WEBVTT
Language: en-US

1
00:00:05.369 --> 00:00:10.720
<v Nick Leeder>So, you know, when you kind of start in
that small wing approach, you know, again

2
00:00:10.720 --> 00:00:13.039
<v Nick Leeder>back to the lighthouse, um.

3
00:00:13.869 --> 00:00:20.729
<v Nick Leeder>These transformations tend to happen 25 to
50% faster when you kind of start

4
00:00:20.729 --> 00:00:21.809
<v Nick Leeder>small and scaling.

5
00:00:22.809 --> 00:00:28.489
<v Nick Leeder>Um, and that, that, that comes from
remaining focused on what you're trying to

6
00:00:28.489 --> 00:00:28.850
<v Nick Leeder>prove.

7
00:00:29.649 --> 00:00:34.250
<v Nick Leeder>And accepting that you're gonna do this in
an iterative way.

8
00:00:35.240 --> 00:00:39.080
<v Nick Leeder>Because I guess that we've heard this
phrase yet, um.

9
00:00:40.029 --> 00:00:43.689
<v Nick Leeder>So, start small, scale fast or fail fast.

10
00:00:44.430 --> 00:00:49.430
<v Nick Leeder>And I think that's really important
because resources are always going to be

11
00:00:49.430 --> 00:00:50.049
<v Nick Leeder>limited.

12
00:00:50.389 --> 00:00:53.790
<v Nick Leeder>So if it's not gonna work, you don't want
to spend 6 months trying to prove

13
00:00:53.790 --> 00:00:55.619
<v Nick Leeder>something and find it doesn't work then.

14
00:00:55.950 --> 00:01:00.029
<v Nick Leeder>You want to prove it within 2 or 2 or 3
months that it's gonna work or not.

15
00:01:00.389 --> 00:01:04.669
<v Nick Leeder>Have a plan to scale it or a plan to say
look, let's be honest with ourselves, this

16
00:01:04.669 --> 00:01:05.510
<v Nick Leeder>isn't working.

17
00:01:06.319 --> 00:01:11.529
<v Speaker 2>Welcome to the trend detection podcast
powered by Siemens, and now your host, Nel

18
00:01:11.529 --> 00:01:12.120
<v Speaker 2>Sullivan.

19
00:01:14.470 --> 00:01:19.010
<v Niall Sullivan>Hello and welcome to the latest episode of
the trend detection podcast.

20
00:01:19.510 --> 00:01:23.510
<v Niall Sullivan>Um, I'm really pleased to be joined once
again by Nick Leader, a familiar face to

21
00:01:23.510 --> 00:01:24.330
<v Niall Sullivan>the podcast now.

22
00:01:24.430 --> 00:01:25.230
<v Niall Sullivan>How are you, Nick?

23
00:01:26.110 --> 00:01:30.910
<v Nick Leeder>All good, trying to, trying to, trying to
remain cool on this really hot summery day

24
00:01:30.910 --> 00:01:31.839
<v Nick Leeder>in the UK.

25
00:01:32.459 --> 00:01:35.750
<v Niall Sullivan>It is, I don't know, I mean, most people
will be listening to this actually, but

26
00:01:35.750 --> 00:01:39.230
<v Niall Sullivan>yeah, if you happen to see us, we, we
might slowly melt through this podcast as we,

27
00:01:39.349 --> 00:01:42.910
<v Niall Sullivan>as we get through, but we'll, we'll try,
try and survive the best we can.

28
00:01:43.319 --> 00:01:49.169
<v Niall Sullivan>Um, but today, yeah, today we're talking
about, um, a, a not unfamiliar topic, AI,

29
00:01:49.449 --> 00:01:53.029
<v Niall Sullivan>but hopefully we're gonna have a new sort
of angle on it, cos we're talking about AI

30
00:01:53.029 --> 00:01:57.440
<v Niall Sullivan>beyond the hype, which I think is a very
challenging topic actually, Nick, cos it's

31
00:01:57.440 --> 00:02:00.069
<v Niall Sullivan>impossible not to hype AI, I think at the
moment.

32
00:02:00.730 --> 00:02:06.250
<v Niall Sullivan>Um, but, um, but yeah, so I guess let's
dive straight in really around this topic.

33
00:02:06.330 --> 00:02:10.229
<v Niall Sullivan>I mean, why did you, you, I mean, you came
to me with this topic as a suggestion.

34
00:02:10.639 --> 00:02:13.619
<v Niall Sullivan>What sort of got that sort of thinking in
your head, Nick?

35
00:02:14.369 --> 00:02:20.169
<v Nick Leeder>Yeah, well, to be, to be honest, when you
say hype, I mean there's hype and then

36
00:02:20.169 --> 00:02:21.389
<v Nick Leeder>there's hyper hype.

37
00:02:21.899 --> 00:02:28.869
<v Nick Leeder>And I think where we are right now is that
everybody, vendors, my, my clients

38
00:02:28.869 --> 00:02:34.589
<v Nick Leeder>in the manufacturing industry, everybody
needs to, or they feel at least they need

39
00:02:34.589 --> 00:02:38.490
<v Nick Leeder>to be seen to be doing something with um
AI.

40
00:02:39.199 --> 00:02:42.800
<v Nick Leeder>And you know, even some of the execs I've
been speaking with, I mean I was speaking

41
00:02:42.800 --> 00:02:47.440
<v Nick Leeder>with the CTO of a UK PLC company and he's
going, he's sitting in board meetings and

42
00:02:47.440 --> 00:02:50.580
<v Nick Leeder>they're talking about what are the AI
projects that they need to do.

43
00:02:52.070 --> 00:02:58.350
<v Nick Leeder>Not, what are the problems we need to
solve and how a, how AI can help us solve

44
00:02:58.350 --> 00:03:04.259
<v Nick Leeder>those problems or change the customer
experience or provide new functionality that

45
00:03:04.259 --> 00:03:06.589
<v Nick Leeder>hasn't existed or insights.

46
00:03:07.029 --> 00:03:13.990
<v Nick Leeder>So uh the, the thing that's frustrating I
guess for me is that AI however

47
00:03:13.990 --> 00:03:19.029
<v Nick Leeder>powerful it is, I don't think we've yet
really understood the true value and

48
00:03:19.029 --> 00:03:20.369
<v Nick Leeder>potential, and we're probably.

49
00:03:21.089 --> 00:03:23.220
<v Nick Leeder>Spinning it in the wrong way.

50
00:03:27.750 --> 00:03:28.389
<v Niall Sullivan>And that's.

51
00:03:28.500 --> 00:03:31.649
<v Niall Sullivan>I mean, you, you've sort of hit the nail
on the head there in terms of, you know,

52
00:03:32.669 --> 00:03:34.270
<v Niall Sullivan>understanding the value and the value.

53
00:03:34.500 --> 00:03:38.949
<v Niall Sullivan>Do you think, and that, that must be an
issue where people hear about AI and needing

54
00:03:38.949 --> 00:03:39.940
<v Niall Sullivan>to do AI projects.

55
00:03:39.949 --> 00:03:44.110
<v Niall Sullivan>Is it sort of going too quickly ahead with
things without actually thinking about,

56
00:03:44.220 --> 00:03:47.669
<v Niall Sullivan>you know, how, how can we apply this to,
you know, business challenges?

57
00:03:47.880 --> 00:03:48.020
<v Niall Sullivan>Yeah,

58
00:03:48.910 --> 00:03:54.669
<v Nick Leeder>there's a phrase actually I'm, I'm hearing
more and more these days, um, or at least

59
00:03:54.669 --> 00:03:56.449
<v Nick Leeder>I'm saying it more and more, which is.

60
00:03:56.949 --> 00:04:02.100
<v Nick Leeder>The capability of these tools is
outstripping our ability to take the value from it.

61
00:04:03.160 --> 00:04:04.639
<v Nick Leeder>So, um.

62
00:04:05.529 --> 00:04:09.789
<v Nick Leeder>What I, what I mean by that is, uh, you
know, I, I think that.

63
00:04:10.889 --> 00:04:15.460
<v Nick Leeder>Just because a tool can do something
doesn't mean we necessarily have to use it.

