v Speaker 1>information lookup and, um, synthesis of
information, right?

223
00:12:31.580 --> 00:12:36.270
<v Speaker 1>You know, to help with standard
application tasks like spotting problems, explaining

224
00:12:36.270 --> 00:12:37.989
<v Speaker 1>what might be wrong, um.

225
00:12:38.770 --> 00:12:40.690
<v Speaker 1>And, and, and so forth, right?

226
00:12:41.090 --> 00:12:45.250
<v Speaker 1>Um, you can, OK, so, so you can think of
in terms of the sort of things the product

227
00:12:45.250 --> 00:12:48.710
<v Speaker 1>has, has already done, um, like that.

228
00:12:49.179 --> 00:12:55.409
<v Speaker 1>Um, in terms of exploratory things, which
AgenicIA really does excel at, I mean, we

229
00:12:55.409 --> 00:12:59.969
<v Speaker 1>have, we have this basic task of trying to
figure out what, what's going on, right?

230
00:13:00.090 --> 00:13:01.650
<v Speaker 1>So, so maybe the system is.

231
00:13:02.469 --> 00:13:07.000
<v Speaker 1>Raised some alerts, giving you some
possible explanations and things and back that

232
00:13:07.000 --> 00:13:07.780
<v Speaker 1>up, right?

233
00:13:08.200 --> 00:13:13.229
<v Speaker 1>And then, you know, yeah, maybe, you know,
you, you want to ask some questions, the

234
00:13:13.229 --> 00:13:16.239
<v Speaker 1>documentation or something to decide
whether to go and do, um.

235
00:13:17.179 --> 00:13:19.580
<v Speaker 1>You know, um, an, an investigation or what
have you.

236
00:13:19.739 --> 00:13:23.099
<v Speaker 1>When you go and do an investigation, you
take some photos or something like this.

237
00:13:23.340 --> 00:13:25.619
<v Speaker 1>You can imagine uploading those into the
system.

238
00:13:25.940 --> 00:13:28.780
<v Speaker 1>The system then takes into account what
it's seen, right?

239
00:13:29.140 --> 00:13:32.539
<v Speaker 1>Um, you know, goes off and, and just
generally does that in a more sophisticated

240
00:13:32.539 --> 00:13:33.159
<v Speaker 1>way.

241
00:13:33.539 --> 00:13:33.890
<v Speaker 1>OK?

242
00:13:34.140 --> 00:13:38.059
<v Speaker 1>So it's sort of like debugging problems,
um, kind of thing.

243
00:13:38.479 --> 00:13:41.719
<v Speaker 1>And, but, but you, you can also imagine a
little bit, um, more broadly.

244
00:13:42.190 --> 00:13:48.500
<v Speaker 1>Um, you, you can give these systems, um,
yeah, very general and sort of user, um,

245
00:13:48.630 --> 00:13:52.830
<v Speaker 1>you know, very, very general tasks coming
from, from the users, um, needs or

246
00:13:52.830 --> 00:13:55.909
<v Speaker 1>interests that we've, we've, you know,
not, not built into the app in terms of

247
00:13:55.909 --> 00:13:56.440
<v Speaker 1>structures, right?

248
00:13:56.590 --> 00:14:00.229
<v Speaker 1>So, so one of those examples would be, is
like, um, optimizing your maintenance

249
00:14:00.229 --> 00:14:01.309
<v Speaker 1>strategy, right?

250
00:14:01.549 --> 00:14:06.059
<v Speaker 1>You know, maybe you're monitoring 1000
robots, you have all these maintenance events

251
00:14:06.059 --> 00:14:09.140
<v Speaker 1>in the sense I app, all this data, all
these things happening.

252
00:14:09.429 --> 00:14:10.909
<v Speaker 1>There's just a lot of stuff, right?

253
00:14:11.429 --> 00:14:15.150
<v Speaker 1>You know, but you, you can have, uh, an
agentic system, you can give it that general

254
00:14:15.150 --> 00:14:18.049
<v Speaker 1>task, you can give it access to all of
that information.

255
00:14:18.500 --> 00:14:23.510
<v Speaker 1>Yeah, and you can just have it, browse
around, think about it, reason, you know,

256
00:14:23.630 --> 00:14:29.109
<v Speaker 1>come up with some hypotheses about how
you, you might improve things, go and follow

257
00:14:29.109 --> 00:14:33.140
<v Speaker 1>up on those hypotheses, da, da, da, da,
da, and in the background, you know, can

258
00:14:33.140 --> 00:14:34.669
<v Speaker 1>compile you some report, right?

259
00:14:35.099 --> 00:14:38.820
<v Speaker 1>So that, that would be an example of
things where that's in some sense, not the core

260
00:14:38.820 --> 00:14:42.349
<v Speaker 1>purpose of the app, but the application
has all this information there.

261
00:14:42.609 --> 00:14:47.210
<v Speaker 1>And if you, you think, oh, you know, I
wonder how I could schedule my oil checks on

262
00:14:47.210 --> 00:14:50.429
<v Speaker 1>my robots a little bit better, is there
anything you could help me with that?

263
00:14:51.039 --> 00:14:55.320
<v Speaker 1>Kick off for an hour using all our data
and come back with some, you know, very,

264
00:14:55.409 --> 00:14:59.130
<v Speaker 1>very detailed, sophisticated analysis, you
know, ultimately of.

265
00:14:59.700 --> 00:15:00.890
<v Speaker 1>How, how to address that, right?

266
00:15:01.500 --> 00:15:05.640
<v Speaker 1>And, further down the line, um, there are,
there are things that you can do.

267
00:15:05.750 --> 00:15:10.070
<v Speaker 1>So, for example, you can imagine using
this kind of systems to help, um, configure

268
00:15:10.070 --> 00:15:11.510
<v Speaker 1>the application, right?

269
00:15:12.109 --> 00:15:17.840
<v Speaker 1>You know, so, you know, you just got some
robots, um, with some sensor on it and you

270
00:15:17.840 --> 00:15:20.049
<v Speaker 1>just want to set it up and get some good
results.

271
00:15:20.429 --> 00:15:22.929
<v Speaker 1>And, you know, maybe rather than talk to
somebody at Sensa.

272
00:15:23.440 --> 00:15:27.159
<v Speaker 1>You know, you, you can have the system
that, again, go off and, and dig into this,

273
00:15:27.239 --> 00:15:27.320
<v Speaker 1>right?

274
00:15:27.400 --> 00:15:31.510
<v Speaker 1>You can understand the analytics, have the
ability to configure settings and things,

275
00:15:31.840 --> 00:15:35.080
<v Speaker 1>look at historical data, look at your
maintenance records and say, OK, I think this

276
00:15:35.080 --> 00:15:36.559
<v Speaker 1>is a good setup, right?

277
00:15:36.669 --> 00:15:37.619
<v Speaker 1>What do you think of this?

278
00:15:38.080 --> 00:15:41.200
<v Speaker 1>And then actually apply those things, um,
in the app.

279
00:15:41.679 --> 00:15:42.039
<v Speaker 1>Yeah.

280
00:15:42.489 --> 00:15:46.880
<v Speaker 1>And also down the line, you know, once
people have, you know, more trust in these

281
00:15:46.880 --> 00:15:51.200
<v Speaker 1>systems, yeah, you can imagine them doing
more like actual kind of like actions to

282
00:15:51.200 --> 00:15:52.340
<v Speaker 1>do with the physical world, right?

283
00:15:52.479 --> 00:15:53.039
<v Speaker 1>So, um.

284
00:15:53.609 --> 00:15:57.890
<v Speaker 1>You know, um, raising, um, maintenance
orders, right?

285
00:15:58.210 --> 00:16:01.049
<v Speaker 1>Filling them in with the suggestions, and
you know, sending them into an external

286
00:16:01.049 --> 00:16:04.809
<v Speaker 1>system, um, you know, that, that, that
kind of thing, right?

287
00:16:05.640 --> 00:16:10.950
<v Niall Sullivan>Yeah, but I, I guess trust is, um, maybe,
yeah, even, you know, putting agentic AI

288
00:16:11.250 --> 00:16:15.200
<v Niall Sullivan>to the side, you know, even with a
predictive maintenance trust is at the forefront.

289
00:16:15.289 --> 00:16:18.960
<v Niall Sullivan>Do you think, do you think it'll be
harder, harder to build trust in sort of agentic

290
00:16:18.960 --> 00:16:23.020
<v Niall Sullivan>AI agents, or does it just become part of
the natural workflow like generative AI,

291
00:16:23.609 --> 00:16:26.270
<v Niall Sullivan>uh, is now, as it were, in all forms of
life?

292
00:16:28.340 --> 00:16:31.090
<v Speaker 1>Yeah, I, I think all of the same things
apply, right?

293
00:16:31.429 --> 00:16:36.150
<v Speaker 1>And, you know, there's a transition, I
think people make in how they think about the

294
00:16:36.150 --> 00:16:39.070
<v Speaker 1>interactions with the computer when
they're interacting with, um.

295
00:16:39.820 --> 00:16:43.869
<v Speaker 1>You know, a generative AI based system
whether that's a traditional one or a, uh, a,

296
00:16:43.890 --> 00:16:45.059
<v Speaker 1>a genetic one or what have you.

297
00:16:45.419 --> 00:16:49.059
<v Speaker 1>Um, and I think that is thinking it is a
bit more like, um, again, like a colleague,

298
00:16:49.169 --> 00:16:49.690
<v Speaker 1>right?

299
00:16:50.020 --> 00:16:53.440
<v Speaker 1>It's just some system with all these kind
of powers and abilities and knowledge,

300
00:16:53.619 --> 00:16:53.969
<v Speaker 1>right?

301
00:16:54.219 --> 00:16:56.109
<v Speaker 1>But it's not a fool foolproof thing.

302
00:16:56.619 --> 00:16:57.080
<v Speaker 1>Yeah.

303
00:16:57.479 --> 00:16:59.659
<v Speaker 1>Um, and yeah, just like, um.

304
00:17:00.049 --> 00:17:05.449
<v Speaker 1>Your, your human colleagues or what have
you, you have some issue that you, you want

305
00:17:05.449 --> 00:17:08.280
<v Speaker 1>to solve, you might go to someone who you
think knows about that.

306
00:17:08.599 --> 00:17:12.099
<v Speaker 1>You listen to what they say and you put
some value on it, right?

307
00:17:12.199 --> 00:17:14.439
<v Speaker 1>But you don't treat it as, as gospel,
right?

308
00:17:14.810 --> 00:17:18.760
<v Speaker 1>You know, even experts in any area make
mistakes and what have you, right?

309
00:17:19.130 --> 00:17:22.810
<v Speaker 1>But, you know, you, you, you go to
somebody who, who has a, has a relevant

310
00:17:22.810 --> 00:17:26.530
<v Speaker 1>knowledge, you listen, you understand, and
you, you choose how to act based, based

311
00:17:26.530 --> 00:17:27.189
<v Speaker 1>on that.

312
00:17:27.890 --> 00:17:31.729
<v Speaker 1>And I think, you know, that's a sensible
way to, to interact with any generative AI

313
00:17:31.729 --> 00:17:36.170
<v Speaker 1>system, you know, um, chat GPT or all the
way to the things that we have in our, on

314
00:17:36.170 --> 00:17:37.359
<v Speaker 1>our app, right?

315
00:17:37.540 --> 00:17:42.969
<v Speaker 1>It's fundamentally an advisory system, um,
but one with great, great power and, you

316
00:17:42.969 --> 00:17:44.150
<v Speaker 1>know, intelligence.

317
00:17:44.760 --> 00:17:49.050
<v Speaker 1>And basically, like at the moment, you
know, you, you make some suggestions, you

318
00:17:49.050 --> 00:17:53.150
<v Speaker 1>very clearly say this is just to support
you or to, to help you, and then you give

319
00:17:53.489 --> 00:17:57.130
<v Speaker 1>reference information to support that and
you let the user understand to make that

320
00:17:57.130 --> 00:17:57.329
<v Speaker 1>decision.

321
00:17:57.819 --> 00:18:02.099
<v Speaker 1>And I think you would just follow a curve
over time, um, when these systems just

322
00:18:02.099 --> 00:18:03.900
<v Speaker 1>become naturally better, right?

323
00:18:03.979 --> 00:18:09.130
<v Speaker 1>You have better models and implemented in
better ways, people learn more about them,

324
00:18:09.589 --> 00:18:13.000
<v Speaker 1>and more comfortable with them and just
offload more and more onto them.

325
00:18:13.420 --> 00:18:13.979
<v Speaker 1>OK?