64
00:04:15.960 --> 00:04:21.559
<v Nick Leeder>Um, and there are so many clever things in
simple things we can do in our operations

65
00:04:21.559 --> 00:04:26.640
<v Nick Leeder>in our facilities that take us a step
forward, that AI as a second or maybe third

66
00:04:26.640 --> 00:04:32.529
<v Nick Leeder>step could be incredibly powerful or
incredibly valuable, but once we've established

67
00:04:32.529 --> 00:04:37.559
<v Nick Leeder>perhaps core data that is trusted, that is
contextualized in the right way that you

68
00:04:37.559 --> 00:04:40.079
<v Nick Leeder>can train off effectively.

69
00:04:40.760 --> 00:04:44.959
<v Nick Leeder>I, I think those things that just having
that foundation actually takes you a

70
00:04:44.959 --> 00:04:45.859
<v Nick Leeder>massive step forward.

71
00:04:46.399 --> 00:04:51.079
<v Nick Leeder>So we're jumping, we're jumping the
technology, so technology essentially is

72
00:04:51.079 --> 00:04:55.000
<v Nick Leeder>outstripping our ability to consume it.

73
00:04:55.079 --> 00:04:58.089
<v Nick Leeder>I'm sorry there's a little bit of
background noise perhaps going on, but you know it

74
00:04:58.119 --> 00:05:03.279
<v Nick Leeder>it's outstripping our ability to consume
it and therefore we are now, we're almost

75
00:05:03.279 --> 00:05:07.920
<v Nick Leeder>kind of being driven by the technology
pace rather than our pace to.

76
00:05:08.309 --> 00:05:13.309
<v Nick Leeder>Both, uh, capitalize on it, but also, you
know, let's face it, we've, we've been

77
00:05:13.309 --> 00:05:17.790
<v Nick Leeder>living in this world of digital
transformation for, um, I, I, I'm, if you can't, you

78
00:05:17.790 --> 00:05:22.029
<v Nick Leeder>can't see this, but I'm doing air quotes,
yeah, digital transformation, um.

79
00:05:22.910 --> 00:05:26.630
<v Nick Leeder>We've been living in it for, well, at
least I've been doing it since the year 2000,

80
00:05:26.750 --> 00:05:30.209
<v Nick Leeder>you know, 25, nearly 25 years into this
realm.

81
00:05:30.630 --> 00:05:34.250
<v Nick Leeder>We've all got a little bit of change
fatigue and then, you know, getting pushed now

82
00:05:34.709 --> 00:05:37.730
<v Nick Leeder>to do the next thing, the next thing,
because the technology can do it.

83
00:05:38.190 --> 00:05:42.950
<v Nick Leeder>Actually, we've got to get back to, right,
how is this gonna improve the quality of

84
00:05:42.950 --> 00:05:45.589
<v Nick Leeder>my product, the reliability of my
machines?

85
00:05:45.829 --> 00:05:49.019
<v Nick Leeder>How is this increasing the throughput
without increasing my capacity?

86
00:05:49.269 --> 00:05:51.829
<v Nick Leeder>How can I avoid the net Capital
expenditure?

87
00:05:52.140 --> 00:05:56.339
<v Nick Leeder>You know, those things, how can you know,
how can I, it's very topical from a UK

88
00:05:56.339 --> 00:06:01.420
<v Nick Leeder>point of view, how can I optimize
machines' performance to use less energy to

89
00:06:01.420 --> 00:06:02.619
<v Nick Leeder>produce the same output.

90
00:06:03.679 --> 00:06:09.000
<v Nick Leeder>Let's focus on those core problems and
then see how the technology supports us

91
00:06:09.200 --> 00:06:10.399
<v Nick Leeder>rather than the other way around.

92
00:06:11.940 --> 00:06:15.600
<v Niall Sullivan>And you mentioned some of those, um, some
of those use cases, so around energy

93
00:06:15.600 --> 00:06:19.269
<v Niall Sullivan>optimization, predictive maintenance, and,
and things like that.

94
00:06:19.309 --> 00:06:24.029
<v Niall Sullivan>I mean, do these use cases often spring to
mind for manufacturers or is it something

95
00:06:24.029 --> 00:06:28.470
<v Niall Sullivan>usually that needs to be, you know, it's
not there until it's sort of in front of

96
00:06:28.470 --> 00:06:29.380
<v Niall Sullivan>their face kind of thing.

97
00:06:29.459 --> 00:06:31.149
<v Niall Sullivan>I'm just wondering whether it's like an
obvious thing.

98
00:06:31.649 --> 00:06:34.970
<v Nick Leeder>It, it, it kind of, it kind of is and it
kind of isn't.

99
00:06:35.489 --> 00:06:36.309
<v Nick Leeder>I was about to say.

100
00:06:36.679 --> 00:06:41.609
<v Nick Leeder>If you're a manufacturer that goes to
trade shows, I mean, I, I guess most companies

101
00:06:41.609 --> 00:06:45.410
<v Nick Leeder>if they have in the UK, if they've been in
the UK, the big show is at the beginning

102
00:06:45.410 --> 00:06:50.329
<v Nick Leeder>of June at the NEC with the digital
manufacturing, um, week, um.

103
00:06:50.769 --> 00:06:56.450
<v Nick Leeder>And you could not move for AI stands,
yeah.

104
00:06:56.649 --> 00:07:02.730
<v Nick Leeder>Oh, we do AI this, we kind of have AI
enabled tissue dispensers for your toilets,

105
00:07:02.809 --> 00:07:04.200
<v Nick Leeder>you know, all this sort of stuff.

106
00:07:04.369 --> 00:07:09.160
<v Nick Leeder>You can track the paper usage from it, um,
it, it's gone mad.

107
00:07:09.399 --> 00:07:16.130
<v Nick Leeder>It has gone mad and it's, it's like there
are some very, very important things

108
00:07:16.130 --> 00:07:17.750
<v Nick Leeder>I think that.

109
00:07:19.029 --> 00:07:22.359
<v Nick Leeder>Are best suited, I guess, to AI
applications.

110
00:07:23.140 --> 00:07:28.059
<v Nick Leeder>So you know we talk about predictive
maintenance, you know, predictive maintenance,

111
00:07:28.140 --> 00:07:34.859
<v Nick Leeder>as long as you can get constant signals
from your sensors, so vibrational

112
00:07:34.859 --> 00:07:38.640
<v Nick Leeder>sensors, uh, thermal sensors, oil quality,
whatever you need.

113
00:07:40.309 --> 00:07:46.429
<v Nick Leeder>Link that back to what you were producing
at that time because there are variables

114
00:07:46.429 --> 00:07:49.470
<v Nick Leeder>there and that's the important thing is
understanding and building the correct

115
00:07:49.470 --> 00:07:51.130
<v Nick Leeder>variables into these models.

116
00:07:52.089 --> 00:07:57.369
<v Nick Leeder>Then what we've proven and then what
companies have proven over time is that you can

117
00:07:57.369 --> 00:08:03.450
<v Nick Leeder>build gradually more models that have
higher degrees of confidence to predict

118
00:08:03.450 --> 00:08:04.769
<v Nick Leeder>potential failures.

119
00:08:06.269 --> 00:08:10.709
<v Nick Leeder>Now the word predict is really, really key
here.

120
00:08:11.660 --> 00:08:14.619
<v Nick Leeder>Because it's not prescriptive, it's a
prediction.

121
00:08:15.140 --> 00:08:21.649
<v Nick Leeder>So that you still need people to first of
all, go in, look at the data,

122
00:08:21.859 --> 00:08:24.339
<v Nick Leeder>interpret that data to understand yes.

123
00:08:25.040 --> 00:08:27.100
<v Nick Leeder>OK, I'm looking at these various signals.

124
00:08:27.209 --> 00:08:29.380
<v Nick Leeder>Perhaps my peak acceleration's gone up.

125
00:08:29.799 --> 00:08:34.659
<v Nick Leeder>I've done some analysis and looked at the
spectral, uh, you know, at the FFTs and

126
00:08:34.880 --> 00:08:39.159
<v Nick Leeder>going, OK, yeah, I can see that there's a
probably a wear on the inner raceway or

127
00:08:39.159 --> 00:08:40.500
<v Nick Leeder>shaft misalignment.

128
00:08:41.119 --> 00:08:41.880
<v Nick Leeder>And, and then.

129
00:08:44.010 --> 00:08:49.409
<v Nick Leeder>Use that data to go and verify and then
take corrective action.

130
00:08:50.599 --> 00:08:55.469
<v Nick Leeder>Which is better than waiting for something
to break than basically companies do

131
00:08:55.469 --> 00:08:55.789
<v Nick Leeder>today.

132
00:08:56.900 --> 00:09:02.219
<v Nick Leeder>But we're not yet in that full
prescriptive view, and I think there's actually a a

133
00:09:02.219 --> 00:09:03.919
<v Nick Leeder>confusion here, um.

134
00:09:05.150 --> 00:09:10.510
<v Nick Leeder>Where people think also for AI I think
it's a vendor's role, so apologies to

135
00:09:10.510 --> 00:09:15.349
<v Nick Leeder>everybody from Siemens listening here, but
I think there's also an important role

136
00:09:15.349 --> 00:09:16.070
<v Nick Leeder>for the vendor.

137
00:09:16.859 --> 00:09:19.369
<v Nick Leeder>So let's get back to the predictive
maintenance piece.

138
00:09:19.659 --> 00:09:24.140
<v Nick Leeder>You know, I, I've mentioned FFTs, I've
mentioned, you know, looking and analyzing

139
00:09:24.140 --> 00:09:25.340
<v Nick Leeder>the data here.

140
00:09:25.820 --> 00:09:29.820
<v Nick Leeder>That requires a certain level of
competence and skills and training.

141
00:09:30.869 --> 00:09:35.349
<v Nick Leeder>Whereas if we use AI correctly, we can use
that knowledge, we can embed that

142
00:09:35.349 --> 00:09:41.109
<v Nick Leeder>knowledge into the tool and actually then
produce better insights for people to take

143
00:09:41.109 --> 00:09:43.869
<v Nick Leeder>action without needing that detailed
information.

144
00:09:44.309 --> 00:09:48.070
<v Nick Leeder>Isn't it better to say, hey guys, this
machine's gonna potentially fail cos the

145
00:09:48.070 --> 00:09:52.950
<v Nick Leeder>shafts are misaligned or you need to go
and check the lubrication on the machine.

146
00:09:54.690 --> 00:10:01.489
<v Nick Leeder>And just have it as a statement like that
rather than, hey, there's a warning on you

147
00:10:01.489 --> 00:10:04.729
<v Nick Leeder>know your RMS or something else like that,
you've gotta go off and do something.

148
00:10:05.330 --> 00:10:10.349
<v Nick Leeder>So we need to take uh the world of
complexity and I've had, by the way, I've had a

149
00:10:10.890 --> 00:10:14.969
<v Nick Leeder>recent report or recent blog written and
uh published by the World Economic Forum on

150
00:10:14.969 --> 00:10:19.130
<v Nick Leeder>this, which is, do we have a skills crisis
or do we have a usability crisis?

151
00:10:19.570 --> 00:10:22.869
<v Nick Leeder>And I think this is where AI could put
massive benefit.

152
00:10:23.580 --> 00:10:29.780
<v Nick Leeder>Is to simplify the complexity of all this
data to simple actions and information

153
00:10:29.780 --> 00:10:34.059
<v Nick Leeder>that people can understand irrespective of
whether you're a manager, whether you're

154
00:10:34.059 --> 00:10:39.820
<v Nick Leeder>an a line technician or whether you're a
detail VA2, um, in the maintenance team.

155
00:10:42.090 --> 00:10:48.210
<v Niall Sullivan>I guess there's another, uh, I, I
completely get, um, the, the points you're making,

156
00:10:48.289 --> 00:10:52.869
<v Niall Sullivan>but I guess in order to take action from
AI insights, there has to be a high level

157
00:10:53.530 --> 00:10:57.169
<v Niall Sullivan>degree of trust in AI as well, because
there's one thing saying AI is everywhere and

158
00:10:57.169 --> 00:11:01.489
<v Niall Sullivan>we must use AI, but that human instinct to
actually trust what the AI is saying is

159
00:11:01.489 --> 00:11:04.250
<v Niall Sullivan>crucial as well, right, to the adoption of
these tools.

160
00:11:05.109 --> 00:11:07.140
<v Nick Leeder>Yeah, and it, it occurs over time.

161
00:11:07.880 --> 00:11:09.039
<v Nick Leeder>So, um.

162
00:11:09.969 --> 00:11:15.570
<v Nick Leeder>You know, I, I think it's good by the way,
to be skeptical, to ask for the tools to

163
00:11:15.570 --> 00:11:17.270
<v Nick Leeder>prove themselves to be right.

164
00:11:17.770 --> 00:11:22.530
<v Nick Leeder>Um, but actually, particularly with that
usability question answered.

165
00:11:23.340 --> 00:11:30.140
<v Nick Leeder>I, I've seen real success stories happen
where people, you know, operators get and

166
00:11:30.140 --> 00:11:34.760
<v Nick Leeder>use um use the tool, the barrier to entry
is lower.

167
00:11:35.219 --> 00:11:39.780
<v Nick Leeder>They actually go and find out yes the tool
was right or sometimes the tool was

168
00:11:39.780 --> 00:11:40.299
<v Nick Leeder>almost right.

169
00:11:40.380 --> 00:11:45.340
<v Nick Leeder>I mean I've got examples of where we think
there's a a bearing getting loose on a

170
00:11:45.340 --> 00:11:45.619
<v Nick Leeder>fan.

171
00:11:46.210 --> 00:11:49.609
<v Nick Leeder>And when we actually look at the fan,
what's happened is that the fan's imbalance

172
00:11:49.609 --> 00:11:55.010
<v Nick Leeder>because of some um debris on the fan
blades, but it gave the same signal as a kind

173
00:11:55.010 --> 00:11:58.450
<v Nick Leeder>of vibrational signal as you would do with
a a a warm bearing.

174
00:11:59.619 --> 00:12:03.979
<v Nick Leeder>So You know, I think there's trust and I
think that you need to understand the

175
00:12:03.979 --> 00:12:06.260
<v Nick Leeder>limitations of these things as well.

176
00:12:06.580 --> 00:12:09.289
<v Nick Leeder>And as you know, you can only do that
through experience.

177
00:12:09.539 --> 00:12:12.650
<v Nick Leeder>That's not saying I'm not gonna use it
because I don't trust it.

178
00:12:12.900 --> 00:12:18.219
<v Nick Leeder>It means that I have to take everything
with a degree of salt, a pinch of salt until

179
00:12:18.219 --> 00:12:22.070
<v Nick Leeder>it's proven and repeatedly proven to give
me the right information.

180
00:12:22.500 --> 00:12:27.119
<v Nick Leeder>And I think that's my gripe a little bit
from these trade shows cos you go around,

181
00:12:27.179 --> 00:12:30.419
<v Nick Leeder>you speak to the vendors touting their
latest AI solution.

182
00:12:31.390 --> 00:12:36.549
<v Nick Leeder>And they don't necessarily have the, the
process knowledge, the detailed process

183
00:12:36.549 --> 00:12:36.849
<v Nick Leeder>knowledge.

184
00:12:37.770 --> 00:12:44.309
<v Nick Leeder>To actually explain really how it works,
um, as opposed to, well that we have models

185
00:12:44.320 --> 00:12:48.909
<v Nick Leeder>to do this, you know, there's a, there's a
big discrepancy between again between

186
00:12:49.469 --> 00:12:51.609
<v Nick Leeder>technology capability and practical
application.

187
00:12:52.830 --> 00:12:56.390
<v Niall Sullivan>And I, but I think it comes back to
another word which I know you use quite

188
00:12:56.390 --> 00:12:58.619
<v Niall Sullivan>regularly, Nick, is about buzzwords as
well.

189
00:12:58.869 --> 00:13:01.690
<v Niall Sullivan>So if you talk about algorithms and things
like that, I guess people think.

190
00:13:02.119 --> 00:13:05.630
<v Niall Sullivan>Or vendors, some vendors are guilty of
thinking that that impresses people, right?