326
00:18:14.300 --> 00:18:17.380
<v Speaker 1>Um, and it's, it's just a curve that I
think would just be followed every time all

327
00:18:17.380 --> 00:18:18.040
<v Speaker 1>over the place.

328
00:18:19.089 --> 00:18:20.589
<v Speaker 1>Um, but it's a smooth curve, right?

329
00:18:20.660 --> 00:18:24.910
<v Speaker 1>There's no point where these systems in a
sense, become, you know, completely

330
00:18:24.910 --> 00:18:25.680
<v Speaker 1>reliable, right?

331
00:18:25.829 --> 00:18:26.290
<v Speaker 1>So.

332
00:18:27.130 --> 00:18:30.270
<v Speaker 1>You know what I mean, you should go from
ignoring it to doing exactly what it says,

333
00:18:30.349 --> 00:18:30.680
<v Speaker 1>right?

334
00:18:30.849 --> 00:18:33.180
<v Speaker 1>It's just a, it's just a smooth
transition.

335
00:18:33.650 --> 00:18:36.890
<v Niall Sullivan>No, no, exactly, exactly, and there's back
and forths and there's successes and

336
00:18:36.890 --> 00:18:39.479
<v Niall Sullivan>things that don't rub below, so nothing's
perfect, right?

337
00:18:39.569 --> 00:18:41.339
<v Niall Sullivan>Um, and it's interesting what you say.

338
00:18:41.410 --> 00:18:45.000
<v Niall Sullivan>It's something we emphasize that sense,
isn't it, but, you know, about, you know,

339
00:18:45.569 --> 00:18:49.089
<v Niall Sullivan>decision support, you know, it's about,
you know, pointing people to a problem that,

340
00:18:49.170 --> 00:18:51.449
<v Niall Sullivan>that the human still needs to investigate.

341
00:18:51.969 --> 00:18:55.209
<v Niall Sullivan>Um, but it's all, yeah, it's interesting
because we've, we've talked about this, you

342
00:18:55.209 --> 00:18:59.849
<v Niall Sullivan>know, um, since I was like a, I've heard
it before, you know, like a re reliability,

343
00:18:59.930 --> 00:19:02.469
<v Niall Sullivan>a virtual reliability engineer on your
team.

344
00:19:03.170 --> 00:19:07.849
<v Niall Sullivan>So, um, and so we're certainly talk about
today, this is to help augment a team

345
00:19:07.849 --> 00:19:09.750
<v Niall Sullivan>because obviously that's that, well.

346
00:19:10.489 --> 00:19:14.650
<v Niall Sullivan>Um, said line about AI taking jobs and all
this kind of thing, but we really are

347
00:19:14.650 --> 00:19:19.209
<v Niall Sullivan>just talking about making, uh, you know,
making, um, maintenance seems more

348
00:19:19.209 --> 00:19:20.329
<v Niall Sullivan>efficient, basically.

349
00:19:22.369 --> 00:19:23.000
<v Speaker 1>Yeah, I think so.

350
00:19:23.010 --> 00:19:24.670
<v Speaker 1>I mean, that's been our experience.

351
00:19:24.770 --> 00:19:29.760
<v Speaker 1>I, I think with, with Sensai, you know,
that maintenance teams have adopted it and

352
00:19:29.760 --> 00:19:32.520
<v Speaker 1>have found themselves being able to do
more, right?

353
00:19:32.849 --> 00:19:38.489
<v Speaker 1>To monitor more machines, uh, monitor
machines more, more precisely and to repair

354
00:19:38.489 --> 00:19:40.949
<v Speaker 1>machines at the, at better times, right?

355
00:19:41.329 --> 00:19:45.520
<v Speaker 1>And ultimately to, you know, sort of save
money, um, through, you know, less

356
00:19:45.520 --> 00:19:48.189
<v Speaker 1>unplanned downtime and, and things like
that, right?

357
00:19:48.469 --> 00:19:50.130
<v Speaker 1>So I often think about it like that.

358
00:19:50.209 --> 00:19:50.949
<v Speaker 1>You have a team.

359
00:19:51.510 --> 00:19:55.599
<v Speaker 1>You bring in Sensi, it supplements them,
um, and it helps to make a bigger

360
00:19:55.599 --> 00:19:58.069
<v Speaker 1>contribution to the, the organization,
right?

361
00:19:58.180 --> 00:20:01.209
<v Speaker 1>Rather than, you know, replacing people.

362
00:20:01.469 --> 00:20:01.910
<v Speaker 1>OK?

363
00:20:02.150 --> 00:20:08.239
<v Speaker 1>Um, and I, I think that, yeah, AI in, in
general, you know, these systems we're,

364
00:20:08.270 --> 00:20:11.250
<v Speaker 1>we're talking about today in the short
term, I think they.

365
00:20:12.089 --> 00:20:15.089
<v Speaker 1>They they just follow, follow this path,
right?

366
00:20:15.170 --> 00:20:15.640
<v Speaker 1>OK.

367
00:20:15.890 --> 00:20:19.250
<v Speaker 1>It's just, you know, more sophisticated
colleagues, right?

368
00:20:19.650 --> 00:20:23.949
<v Speaker 1>Um, and it's, it's interesting with, with
maintenance, you know, maintenance people,

369
00:20:24.209 --> 00:20:29.329
<v Speaker 1>yeah, have a lot of, um, like on the
ground, uh, knowledge of what's happening,

370
00:20:29.410 --> 00:20:29.569
<v Speaker 1>right?

371
00:20:29.680 --> 00:20:30.770
<v Speaker 1>They're in the factory.

372
00:20:31.170 --> 00:20:35.530
<v Speaker 1>They know the actual machines, they can
hear them, you know, and they're the people

373
00:20:35.530 --> 00:20:39.910
<v Speaker 1>who are actually going to those machines
and, you know, maybe removing components or

374
00:20:40.150 --> 00:20:41.689
<v Speaker 1>all, all of this kind of stuff, you know.

375
00:20:41.989 --> 00:20:45.510
<v Speaker 1>I mean, there's big chunks of the jobs of
maintenance engineers, which are, are not

376
00:20:45.510 --> 00:20:49.949
<v Speaker 1>going to be replaced by AI anytime soon,
right?

377
00:20:50.390 --> 00:20:55.189
<v Speaker 1>Um, you know, it's the jobs of knowledge
workers like, like me and you, I think are,

378
00:20:55.420 --> 00:20:58.530
<v Speaker 1>are, are more at risk ultimately than,
than maintenance engineers.