191
00:13:05.679 --> 00:13:09.359
<v Niall Sullivan>If I say, oh, we've got these algorithms,
people are like, oh, brilliant, that's

192
00:13:09.359 --> 00:13:11.979
<v Niall Sullivan>what ding, that's what we need, that's
what we need.

193
00:13:12.200 --> 00:13:14.750
<v Niall Sullivan>Is it that, it's that kind of thing as
well, right, that's playing into it?

194
00:13:15.039 --> 00:13:18.000
<v Nick Leeder>Yeah, I mean, obviously if you've got
algorithms, there's creams for that, we can,

195
00:13:18.039 --> 00:13:23.640
<v Nick Leeder>uh, solve, solve problems, but in reality,
I, I mean, something like AI, if we kind

196
00:13:23.640 --> 00:13:28.200
<v Nick Leeder>of boil it down, it, it isn't a thing, it
isn't a product, yeah.

197
00:13:28.880 --> 00:13:33.320
<v Nick Leeder>It it's a capability that you're gonna use
and develop and it, it's about turning

198
00:13:33.320 --> 00:13:33.739
<v Nick Leeder>data.

199
00:13:34.679 --> 00:13:37.239
<v Nick Leeder>Into a way that you can make better
decisions.

200
00:13:38.070 --> 00:13:42.549
<v Nick Leeder>So again, let's let's strip these
buzzwords back to actually what it's actually

201
00:13:42.549 --> 00:13:43.530
<v Nick Leeder>doing for you.

202
00:13:44.619 --> 00:13:47.380
<v Nick Leeder>You know, it's algorithm, I mean who cares
about an algorithm?

203
00:13:47.739 --> 00:13:51.250
<v Nick Leeder>It's about helping you use the data you've
got to make better decisions, yeah.

204
00:13:51.619 --> 00:13:57.219
<v Nick Leeder>It's about augmenting the team um and
providing, you know, what AI is brilliant at

205
00:13:57.219 --> 00:14:03.250
<v Nick Leeder>doing is spotting patterns of things and
often patterns with disparate pieces of

206
00:14:03.250 --> 00:14:03.520
<v Nick Leeder>information.

207
00:14:03.859 --> 00:14:09.140
<v Nick Leeder>Is it a combination of the machine running
speed, the, the humidity or the

208
00:14:09.140 --> 00:14:12.539
<v Nick Leeder>temperature in the in in the factory at
the time and the mach what you're actually

209
00:14:12.539 --> 00:14:13.030
<v Nick Leeder>producing.

210
00:14:13.619 --> 00:14:18.140
<v Nick Leeder>And those are quite disparate bits of
information, but it can connect those things

211
00:14:18.140 --> 00:14:23.979
<v Nick Leeder>together and predict outcomes because it
starts seeing patterns of failure or issues

212
00:14:23.979 --> 00:14:27.979
<v Nick Leeder>that occur in specific situations, um.

213
00:14:28.979 --> 00:14:33.039
<v Nick Leeder>And you know, I think when you think about
an approach which is can you sense it,

214
00:14:33.099 --> 00:14:36.179
<v Nick Leeder>can you decide what you need to do and
then can you take action from it.

215
00:14:36.900 --> 00:14:42.619
<v Nick Leeder>You know, tools like AI kind of bring the
horsepower to do this faster, where

216
00:14:42.619 --> 00:14:46.210
<v Nick Leeder>traditionally you'd have you imagine the
amount of Excel spreadsheets you would have

217
00:14:46.210 --> 00:14:49.330
<v Nick Leeder>had to gone through that previous scenario
to try and do that.

218
00:14:49.619 --> 00:14:53.739
<v Nick Leeder>You know, you can speed it up, you can do
it more accurately, and you can bring many

219
00:14:53.739 --> 00:14:58.080
<v Nick Leeder>more variables in that a human
traditionally couldn't analyze effectively.

220
00:14:58.700 --> 00:15:03.820
<v Nick Leeder>So I think when we talk about words like
algorithm, let's just distill it down into

221
00:15:03.820 --> 00:15:10.109
<v Nick Leeder>those simple things that are really
relatable to the operating environment of um

222
00:15:10.109 --> 00:15:12.750
<v Nick Leeder>your your customers and the people
listening to this podcast.

223
00:15:15.440 --> 00:15:17.840
<v Niall Sullivan>Yeah, no, absolutely, absolutely.

224
00:15:18.070 --> 00:15:18.520
<v Niall Sullivan>Um.

225
00:15:19.140 --> 00:15:23.789
<v Niall Sullivan>And OK, another thing sort of stands out
to me as well because you, you mentioned,

226
00:15:23.989 --> 00:15:28.830
<v Niall Sullivan>you know, the, um, Smart factory event
where, you know, there's AI everywhere, and I

227
00:15:28.830 --> 00:15:32.219
<v Niall Sullivan>had a similar experience at the same
event, not, not jumping on this event in

228
00:15:32.219 --> 00:15:36.349
<v Niall Sullivan>particular, but, um, around predictive
maintenance as well, where predictive

229
00:15:36.349 --> 00:15:39.849
<v Niall Sullivan>maintenance was on everything, like
everywhere as well.

230
00:15:40.450 --> 00:15:44.890
<v Niall Sullivan>So it sort of leads me to think when
people are trying to select tools to use, you

231
00:15:44.890 --> 00:15:48.530
<v Niall Sullivan>know, where do they even begin to start
when everything's AI, everything's got

232
00:15:48.530 --> 00:15:53.530
<v Niall Sullivan>algorithms, everything, you know, what,
what's the key criteria that people should

233
00:15:53.530 --> 00:15:55.169
<v Niall Sullivan>focus on, um,

234
00:15:55.729 --> 00:15:56.049
<v Nick Leeder>yeah,

235
00:15:56.369 --> 00:15:57.419
<v Niall Sullivan>yeah, where do they start?

236
00:15:58.880 --> 00:16:01.320
<v Nick Leeder>It was like, it's like that kid in the
sweet shop moment, isn't it?

237
00:16:01.330 --> 00:16:03.840
<v Nick Leeder>You walk in and go, wow, which one, which
one do I want?

238
00:16:04.289 --> 00:16:08.489
<v Nick Leeder>So, like a kid would do in a sweet shop,
the first thing is decide.

239
00:16:09.030 --> 00:16:13.619
<v Nick Leeder>What sort of suite you're really after,
and then lets me translate that into into

240
00:16:13.619 --> 00:16:15.080
<v Nick Leeder>more meaningful terms, yeah?

241
00:16:15.619 --> 00:16:19.119
<v Nick Leeder>What's the problem you're trying to solve?

242
00:16:19.690 --> 00:16:21.059
<v Nick Leeder>Is it about energy use?

243
00:16:21.140 --> 00:16:22.419
<v Nick Leeder>Is it about quality?

244
00:16:22.500 --> 00:16:26.010
<v Nick Leeder>Is it about maintenance and unplanned
downtime or availability, yeah?

245
00:16:27.150 --> 00:16:29.989
<v Nick Leeder>What's the biggest problem you're trying
to solve?

246
00:16:30.830 --> 00:16:31.820
<v Nick Leeder>Understand.

247
00:16:33.200 --> 00:16:36.419
<v Nick Leeder>What wouldn't, what good would look like?

248
00:16:37.900 --> 00:16:43.260
<v Nick Leeder>And then if you've got those two things,
then I think you can start narrowing down

249
00:16:43.260 --> 00:16:49.099
<v Nick Leeder>because you're, you've got a real focus on
an outcome, not a piece of tech.

250
00:16:50.000 --> 00:16:55.320
<v Nick Leeder>Um, and therefore not all immediately, not
all the tech is gonna help, you know, do

251
00:16:55.320 --> 00:17:01.039
<v Nick Leeder>you want that sweet to be sweet, uh, do
you want it to be a gel kind of based thing,

252
00:17:01.109 --> 00:17:02.630
<v Nick Leeder>do you want it to be a sour?

253
00:17:02.989 --> 00:17:07.280
<v Nick Leeder>You know, those are your problem kind of
comparables, and therefore you're narrowing

254
00:17:07.280 --> 00:17:08.560
<v Nick Leeder>down the sweets in the sweet shop.

255
00:17:09.650 --> 00:17:15.560
<v Nick Leeder>Um, And then I think there's other
questions you need to think about um like how

256
00:17:15.560 --> 00:17:20.280
<v Nick Leeder>interoperable are these tools, can it talk
to your systems you have, you might have

257
00:17:20.280 --> 00:17:26.760
<v Nick Leeder>an MES system, a SCA um an ERP system, can
it connect to those things really easily?

258
00:17:27.239 --> 00:17:32.839
<v Nick Leeder>Because you're gonna need that data to
feed these sort of AI based tools with more

259
00:17:32.839 --> 00:17:34.020
<v Nick Leeder>contextual data.

260
00:17:34.760 --> 00:17:39.640
<v Nick Leeder>Um, do you need a PhD in data science to
use it?

261
00:17:40.380 --> 00:17:44.380
<v Nick Leeder>Or is this kind of a low code, no code
type deployment, yeah?

262
00:17:45.189 --> 00:17:49.550
<v Nick Leeder>That's another kind of criteria, you know,
I, I remember going to an event and then

263
00:17:49.550 --> 00:17:50.869
<v Nick Leeder>going we need more.

264
00:17:51.030 --> 00:17:54.589
<v Nick Leeder>The problem with manufacturing is we can't
attract enough data scientists because

265
00:17:54.589 --> 00:17:57.770
<v Nick Leeder>the financial services sector is taking
them all.

266
00:17:58.150 --> 00:17:59.849
<v Nick Leeder>And I'm going, hold on, this is wrong.

267
00:18:00.270 --> 00:18:03.849
<v Nick Leeder>We need AI to be used and data sciences to
be used by everybody.

268
00:18:04.310 --> 00:18:06.969
<v Nick Leeder>You don't need that, we shouldn't need
those sort of skills.

269
00:18:07.780 --> 00:18:10.939
<v Nick Leeder>And the, the final question is, what's the
ROI?

270
00:18:11.089 --> 00:18:16.739
<v Nick Leeder>Is the problem big enough and worth it to
solve, you know, are, if you deploy this,

271
00:18:16.819 --> 00:18:21.530
<v Nick Leeder>will the ROI be 6 months, 12 months or 18
months?

272
00:18:22.020 --> 00:18:25.979
<v Nick Leeder>You know, if you get the, if, if you get a
6 month ROI on this, you've got a big

273
00:18:25.979 --> 00:18:29.199
<v Nick Leeder>problem that can be solved with a very
rapid return.

274
00:18:30.150 --> 00:18:34.310
<v Nick Leeder>That, that helps your case to get the
funding to actually launch these initiatives

275
00:18:34.310 --> 00:18:34.770
<v Nick Leeder>as well.

276
00:18:36.780 --> 00:18:42.300
<v Niall Sullivan>And I wonder with, with AI as well,
whether it, it also comes with expectations of

277
00:18:42.300 --> 00:18:46.349
<v Niall Sullivan>success because everyone knows how
amazingly productive AI makes everyone.

278
00:18:46.579 --> 00:18:51.969
<v Niall Sullivan>But does that also create Yeah, an issue
sometimes where you expect instant results

279
00:18:51.969 --> 00:18:55.150
<v Niall Sullivan>or otherwise, I'll no, that's, that, that,
that obviously doesn't work.

280
00:18:55.930 --> 00:18:59.050
<v Niall Sullivan>There's a, there's not, there's a bit of
patience required, right, for depends on

281
00:18:59.050 --> 00:19:00.689
<v Niall Sullivan>the technology we're talking about, of
course,

282
00:19:00.810 --> 00:19:05.130
<v Nick Leeder>but there, there, there is, but I, I, I
guess, you know, I don't think there's ever

283
00:19:05.130 --> 00:19:06.869
<v Nick Leeder>been instant payback on anything.

284
00:19:07.329 --> 00:19:13.290
<v Nick Leeder>Um, I've seen some phenomenal paybacks or
at least phenomenal results, uh, with,

285
00:19:13.410 --> 00:19:17.729
<v Nick Leeder>with some of these tools, you know, and
going just simply measuring stuff, you start

286
00:19:17.729 --> 00:19:20.079
<v Nick Leeder>to see stuff you didn't know existed.

287
00:19:20.449 --> 00:19:25.770
<v Nick Leeder>So for example, I worked with a customer
that the HVAC system was working at, um, 33

288
00:19:25.770 --> 00:19:26.050
<v Nick Leeder>a.m.

289
00:19:26.050 --> 00:19:27.250
<v Nick Leeder>in the morning when the factory was shut.

290
00:19:28.530 --> 00:19:32.290
<v Nick Leeder>It's costing them money, you know, that's
not AI by the way, that's just simple.

291
00:19:32.359 --> 00:19:35.500
<v Nick Leeder>Well, what's, what's that system turning
on at that time in the morning?

292
00:19:35.880 --> 00:19:36.400
<v Nick Leeder>But um.

293
00:19:37.589 --> 00:19:42.589
<v Nick Leeder>You know, I, I think the problem here, the
other problem is when we start initiating

294
00:19:42.589 --> 00:19:46.530
<v Nick Leeder>these projects, we, we make the projects
also too big.

295
00:19:47.540 --> 00:19:54.500
<v Nick Leeder>So are we trying to eat the elephant in
one go, you know, or do you just take bites

296
00:19:54.500 --> 00:19:55.229
<v Nick Leeder>to prove.

297
00:19:56.280 --> 00:20:00.439
<v Nick Leeder>And starting with very, you know, starting
with something smaller.

298
00:20:01.229 --> 00:20:05.530
<v Nick Leeder>With very clear metrics that you can
demonstrate whether that's like reduced and

299
00:20:05.530 --> 00:20:12.180
<v Nick Leeder>unplanned down um, unplanned downtime, uh,
yield or energy, yeah, um,

300
00:20:12.420 --> 00:20:18.390
<v Nick Leeder>run experiments as well so that in, you
know, every 90 days you're kind of moving it

301
00:20:18.390 --> 00:20:20.229
<v Nick Leeder>forward, you're capturing the benefit.

302
00:20:20.630 --> 00:20:25.760
<v Nick Leeder>You're able to share the story of what
you've done and the results that you've seen,

303
00:20:25.839 --> 00:20:28.689
<v Nick Leeder>that builds confidence and buy-in within
the organization.

304
00:20:29.390 --> 00:20:35.020
<v Nick Leeder>Um, and by solving one problem at a time
and gradually building, this is something

305
00:20:35.199 --> 00:20:41.050
<v Nick Leeder>also really important for AI, yeah, is
gradually building up the training data and

306
00:20:41.050 --> 00:20:44.760
<v Nick Leeder>showing that it's actually working and
correlating correctly.

307
00:20:45.239 --> 00:20:50.040
<v Nick Leeder>You know, building that momentum, one
problem at a time, you know, perhaps one per

308
00:20:50.040 --> 00:20:50.719
<v Nick Leeder>quarter.

309
00:20:51.939 --> 00:20:55.319
<v Nick Leeder>That I think is the big difference here.

310
00:20:55.650 --> 00:20:59.609
<v Nick Leeder>So it's not the, the big transformation
that is gonna take, you know, the classic

311
00:20:59.609 --> 00:21:04.630
<v Nick Leeder>world of ERP, yeah, 18 to 24 months at
least, it never works first time round.

312
00:21:05.319 --> 00:21:08.209
<v Nick Leeder>Start smaller, prove it, and gradually
build it.

313
00:21:08.569 --> 00:21:11.189
<v Nick Leeder>But I think it's important, you know, we,
we talk about.

314
00:21:12.300 --> 00:21:17.040
<v Nick Leeder>A little bit about, I, I mentioned eating
the elephant, you need to at least know

315
00:21:17.339 --> 00:21:21.160
<v Nick Leeder>what your roadmap is and where you're
gonna go with it.

316
00:21:21.660 --> 00:21:26.300
<v Nick Leeder>Cos if you only have a small project to
prove, that project could well prove but may

317
00:21:26.300 --> 00:21:29.469
<v Nick Leeder>well never scale because you don't know
where they're.