379
00:20:59.069 --> 00:21:03.270
<v Speaker 1>But, but, but AI for, you know, in a, in a
maintenance shop, you know, it, it has

380
00:21:03.270 --> 00:21:07.550
<v Speaker 1>the ability and sort of improve
productivity a lot and help these people to make

381
00:21:07.550 --> 00:21:10.430
<v Speaker 1>much more of a, much more of a
contribution, right?

382
00:21:10.910 --> 00:21:17.290
<v Speaker 1>And then down the line, you know,
obviously AI is going to cause Large-scale changes

383
00:21:17.290 --> 00:21:19.640
<v Speaker 1>to employment and society and things like
that.

384
00:21:20.130 --> 00:21:24.930
<v Speaker 1>Um, but that is a slightly longer-term
thing and, um, yeah, maintenance engineers

385
00:21:24.930 --> 00:21:26.989
<v Speaker 1>may be more protected from that than most
people.

386
00:21:28.270 --> 00:21:30.599
<v Niall Sullivan>Yeah, no, it, it makes a lot of sense.

387
00:21:30.640 --> 00:21:35.109
<v Niall Sullivan>There's a lot of skill, um, yeah, a lot of
skills you can't simply replace, right?

388
00:21:35.199 --> 00:21:36.800
<v Niall Sullivan>So that's what we're talking about here.

389
00:21:36.920 --> 00:21:37.500
<v Niall Sullivan>I guess.

390
00:21:38.000 --> 00:21:42.260
<v Niall Sullivan>Another thing with AI agents, we talked
about it in a predictive maintenance

391
00:21:42.260 --> 00:21:47.750
<v Niall Sullivan>context, but of course, you know, as the
technology develops within, um, say, um,

392
00:21:48.040 --> 00:21:52.119
<v Niall Sullivan>you know, within our customers and other
organizations, they'll be developing their

393
00:21:52.119 --> 00:21:56.270
<v Niall Sullivan>own, um, agentic AI, um, capabilities.

394
00:21:56.520 --> 00:21:57.640
<v Niall Sullivan>So, again, it.

395
00:21:58.020 --> 00:22:02.790
<v Niall Sullivan>And at that point it comes about how our
AI agents talk to their AI agents.

396
00:22:02.910 --> 00:22:06.290
<v Niall Sullivan>So what, what sort of the challenges
around that, James, or opportunities as well,

397
00:22:06.510 --> 00:22:07.089
<v Niall Sullivan>of course.

398
00:22:08.290 --> 00:22:11.260
<v Speaker 1>Yeah, I mean, I've, I've presented agents
so far as being something that's to do

399
00:22:11.260 --> 00:22:12.209
<v Speaker 1>with depth, right?

400
00:22:12.579 --> 00:22:16.209
<v Speaker 1>So the ability of, um, a self-contained
system to, you know, to go off, use tools,

401
00:22:16.260 --> 00:22:20.000
<v Speaker 1>dig into things and accomplish these sort
of broad tasks.

402
00:22:20.530 --> 00:22:25.140
<v Speaker 1>Um, but you, there are other, other ways
of thinking about it and I think which one,

403
00:22:25.219 --> 00:22:30.140
<v Speaker 1>which is quite salient to people, um, is,
you know, yeah, as you said, thinking

404
00:22:30.140 --> 00:22:31.699
<v Speaker 1>about the interactions of agents, right?

405
00:22:31.819 --> 00:22:36.420
<v Speaker 1>So when, when your agent is doing its
stuff and it's like calling a tool to.

406
00:22:37.170 --> 00:22:41.660
<v Speaker 1>Um, look up some information or something,
it can just as easily, you know, send,

407
00:22:41.699 --> 00:22:47.939
<v Speaker 1>um, a natural language query to an agent
on somebody else's system, you know, asking

408
00:22:47.939 --> 00:22:52.849
<v Speaker 1>that agent if it has anything to, to add
to, to, to, to, to, to the, to the thing at

409
00:22:52.849 --> 00:22:53.500
<v Speaker 1>hand, right?

410
00:22:53.780 --> 00:22:58.170
<v Speaker 1>You know, so you can imagine, you know,
maybe that there is some system which is

411
00:22:58.170 --> 00:23:02.130
<v Speaker 1>handling, um, or modeling your, your
processes or helping control the, control the

412
00:23:02.130 --> 00:23:03.839
<v Speaker 1>processes that are going on in your
machines.

413
00:23:04.959 --> 00:23:08.939
<v Speaker 1>And you have Sensi, which is thinking
about maintenance, but both of these systems

414
00:23:08.939 --> 00:23:11.160
<v Speaker 1>are ultimately to do with machines, right?

415
00:23:11.880 --> 00:23:15.839
<v Speaker 1>Um, and you can think of agents which are
connected to both of those systems, being

416
00:23:15.839 --> 00:23:20.719
<v Speaker 1>able to do interesting tasks, looking up
information from their own data sources and

417
00:23:20.719 --> 00:23:21.500
<v Speaker 1>so forth.

418
00:23:21.829 --> 00:23:25.359
<v Speaker 1>Um, and you can imagine them hooking into
each other's data sources in traditional

419
00:23:25.359 --> 00:23:27.729
<v Speaker 1>ways like by calling that APIs, right?

420
00:23:28.160 --> 00:23:34.079
<v Speaker 1>Um, but an agent to agent interaction,
yeah, it's potentially, um, yeah, very cool

421
00:23:34.079 --> 00:23:38.099
<v Speaker 1>and, um, much more flexible than like this
kind of API things, right?

422
00:23:38.540 --> 00:23:39.719
<v Speaker 1>Um, so yeah.

423
00:23:40.250 --> 00:23:43.729
<v Speaker 1>The Sensai agent can completely understand
the Sensai product and everything that's

424
00:23:43.729 --> 00:23:46.959
<v Speaker 1>in it that's connected up, like where's
your process one?

425
00:23:47.250 --> 00:23:52.010
<v Speaker 1>And if there's a process issue, you know,
that system can, yeah, very easily talk to

426
00:23:52.010 --> 00:23:55.449
<v Speaker 1>the Sensai one in principle, you know, to
ask questions about whether there's

427
00:23:55.449 --> 00:23:59.770
<v Speaker 1>anything there that relates to this or
might help, you know, understand what's going

428
00:23:59.770 --> 00:24:02.550
<v Speaker 1>on, um, and then, and then vice versa.

429
00:24:03.010 --> 00:24:03.319
<v Speaker 1>Yeah.

430
00:24:03.689 --> 00:24:08.170
<v Speaker 1>So, so these kinds of systems, so, yeah,
instead of using tools to just like call an

431
00:24:08.170 --> 00:24:09.709
<v Speaker 1>API, you know, to, to do it.

432
00:24:10.280 --> 00:24:16.000
<v Speaker 1>Computational task, you have tools which
are like, you know, um, sending messages in

433
00:24:16.000 --> 00:24:19.359
<v Speaker 1>a conversational way between different
agents, right?

434
00:24:19.680 --> 00:24:25.319
<v Speaker 1>So you can set up these systems in kind of
toy ways, uh, quite easily and there's

435
00:24:25.319 --> 00:24:28.229
<v Speaker 1>certainly, uh, some kind of extreme magic
to it, right?

436
00:24:28.520 --> 00:24:32.119
<v Speaker 1>Which I think is why it's quite salient to
people, you know, to have two computer

437
00:24:32.119 --> 00:24:36.760
<v Speaker 1>systems that chatting with each other,
trying to work out what you want as the user

438
00:24:36.760 --> 00:24:37.920
<v Speaker 1>and how they can help you.

439
00:24:38.300 --> 00:24:41.260
<v Speaker 1>Um, uh, yeah, satisfy that one.

440
00:24:41.329 --> 00:24:44.020
<v Speaker 1>It's very, it's very, very, uh, cool,
right?

441
00:24:45.310 --> 00:24:51.010
<v Speaker 1>Um, OK, so, yeah, again, this is a thing
where this, this technology sort of exists,

442
00:24:51.089 --> 00:24:51.369
<v Speaker 1>right?

443
00:24:51.410 --> 00:24:55.569
<v Speaker 1>Um, and there are nice, um, tech technical
ways for handling this.

444
00:24:55.650 --> 00:24:58.189
<v Speaker 1>There are nice, um, protocols, right?

445
00:24:58.770 --> 00:25:05.329
<v Speaker 1>Um, which, um, you know, help, for
example, the Sensite agent understand in this

446
00:25:05.329 --> 00:25:10.050
<v Speaker 1>world that it operates, what other tools
or agents are available to it that it can,

447
00:25:10.170 --> 00:25:12.329
<v Speaker 1>can talk to and how to interact with them.

448
00:25:13.209 --> 00:25:15.000
<v Speaker 1>Um, yeah.

449
00:25:15.500 --> 00:25:19.260
<v Speaker 1>OK, so this is, um, it's just, it's just a
cool thing, right?

450
00:25:19.339 --> 00:25:23.640
<v Speaker 1>And, uh, but if you think about it in, in,
in, in real products, right?

451
00:25:24.020 --> 00:25:28.219
<v Speaker 1>Um, what you, you have to do to, to, to
make this work is, I mean, firstly, you just

452
00:25:28.219 --> 00:25:29.900
<v Speaker 1>need to have good, good agents, right?

453
00:25:30.099 --> 00:25:35.099
<v Speaker 1>So ones that can do these long sequences
of tasks and get lots of information and

454
00:25:35.099 --> 00:25:40.420
<v Speaker 1>not, not be confused, then you have to
have products that, um, while working

455
00:25:40.420 --> 00:25:41.630
<v Speaker 1>together really makes sense.

456
00:25:42.040 --> 00:25:44.630
<v Speaker 1>You know, whether they, where they really
have, um.

457
00:25:45.729 --> 00:25:47.160
<v Speaker 1>Complementary information.

458
00:25:48.069 --> 00:25:51.660
<v Speaker 1>And you also want systems where it makes
sense to have this very general like

459
00:25:51.660 --> 00:25:53.670
<v Speaker 1>natural language conversation, right?

460
00:25:54.369 --> 00:25:59.800
<v Speaker 1>OK, um, and once you have that, that sort
of setup, and you, you can, you can create

461
00:25:59.800 --> 00:26:06.430
<v Speaker 1>these things, um, in terms of, Tech
technical issues, um, and understanding the

462
00:26:06.430 --> 00:26:10.260
<v Speaker 1>relationship between entities in these
different systems is, is important.

463
00:26:10.760 --> 00:26:11.300
<v Speaker 1>OK?

464
00:26:11.680 --> 00:26:17.400
<v Speaker 1>Um, so, you know, how, how does Sensi,
which is working on a particular asset,

465
00:26:17.959 --> 00:26:21.880
<v Speaker 1>thinking about a particular asset with a
particular ID communicate to some external

466
00:26:21.880 --> 00:26:24.719
<v Speaker 1>system what exactly, what assay is talking
about, right?

467
00:26:25.040 --> 00:26:28.479
<v Speaker 1>That system might have other internal
representations, OK?

468
00:26:28.729 --> 00:26:32.920
<v Speaker 1>So you have these different systems and
you tell them to know what assets are what,

469
00:26:33.150 --> 00:26:34.140
<v Speaker 1>right, so that they can.

470
00:26:34.939 --> 00:26:37.650
<v Speaker 1>Communicate to each other effectively what
they're talking about.

471
00:26:38.099 --> 00:26:41.699
<v Speaker 1>And then you also have, you know, issues
to, to do with security, you know,

472
00:26:41.780 --> 00:26:44.930
<v Speaker 1>authentication and data access and so
forth, right?

473
00:26:45.260 --> 00:26:48.800
<v Speaker 1>So that, you know, one system firing off
questions to the other system.

474
00:26:49.439 --> 00:26:52.739
<v Speaker 1>You know, it needs one way or another to
communicate.

475
00:26:53.780 --> 00:26:57.979
<v Speaker 1>Who its user is, what that user is allowed
to know, and that authentication is

476
00:26:57.979 --> 00:27:01.619
<v Speaker 1>passed to this external system, so that
external system knows what it is allowed to

477
00:27:01.619 --> 00:27:02.819
<v Speaker 1>reply, right?

478
00:27:03.180 --> 00:27:05.979
<v Speaker 1>OK, so to prevent data leakage and, and so
forth.

479
00:27:06.599 --> 00:27:07.479
<v Speaker 1>But yeah, it's one of those things.

480
00:27:07.520 --> 00:27:11.939
<v Speaker 1>It's just a question of finding use cases
that, that really make sense.