318
00:21:30.020 --> 00:21:34.140
<v Nick Leeder>You know, how can you apply this to more
facilities or different lines within your

319
00:21:34.140 --> 00:21:38.619
<v Nick Leeder>factory or different machines, is it
scalable that way or where do you build from

320
00:21:38.619 --> 00:21:42.459
<v Nick Leeder>the next capability on it, so it's
important to have a view of what the elephant

321
00:21:42.459 --> 00:21:45.339
<v Nick Leeder>looks like and the steps to get you there.

322
00:21:46.199 --> 00:21:50.170
<v Nick Leeder>But you don't start trying to deliver it
all in one go, you take the first steps,

323
00:21:50.239 --> 00:21:51.680
<v Nick Leeder>you prove it and you grow it.

324
00:21:53.449 --> 00:21:56.890
<v Niall Sullivan>So, so maybe it's sort of, because, I
mean, the, the title of this episode is around

325
00:21:56.890 --> 00:21:57.969
<v Niall Sullivan>sort of practical strategy.

326
00:21:58.079 --> 00:22:03.729
<v Niall Sullivan>So how, how practically, um, can companies
measure these small wins in order to

327
00:22:03.729 --> 00:22:06.329
<v Niall Sullivan>enable the, you know, the, the effective
goal of scaling it further.

328
00:22:06.400 --> 00:22:10.449
<v Niall Sullivan>Obviously, if it, if there's return on
investment, it wants to be scaled everywhere.

329
00:22:10.800 --> 00:22:14.969
<v Niall Sullivan>One see return on investment, you know,
throughout an organization, but how can

330
00:22:14.969 --> 00:22:16.829
<v Niall Sullivan>companies sort of think about that in
reality?

331
00:22:18.300 --> 00:22:21.560
<v Nick Leeder>So I, I, I guess the first conversation
is.

332
00:22:23.150 --> 00:22:27.859
<v Nick Leeder>This, this, this project we're gonna run,
let's, let's say you've got an extrusion

333
00:22:27.859 --> 00:22:31.770
<v Nick Leeder>machine and you've got 10 extrusion
machines in, in, in the in your factory or 10

334
00:22:31.770 --> 00:22:32.939
<v Nick Leeder>presses or whatever, yeah?

335
00:22:33.939 --> 00:22:40.939
<v Nick Leeder>And Start with one and say right, we're
going to improve the reliability of

336
00:22:40.939 --> 00:22:41.699
<v Nick Leeder>this extruder.

337
00:22:42.530 --> 00:22:45.640
<v Nick Leeder>And we're gonna do this because we're
gonna, we're gonna use condition-based

338
00:22:45.640 --> 00:22:49.130
<v Nick Leeder>monitoring or predictive maintenance, you
know, evolving from that.

339
00:22:49.869 --> 00:22:56.079
<v Nick Leeder>Um, and I'll tell you what, and this is
the CFO conversation or your, or your line

340
00:22:56.079 --> 00:22:56.520
<v Nick Leeder>manager.

341
00:22:57.510 --> 00:23:03.040
<v Nick Leeder>To do all 10 is gonna cost 100,000 or
something.

342
00:23:03.239 --> 00:23:06.000
<v Nick Leeder>What I want is 5000 to start off with.

343
00:23:06.719 --> 00:23:10.479
<v Nick Leeder>Because I don't, you know, we need to be
prepared to make that investment, but we've

344
00:23:10.479 --> 00:23:11.280
<v Nick Leeder>gotta prove it.

345
00:23:12.079 --> 00:23:13.729
<v Nick Leeder>Um, so first of all.

346
00:23:14.619 --> 00:23:18.859
<v Nick Leeder>You've you've lowered the financial
barrier to entry in that conversation.

347
00:23:19.599 --> 00:23:21.930
<v Nick Leeder>Um And then.

348
00:23:23.209 --> 00:23:28.959
<v Nick Leeder>Start by sensorizing it, getting the data
and understanding what what's where you

349
00:23:28.959 --> 00:23:34.949
<v Nick Leeder>are today, so I think it's also really
important to try and baseline the

350
00:23:34.949 --> 00:23:35.339
<v Nick Leeder>performance.

351
00:23:36.510 --> 00:23:41.109
<v Nick Leeder>Cause if you can pageline the performance
right, over the past 6 months we've had 5

352
00:23:41.109 --> 00:23:46.300
<v Nick Leeder>failures on this machine, each has taken
us, um, you know, 4 hours on average to

353
00:23:46.300 --> 00:23:46.510
<v Nick Leeder>fix.

354
00:23:46.589 --> 00:23:49.920
<v Nick Leeder>We had 1 major failure that took the
machine down for 2 days.

355
00:23:50.390 --> 00:23:52.250
<v Nick Leeder>That meant we had this much lost
production.

356
00:23:53.239 --> 00:23:58.300
<v Nick Leeder>OK, because of the failure with these
things, and we believe that using

357
00:23:58.819 --> 00:24:03.839
<v Nick Leeder>condition-based monitoring or predictive
maintenance, we can detect these failures

358
00:24:03.839 --> 00:24:04.439
<v Nick Leeder>earlier.

359
00:24:05.420 --> 00:24:09.900
<v Nick Leeder>Schedule the maintenance in and avoid the
failure.

360
00:24:10.089 --> 00:24:15.219
<v Nick Leeder>Now that's always a tricky one for a
business case, yeah, because the avoidance

361
00:24:15.219 --> 00:24:18.560
<v Nick Leeder>question, this is what AI is really good
at, is avoiding things often.

362
00:24:19.099 --> 00:24:24.560
<v Nick Leeder>It's like an insurance policy, um, but
it's a very valuable insurance policy.

363
00:24:25.140 --> 00:24:31.459
<v Nick Leeder>So I, I guess the, the thing there is have
some very, very clear and

364
00:24:31.459 --> 00:24:34.420
<v Nick Leeder>meaningful metrics that everybody would
understand.

365
00:24:35.040 --> 00:24:39.140
<v Nick Leeder>Um, and please think of those metrics in
two ways.

366
00:24:39.479 --> 00:24:45.119
<v Nick Leeder>The operational KPI, the percentage
improvements that you see, uh, at an ops level,

367
00:24:45.229 --> 00:24:46.910
<v Nick Leeder>yeah, that sit in your dashboard.

368
00:24:47.280 --> 00:24:54.079
<v Nick Leeder>But also translate those into a financial
measure that the CFO would understand.

369
00:24:55.589 --> 00:25:01.390
<v Nick Leeder>And if you can connect those two
operational metrics to financial metrics, when you

370
00:25:01.390 --> 00:25:07.619
<v Nick Leeder>prove that you are improving your KPIs,
you can also then directly correlate the

371
00:25:07.619 --> 00:25:09.689
<v Nick Leeder>financial improvement and benefits of the
business.

372
00:25:10.930 --> 00:25:15.709
<v Nick Leeder>And that will, if you can prove it in that
kind of 90 day window I talked about,

373
00:25:16.209 --> 00:25:18.650
<v Nick Leeder>with a relatively low level of investment.

374
00:25:19.800 --> 00:25:22.959
<v Nick Leeder>You can say, right, this is, we've proven
this.

375
00:25:24.069 --> 00:25:27.910
<v Nick Leeder>Why don't we start scaling this across the
rest of the machines, maybe not go to all

376
00:25:27.910 --> 00:25:33.310
<v Nick Leeder>10, but maybe go to another 2, so we've
got 3, and maybe we run this for another 3

377
00:25:33.310 --> 00:25:35.489
<v Nick Leeder>months and we'll see that correlation
again.

378
00:25:35.829 --> 00:25:39.810
<v Nick Leeder>And if you see that correlation again,
well then we're ready to invest.

379
00:25:40.810 --> 00:25:44.650
<v Nick Leeder>So you've kind of, you're scaling, you're
proving at each stage and you've got

380
00:25:44.650 --> 00:25:48.469
<v Nick Leeder>something that's really meaningful for
your KPIs.

381
00:25:48.849 --> 00:25:52.260
<v Nick Leeder>um and just use as much stuff off the
shelf.

382
00:25:52.489 --> 00:25:54.250
<v Nick Leeder>The other, the other habit I see is.

383
00:25:55.300 --> 00:25:59.250
<v Nick Leeder>Particularly when people start playing
with these tools.