481
00:27:12.280 --> 00:27:18.280
<v Speaker 1>Yeah, um, implementing things sensibly
with, with good models and making sure that

482
00:27:18.280 --> 00:27:24.060
<v Speaker 1>you handle these, um, authentication and,
you know, entity matching sort of problems

483
00:27:24.400 --> 00:27:25.699
<v Speaker 1>in, in sensible ways.

484
00:27:26.660 --> 00:27:27.060
<v Speaker 1>Right.

485
00:27:27.420 --> 00:27:29.500
<v Speaker 1>And this is just really being worked out
at the moment, right?

486
00:27:29.579 --> 00:27:33.760
<v Speaker 1>The best way, particularly those, those
last items I mentioned, that the best way to

487
00:27:34.219 --> 00:27:35.290
<v Speaker 1>approach those problems.

488
00:27:35.859 --> 00:27:39.900
<v Speaker 1>But if you look further down the road,
yeah, you, you can very easily imagine, um, a

489
00:27:39.900 --> 00:27:42.589
<v Speaker 1>customer having, you know, their AI,
right?

490
00:27:42.660 --> 00:27:47.140
<v Speaker 1>And it is like the, the entry point to
that system, and that's what users experience

491
00:27:47.140 --> 00:27:49.180
<v Speaker 1>in the first case, right?

492
00:27:49.260 --> 00:27:53.369
<v Speaker 1>And then Sensi becomes some sort of system
that sits behind that, that it can

493
00:27:53.369 --> 00:27:54.819
<v Speaker 1>communicate with, right?

494
00:27:55.420 --> 00:27:58.949
<v Speaker 1>OK, and can, you know, yeah, maybe.

495
00:27:59.959 --> 00:28:03.410
<v Speaker 1>Do some back and forth about what's going
on or that central system can hand you off

496
00:28:03.410 --> 00:28:08.829
<v Speaker 1>to Sensa right directly and then
eventually take you into the Sensai application

497
00:28:08.829 --> 00:28:09.650
<v Speaker 1>this kind of thing,

498
00:28:09.890 --> 00:28:13.829
<v Niall Sullivan>yeah, um, yeah, that makes, that makes a
lot of sense as well.

499
00:28:14.420 --> 00:28:18.420
<v Niall Sullivan>Um, I guess another aspect you mentioned
it, I mean just briefly then, James, is

500
00:28:18.420 --> 00:28:21.819
<v Niall Sullivan>around sort of data and how, how does that
fit in?

501
00:28:21.890 --> 00:28:25.500
<v Niall Sullivan>And we know it's a key component of
successful predictive maintenance deployments,

502
00:28:25.560 --> 00:28:30.660
<v Niall Sullivan>but in this context of agentic AI, you
know, how does the data we collect to enable

503
00:28:30.660 --> 00:28:34.180
<v Niall Sullivan>predictive maintenance, how does that fit
together with, with this new technology?

504
00:28:35.760 --> 00:28:38.260
<v Speaker 1>Yeah, I, I think you just have to be able
to access the data.

505
00:28:38.319 --> 00:28:41.959
<v Speaker 1>You have to have relevant information and
in a sense that this is all this time

506
00:28:41.959 --> 00:28:47.239
<v Speaker 1>series data that we collect, all the
observations we make from that, the traditional

507
00:28:47.239 --> 00:28:53.109
<v Speaker 1>machine learning, um, maintenance records
and supplementary information like

508
00:28:53.109 --> 00:28:56.510
<v Speaker 1>manuals, user interactions, all this kind
of stuff, right?

509
00:28:56.760 --> 00:29:00.579
<v Speaker 1>You know, um, so you, you have to have
some, some useful body.

510
00:29:01.369 --> 00:29:06.770
<v Speaker 1>Uh, of data that the, these systems can
use obviously to, to do something useful

511
00:29:06.770 --> 00:29:08.030
<v Speaker 1>along with their general knowledge.

512
00:29:08.750 --> 00:29:13.150
<v Speaker 1>Um, and then you, you just need to
organize that in a way in which these systems can

513
00:29:13.150 --> 00:29:13.709
<v Speaker 1>consume.

514
00:29:14.030 --> 00:29:20.630
<v Speaker 1>And when you're thinking about like, um,
a, a traditional chatbot, very often you

515
00:29:20.630 --> 00:29:24.069
<v Speaker 1>would think of just collecting some of
that information and just kind of putting it

516
00:29:24.069 --> 00:29:25.390
<v Speaker 1>in the chatbot in the background.

517
00:29:25.510 --> 00:29:28.439
<v Speaker 1>Then when you talk to it, you'd just be
talking about what it knows.

518
00:29:28.869 --> 00:29:34.619
<v Speaker 1>And an agentic system, you'll typically
want to, um, put that data behind, uh,

519
00:29:34.630 --> 00:29:35.869
<v Speaker 1>tools, right?

520
00:29:36.290 --> 00:29:41.859
<v Speaker 1>So for example, if the system wants to
know about work history on a particular asset

521
00:29:41.859 --> 00:29:45.930
<v Speaker 1>or what alerts have been raised or it
wants to do a lookup in some documents, they

522
00:29:45.930 --> 00:29:51.449
<v Speaker 1>need to be represented as tools that it
can use, you know, and specify what it wants

523
00:29:51.689 --> 00:29:53.209
<v Speaker 1>and then get some result back.

524
00:29:53.979 --> 00:29:57.079
<v Speaker 1>In a, in a form that it can understand,
you know.

525
00:29:57.579 --> 00:30:01.729
<v Speaker 1>So, I mean, I, there's some work that has
been going on in Sens I really since the

526
00:30:01.729 --> 00:30:07.010
<v Speaker 1>beginning of, um, this generative AI
period, we've always had the right kind of

527
00:30:07.010 --> 00:30:07.800
<v Speaker 1>data, yeah.

528
00:30:08.339 --> 00:30:13.260
<v Speaker 1>What we've been doing, you know, in, in
that thread is, is, is organizing this into,

529
00:30:13.339 --> 00:30:17.739
<v Speaker 1>into ways which, um, get more compatible
with language models that they can

530
00:30:17.739 --> 00:30:18.949
<v Speaker 1>understand and reason over.

531
00:30:19.459 --> 00:30:22.219
<v Speaker 1>Yeah, and, and need to access using tools,
yeah.

532
00:30:23.770 --> 00:30:27.130
<v Niall Sullivan>So that, so that's, but yeah, and that's
basically the, the continued journey, isn't

533
00:30:27.130 --> 00:30:27.239
<v Niall Sullivan>it?