384
00:26:00.130 --> 00:26:04.410
<v Nick Leeder>Um, you get a little cottage industry uh
spinning up on the side.

385
00:26:04.969 --> 00:26:11.910
<v Nick Leeder>So actually take stuff that's, um, as much
off the shelf, as easy to integrate to,

386
00:26:12.250 --> 00:26:18.170
<v Nick Leeder>um, and get that data flowing and start
training that data and try and avoid that

387
00:26:18.170 --> 00:26:22.810
<v Nick Leeder>cottage industry because I think you'll
see results faster and particularly if you

388
00:26:22.810 --> 00:26:27.250
<v Nick Leeder>look at long term results, you don't have
to do all that lifting and then

389
00:26:27.250 --> 00:26:29.069
<v Nick Leeder>maintenance of it afterwards as well.

390
00:26:30.989 --> 00:26:35.660
<v Niall Sullivan>And you, you did mention that before,
because when you talk about metrics and, um,

391
00:26:35.670 --> 00:26:40.400
<v Niall Sullivan>return on investment, I guess another
great, uh, uh, maybe a great way to, to pitch

392
00:26:40.400 --> 00:26:45.959
<v Niall Sullivan>it is you've got all these existing
systems, right, with, with data, you know, that

393
00:26:45.959 --> 00:26:49.680
<v Niall Sullivan>may be underutilized, but maybe you can
use AI to add a layer on top and start

394
00:26:49.680 --> 00:26:52.000
<v Niall Sullivan>getting more value from that data that's
locked in there.

395
00:26:52.119 --> 00:26:55.199
<v Niall Sullivan>I know you did talk about that before, but
I thought that linked nicely in terms of

396
00:26:55.199 --> 00:26:57.119
<v Niall Sullivan>that return on investment conversation.

397
00:26:57.930 --> 00:27:04.699
<v Nick Leeder>Yeah, yeah, I, I mean, every time I hear
about it, I, I, I, I keep forgetting the

398
00:27:04.699 --> 00:27:10.339
<v Nick Leeder>stat because it's a phenomenal stat, but
let's just say manufacturing produces the

399
00:27:10.339 --> 00:27:13.599
<v Nick Leeder>most amount of data almost out of any
industry.

400
00:27:13.859 --> 00:27:19.380
<v Nick Leeder>I guess banking might be the only other
one that's up there, banking and retail, um,

401
00:27:19.699 --> 00:27:23.219
<v Nick Leeder>in the amount of data it produces, but the
question is, is it good data?

402
00:27:23.910 --> 00:27:27.930
<v Nick Leeder>So you know, we don't, we don't worry
about the amount of data, but.

403
00:27:28.819 --> 00:27:35.109
<v Nick Leeder>The system, you know, systems in uh are
within your business where the data resides

404
00:27:35.109 --> 00:27:35.650
<v Nick Leeder>and.

405
00:27:36.790 --> 00:27:39.300
<v Nick Leeder>AI is not going to replace them.

406
00:27:39.790 --> 00:27:46.229
<v Nick Leeder>AI is going to use that information to
enhance what you do and the decisions that

407
00:27:46.229 --> 00:27:51.589
<v Nick Leeder>you make, but it's important to pull that
data in from, for example, your

408
00:27:51.589 --> 00:27:55.550
<v Nick Leeder>manufacturing execution system or your MES
system, um.

409
00:27:56.760 --> 00:28:02.239
<v Nick Leeder>Looking at the SCADA or pulling, reporting
that information back perhaps into ERP

410
00:28:03.079 --> 00:28:08.619
<v Nick Leeder>so, um, a lot of the vendors now have got
AI plugins as well, which makes the

411
00:28:08.619 --> 00:28:14.560
<v Nick Leeder>complexity perhaps at times a bit
confusing again, but um, you know, can you pull

412
00:28:14.560 --> 00:28:18.420
<v Nick Leeder>that data in, can you make better
decisions using AI from that?

413
00:28:18.920 --> 00:28:25.520
<v Nick Leeder>You'll undoubtedly will have production
related data in the MES system, you're gonna

414
00:28:25.520 --> 00:28:25.979
<v Nick Leeder>have.

415
00:28:26.420 --> 00:28:31.130
<v Nick Leeder>Machine related data out your SCRA system,
you're gonna have additional sensors that

416
00:28:31.130 --> 00:28:36.140
<v Nick Leeder>you want to deploy to get more specific
data points around that thing.

417
00:28:37.280 --> 00:28:43.000
<v Nick Leeder>And it's, you know, there has to be some
level of kind of context that you apply it

418
00:28:43.000 --> 00:28:47.109
<v Nick Leeder>to, whether it's the asset itself, the
product, I mean, energy.

419
00:28:47.650 --> 00:28:53.199
<v Nick Leeder>One of the major SKUs that a lot of
company or sorry, KPIs that a lot of companies

420
00:28:53.199 --> 00:28:58.640
<v Nick Leeder>are trying to get to is the kilowatt hour
per SKU produced or the CO2 equivalent per

421
00:28:58.640 --> 00:28:59.550
<v Nick Leeder>SKU produced.

422
00:28:59.959 --> 00:29:01.989
<v Nick Leeder>There's reporting requirements from that.

423
00:29:02.359 --> 00:29:06.339
<v Nick Leeder>So pull that data in, start using that,
um.

424
00:29:07.609 --> 00:29:13.670
<v Nick Leeder>You know, integration is often a major
barrier to it, so there is a technical

425
00:29:13.670 --> 00:29:18.560
<v Nick Leeder>lift that's required, that requires
support from your IT teams.

426
00:29:18.810 --> 00:29:21.170
<v Nick Leeder>It requires support from your IT teams.

427
00:29:21.849 --> 00:29:27.250
<v Nick Leeder>But if you get that integrated correctly
and you pull that data together, then

428
00:29:27.250 --> 00:29:32.650
<v Nick Leeder>that's actually probably the biggest
accelerator that you're gonna get for your AI

429
00:29:32.650 --> 00:29:33.660
<v Nick Leeder>projects, um.

430
00:29:34.729 --> 00:29:38.689
<v Nick Leeder>So one of the things is I often look at
the World Economic Forum lighthouse projects

431
00:29:38.689 --> 00:29:39.089
<v Nick Leeder>and.

432
00:29:39.660 --> 00:29:44.140
<v Nick Leeder>Um, you know, lighthouses have tried, one
of the things we see quite regularly with

433
00:29:44.140 --> 00:29:45.420
<v Nick Leeder>the lighthouses is the.

434
00:29:46.050 --> 00:29:51.130
<v Nick Leeder>If this is, I'm not sure this is a real
word, but assetized for their solutions.

435
00:29:51.530 --> 00:29:56.520
<v Nick Leeder>So they're almost like modular pieces, uh,
and toolkits that they have.

436
00:29:56.849 --> 00:30:00.719
<v Nick Leeder>And that starts ensuring the
interoperability and accelerates the deployment that

437
00:30:00.719 --> 00:30:01.189
<v Nick Leeder>they have.

438
00:30:03.290 --> 00:30:07.650
<v Nick Leeder>Soization, it's not very easy to say and
probably again, not a real world, a real

439
00:30:07.650 --> 00:30:09.290
<v Nick Leeder>word, but, um.

440
00:30:10.619 --> 00:30:14.640
<v Nick Leeder>Has a real kind of acceleration component
to what you're trying to do.

441
00:30:16.819 --> 00:30:22.900
<v Niall Sullivan>And, and if we think of these, these
projects as well, so, Uh, maybe there's a,

442
00:30:23.239 --> 00:30:28.069
<v Niall Sullivan>sometimes a misconception or an
overthought process about how much time, resources.

443
00:30:29.199 --> 00:30:34.180
<v Niall Sullivan>Data budgets are required to actually get
these projects off the ground, but I, I

444
00:30:34.180 --> 00:30:38.319
<v Niall Sullivan>guess what's your view on that, Nick, and
is it possible to launch something, you

445
00:30:38.319 --> 00:30:42.680
<v Niall Sullivan>know, which can deliver a great return on
investment without huge resources,

446
00:30:42.760 --> 00:30:44.380
<v Niall Sullivan>whatever that might be.