534
00:30:27.319 --> 00:30:33.099
<v Niall Sullivan>So you're now applying that to now these
uh to agentic AI um technology, right?

535
00:30:33.290 --> 00:30:35.349
<v Niall Sullivan>That's extending it out to that.

536
00:30:36.359 --> 00:30:40.729
<v Speaker 1>Yeah, it's building at this for its
repertoire of tools, yeah, and then getting

537
00:30:40.729 --> 00:30:43.619
<v Speaker 1>involved with the right models, with the
right tasks.

538
00:30:44.079 --> 00:30:46.900
<v Speaker 1>Yeah, and yeah, building this sort of deep
functionality.

539
00:30:47.660 --> 00:30:50.699
<v Speaker 1>And then with the agent to agent thing
similarly, right?

540
00:30:50.819 --> 00:30:55.099
<v Speaker 1>You know, um, understanding how to, to
build these correct interfaces with, with

541
00:30:55.099 --> 00:30:59.199
<v Speaker 1>outside agents, um, that, yeah, it can be
used like tools.

542
00:31:00.109 --> 00:31:00.119
<v Speaker 1>Yeah,

543
00:31:01.089 --> 00:31:06.140
<v Niall Sullivan>yeah, and I, I guess, you know, as we
come, come towards, I guess the end, um, of

544
00:31:06.140 --> 00:31:09.449
<v Niall Sullivan>our, of our discussion, it's really, it's
been really interesting to learn more

545
00:31:09.449 --> 00:31:12.089
<v Niall Sullivan>about this, this is what I get out of a
lot of these podcasts.

546
00:31:12.130 --> 00:31:17.400
<v Niall Sullivan>I learn a lot too, um, and I know that
being in its infancy, I guess it's very

547
00:31:17.400 --> 00:31:19.569
<v Niall Sullivan>difficult to look too far ahead, James,
but.

548
00:31:19.930 --> 00:31:24.449
<v Niall Sullivan>I know that you're a visionary of sorts,
so I'm sure you have an opinion on where

549
00:31:24.449 --> 00:31:28.369
<v Niall Sullivan>you will, you see things going with
Vigenic AI and predicting maintenance over the

550
00:31:28.369 --> 00:31:33.949
<v Niall Sullivan>next few years and what, what people
should maybe expect to see in their day to day

551
00:31:34.410 --> 00:31:35.349
<v Niall Sullivan>working life, I guess.

552
00:31:37.489 --> 00:31:41.459
<v Speaker 1>Yeah, I, I, I think it's really, it's just
about, um, two things, which is just

553
00:31:41.459 --> 00:31:48.339
<v Speaker 1>much, much greater depth and autonomy, um,
and, you know, um, better integration

554
00:31:48.339 --> 00:31:52.410
<v Speaker 1>between people's predictive maintenance
system and, and their other systems that,

555
00:31:52.739 --> 00:31:54.959
<v Speaker 1>that, that have relevant information, you
know.

556
00:31:55.810 --> 00:31:58.369
<v Speaker 1>Um, I think people can.

557
00:31:58.959 --> 00:32:03.439
<v Speaker 1>Yeah, imagine, yeah, really very sort of
deep and sophisticated output from these

558
00:32:03.439 --> 00:32:08.439
<v Speaker 1>systems, you know, in terms of fixed sort
of output like explaining, oh, yeah, he's

559
00:32:08.439 --> 00:32:11.709
<v Speaker 1>a potential problem, blah blah blah blah,
and then when you're interacting with the

560
00:32:11.709 --> 00:32:17.500
<v Speaker 1>system, I think it's going to And this is
going to feel a lot more intelligent,

561
00:32:17.579 --> 00:32:20.219
<v Speaker 1>right, in sort of real time in front of
you, so.

562
00:32:20.949 --> 00:32:25.589
<v Speaker 1>I'm hoping, you know, in, in the sense
like, you'll be able to sort of see this

563
00:32:25.589 --> 00:32:26.550
<v Speaker 1>going on, right?

564
00:32:26.670 --> 00:32:30.709
<v Speaker 1>You've asked the system something and
you'll be able to see it very quickly pulling

565
00:32:30.709 --> 00:32:36.339
<v Speaker 1>in all the different data, um, you know,
doing analysis of, um, different kinds,

566
00:32:36.550 --> 00:32:39.150
<v Speaker 1>cross-referencing, forming hypothesis,
right?

567
00:32:39.510 --> 00:32:44.530
<v Speaker 1>And then, yeah, coming back to you with
its, um, with it, with its output, right?

568
00:32:44.989 --> 00:32:49.109
<v Speaker 1>Um, yeah, so, so it should just be, just,
it's just deeper and more sophisticated

569
00:32:49.109 --> 00:32:49.569
<v Speaker 1>like that.

570
00:32:49.869 --> 00:32:53.189
<v Speaker 1>And over the kind of time period that
you're talking about, you know, I, I can, I

571
00:32:53.189 --> 00:32:58.270
<v Speaker 1>can't imagine, um, yeah, you know,
interactions with the outside world happening,

572
00:32:58.709 --> 00:33:00.579
<v Speaker 1>um, and, and being reliable.

573
00:33:00.930 --> 00:33:04.949
<v Speaker 1>So, you know, the ability to sort of
managing, uh, managing working work orders and,

574
00:33:05.069 --> 00:33:06.660
<v Speaker 1>and, and, and things like this.

575
00:33:07.109 --> 00:33:07.270
<v Speaker 1>Yeah.

576
00:33:08.219 --> 00:33:09.770
<v Speaker 1>Uh, um, yeah.

577
00:33:11.420 --> 00:33:16.819
<v Speaker 1>That's a, that's a bit of visionary
things, right, but I, I think it, it just, you

578
00:33:16.819 --> 00:33:21.579
<v Speaker 1>know, you, you can see like, um, if you
have access to the very fancy models or in a

579
00:33:21.579 --> 00:33:25.500
<v Speaker 1>research setting, you can just see these
very magical things when systems are doing

580
00:33:25.500 --> 00:33:25.739
<v Speaker 1>this.

581
00:33:25.819 --> 00:33:30.939
<v Speaker 1>You give them the right information, right
direction, um, and the freedom to pursue

582
00:33:30.939 --> 00:33:33.979
<v Speaker 1>over these periods of time, and, um, yeah,
they really do magical things.

583
00:33:34.140 --> 00:33:36.239
<v Speaker 1>So this in my head is what, what I would
want to bring into.

584
00:33:37.439 --> 00:33:39.160
<v Speaker 1>Into our product and yeah,

585
00:33:39.680 --> 00:33:43.640
<v Niall Sullivan>yeah, so maybe we should clip that bit and
then come back to it later and see and

586
00:33:43.640 --> 00:33:46.400
<v Niall Sullivan>see if it came to, came to fruition, um.

587
00:33:47.310 --> 00:33:50.920
<v Niall Sullivan>But, um, I, I guess just one final note, I
always do this, I always say there's one

588
00:33:50.920 --> 00:33:55.040
<v Niall Sullivan>final question, then ask a another final
question, but in, in terms of you

589
00:33:55.040 --> 00:33:58.189
<v Niall Sullivan>personally, James, I know you you've been
such a big driving force behind, you know,

590
00:33:58.319 --> 00:34:00.099
<v Niall Sullivan>um, our James with AI and.

591
00:34:00.900 --> 00:34:04.670
<v Niall Sullivan>Um, and now genic AI, you know, could you
sort of put into words sort of your

592
00:34:04.670 --> 00:34:09.030
<v Niall Sullivan>excitement around this pace and its
potential, you know, and then not just within

593
00:34:09.030 --> 00:34:11.530
<v Niall Sullivan>industrial environments, but, you know,
beyond that as well.

594
00:34:13.689 --> 00:34:17.409
<v Speaker 1>Yeah, I, I just think it's just, it's just
a world-changing technology and it's been

595
00:34:17.409 --> 00:34:22.550
<v Speaker 1>very interesting to see over the past few
years, um, you know, how.

596
00:34:23.408 --> 00:34:28.878
<v Speaker 1>Yeah, people's expectations in terms of
its capabilities have just sort of been,

597
00:34:28.989 --> 00:34:31.638
<v Speaker 1>been met and, and generally exceeded,
right?

598
00:34:32.158 --> 00:34:36.858
<v Speaker 1>There's no feeling that the progress will
sort of, will slow or is slowing down.

599
00:34:37.398 --> 00:34:37.979
<v Speaker 1>OK?

600
00:34:38.458 --> 00:34:40.858
<v Speaker 1>Um, and yeah, there's some.

601
00:34:41.909 --> 00:34:46.030
<v Speaker 1>Long process in which these things get,
get pushed out into, into products and into

602
00:34:46.030 --> 00:34:47.409
<v Speaker 1>wider society, right?

603
00:34:47.510 --> 00:34:52.080
<v Speaker 1>But I, I really think, you know, this is,
um, yeah, um, an absolutely

604
00:34:52.270 --> 00:34:56.889
<v Speaker 1>world-changing, um, technological moment
that, that we're in.

605
00:34:57.330 --> 00:35:02.149
<v Speaker 1>Um, and it's just, yeah, it's sort of
exciting to sort of be part of that, to see

606
00:35:02.149 --> 00:35:05.550
<v Speaker 1>what's happening now, what's going to,
obviously going to come down the line and to

607
00:35:05.870 --> 00:35:08.229
<v Speaker 1>speculate about what's coming sort of
after that.

608
00:35:08.979 --> 00:35:11.070
<v Speaker 1>Um, and some, somehow be part of it,
right?

609
00:35:11.350 --> 00:35:13.590
<v Speaker 1>But I, I think it's generally just a very
positive thing.

610
00:35:13.669 --> 00:35:16.870
<v Speaker 1>I mean, these sort of technologies can
cause, uh, social disruptions of all

611
00:35:16.870 --> 00:35:22.219
<v Speaker 1>different kinds, um, you know, but
ultimately, these are things that should lead

612
00:35:22.219 --> 00:35:28.110
<v Speaker 1>into, um, you know, human beings as a
whole being an awful lot more efficient and

613
00:35:28.110 --> 00:35:28.969
<v Speaker 1>productive.

614
00:35:29.429 --> 00:35:33.949
<v Speaker 1>We should produce a lot of economic growth
and hopefully ends up producing, um, much

615
00:35:33.949 --> 00:35:37.600
<v Speaker 1>more affluent and, Comfortable world for
everyone, right?

616
00:35:38.409 --> 00:35:44.199
<v Speaker 1>Um, but yeah, it's just, just a very, you
know, interesting historical moment and

617
00:35:44.199 --> 00:35:44.750
<v Speaker 1>interesting to be.

618
00:35:45.739 --> 00:35:49.909
<v Speaker 1>You know, part of it in a, in a small way
using the technology in my view.

619
00:35:51.370 --> 00:35:51.830
<v Niall Sullivan>No, exactly.

620
00:35:52.110 --> 00:35:57.290
<v Niall Sullivan>No, it, no, it is very exciting, and of
course, on this podcast and elsewhere, we'll

621
00:35:57.290 --> 00:35:58.790
<v Niall Sullivan>continue to sort of follow this journey.

622
00:35:58.959 --> 00:36:03.129
<v Niall Sullivan>Um, it's amazing to see what, you know,
how far the app's come, even in my 4 years

623
00:36:03.129 --> 00:36:05.810
<v Niall Sullivan>if I say at Senign now, Siemens as well.

624
00:36:05.889 --> 00:36:07.870
<v Niall Sullivan>So we look forward to seeing what's coming
next.

625
00:36:08.050 --> 00:36:12.070
<v Niall Sullivan>But, um, but for today, James, um, it's
been really great to chat with you again.

626
00:36:12.449 --> 00:36:15.030
<v Niall Sullivan>Um, thank you for sharing your insights on
this topic.

627
00:36:15.169 --> 00:36:19.239
<v Niall Sullivan>I'm sure we'll check in again on it, um,
as it develops within, you know, the, the

628
00:36:19.239 --> 00:36:20.350
<v Niall Sullivan>context of Sensai.

629
00:36:20.989 --> 00:36:25.350
<v Niall Sullivan>But um but yeah for today um thank you
very much for joining us and thank you to

630
00:36:25.350 --> 00:36:28.979
<v Niall Sullivan>everyone listening wherever you are and
we'll see you on the next episode.

631
00:36:31.260 --> 00:36:36.199
<v Speaker 2>Please like and subscribe on all major
podcast channels and find out more about

632
00:36:36.199 --> 00:36:39.580
<v Speaker 2>Sensi predictive maintenance at
Siemens.com.