447
00:30:47.260 --> 00:30:51.260
<v Nick Leeder>Yeah, I, I mean, I, I think if we go back
to what I talked about, um.

448
00:30:52.189 --> 00:30:57.910
<v Nick Leeder>In the beginning, there are factors that
influence that, you know, the scale of your

449
00:30:57.910 --> 00:31:04.510
<v Nick Leeder>aspiration drives the budget, drives that,
the cost uh and the resource

450
00:31:04.510 --> 00:31:08.869
<v Nick Leeder>requirements, so starting with relatively
smaller things that you can prove but have

451
00:31:08.869 --> 00:31:12.449
<v Nick Leeder>a disproportionate impact to the way
you're performing.

452
00:31:13.140 --> 00:31:15.199
<v Nick Leeder>That's a, that's a way of limiting it.

453
00:31:15.290 --> 00:31:19.920
<v Nick Leeder>And another factor, and I've talked about
that already is the usability of the tool.

454
00:31:20.439 --> 00:31:26.900
<v Nick Leeder>Um, you know, the skills that you need, if
you need data scientists, um, your

455
00:31:26.900 --> 00:31:31.140
<v Nick Leeder>OT team and your IT team to get this stuff
up and running.

456
00:31:31.579 --> 00:31:36.930
<v Nick Leeder>There's a lot of external pulls that need
to happen and collaboration that needs to

457
00:31:36.930 --> 00:31:39.390
<v Nick Leeder>happen to make these things a reality in
your business.

458
00:31:39.920 --> 00:31:46.229
<v Nick Leeder>So can you just connect to a machine
relatively simply, um, use some, you know,

459
00:31:46.310 --> 00:31:51.469
<v Nick Leeder>retrofitting sensors that are either wired
or running off Bluetooth or Laura WAN or

460
00:31:51.469 --> 00:31:54.670
<v Nick Leeder>something, connect it to a gateway and
pull the data that way.

461
00:31:55.569 --> 00:32:01.729
<v Nick Leeder>That actually removes some of those
technology barriers which pull that resource uh

462
00:32:01.729 --> 00:32:02.530
<v Nick Leeder>position in.

463
00:32:03.420 --> 00:32:10.209
<v Nick Leeder>Um, So, you know, when you kind of start
in that small win

464
00:32:10.209 --> 00:32:13.630
<v Nick Leeder>approach, you know, again back to the
lighthouse, um.

465
00:32:14.420 --> 00:32:21.300
<v Nick Leeder>These transformations tend to happen 25 to
50% faster when you kind of start

466
00:32:21.300 --> 00:32:22.640
<v Nick Leeder>small and scaling that.

467
00:32:23.430 --> 00:32:29.069
<v Nick Leeder>Um, and that, that, that comes from
remaining focused on what you're trying to

468
00:32:29.069 --> 00:32:29.489
<v Nick Leeder>prove.

469
00:32:30.199 --> 00:32:34.880
<v Nick Leeder>And accepting that you're gonna do this in
an iterative way.

470
00:32:35.959 --> 00:32:42.640
<v Nick Leeder>Because I guess we've heard this phrase
yet, um, you know, start small, scale

471
00:32:42.640 --> 00:32:44.180
<v Nick Leeder>fast or fail fast.

472
00:32:44.959 --> 00:32:50.000
<v Nick Leeder>And I think that's really important
because resources are always going to be

473
00:32:50.000 --> 00:32:50.619
<v Nick Leeder>limited.

474
00:32:50.959 --> 00:32:54.359
<v Nick Leeder>So if it's not gonna work, you don't want
to spend 6 months trying to prove

475
00:32:54.359 --> 00:32:56.170
<v Nick Leeder>something and find it doesn't work then.

476
00:32:56.520 --> 00:33:00.500
<v Nick Leeder>You want to prove it within 2 or 2 or 3
months that it's gonna work or not.

477
00:33:00.920 --> 00:33:05.239
<v Nick Leeder>Have a plan to scale it or a plan to say
look, let's be honest with ourselves, this

478
00:33:05.239 --> 00:33:05.920
<v Nick Leeder>isn't working.

479
00:33:08.729 --> 00:33:12.439
<v Niall Sullivan>Yeah, and I think, I think that's a key
part of it as well, because I think.

480
00:33:13.530 --> 00:33:16.709
<v Niall Sullivan>Yeah, although it's just an interesting
point about the honesty part, because there

481
00:33:16.709 --> 00:33:21.339
<v Niall Sullivan>is a, you know, sometimes when time's been
invested or resources have been invested,

482
00:33:21.400 --> 00:33:26.140
<v Niall Sullivan>it's hard to let projects go, but I guess
that means you can allocate resources to

483
00:33:26.140 --> 00:33:29.199
<v Niall Sullivan>another project, which is maybe a better
fit for your business, right?

484
00:33:30.099 --> 00:33:31.640
<v Nick Leeder>Have a higher probability of success.

485
00:33:31.699 --> 00:33:36.650
<v Nick Leeder>I mean, let, let's face it, the world
we're in right now, um, with all the, um,

486
00:33:36.699 --> 00:33:41.760
<v Nick Leeder>uncertainties, getting any funding, uh,
particularly large scale funding.

487
00:33:42.219 --> 00:33:48.800
<v Nick Leeder>It, it is difficult and um you've gotta
have almost that bulletproof.

488
00:33:49.589 --> 00:33:54.109
<v Nick Leeder>Uh, business case to the CFO to get
anything large off the ground, so you, you know,

489
00:33:54.310 --> 00:33:59.790
<v Nick Leeder>start that smaller position, um, but also,
you know, I don't know, I've come from

490
00:33:59.790 --> 00:34:02.209
<v Nick Leeder>running these programs in industry for a
long time.

491
00:34:02.540 --> 00:34:07.810
<v Nick Leeder>If I see a report that is all green, I
immediately become suspicious, yeah.

492
00:34:09.060 --> 00:34:11.149
<v Nick Leeder>It's like, are you really being honest
with me?

493
00:34:11.469 --> 00:34:14.780
<v Nick Leeder>You know, was a, what we've got to have at
least an amber somewhere, and you're

494
00:34:14.780 --> 00:34:16.239
<v Nick Leeder>telling me there's no red issues.

495
00:34:16.550 --> 00:34:19.790
<v Nick Leeder>And it's, it's like running in the uh just
looking at your line reports as well,

496
00:34:19.870 --> 00:34:24.070
<v Nick Leeder>yeah, and if there's no reds, everything's
working perfectly, but we know the world

497
00:34:24.070 --> 00:34:25.250
<v Nick Leeder>doesn't work that way.

498
00:34:26.079 --> 00:34:32.898
<v Nick Leeder>So, um, you know, failure is that I, I,
again I, I've, I've worked with James Dyson

499
00:34:32.898 --> 00:34:37.839
<v Nick Leeder>in my career, yeah, and I worked for Dyson
and you know it's, there was one kind of

500
00:34:37.839 --> 00:34:43.358
<v Nick Leeder>mantra and spoken mantra there which is,
um, making a mistake is not a problem,

501
00:34:43.398 --> 00:34:44.839
<v Nick Leeder>yeah, cos you learn from it.

502
00:34:46.148 --> 00:34:52.229
<v Nick Leeder>Making the same mistake twice starts
becoming a problem because actually you haven't

503
00:34:52.229 --> 00:34:55.229
<v Nick Leeder>learned, you haven't adopted adapted to
the situation.

504
00:34:55.749 --> 00:34:58.849
<v Nick Leeder>So the same applies to these sorts of
projects.

505
00:34:59.229 --> 00:35:03.898
<v Nick Leeder>Getting it right all the time doesn't
actually help, to be honest.

506
00:35:04.388 --> 00:35:09.779
<v Nick Leeder>Having some failures, sometimes by having
a failure, a new opportunity emerges from

507
00:35:09.779 --> 00:35:13.019
<v Nick Leeder>it you didn't predict in the first place.

508
00:35:13.280 --> 00:35:17.280
<v Nick Leeder>You know, you could start off trying to do
predictive maintenance and actually find

509
00:35:17.280 --> 00:35:22.639
<v Nick Leeder>out that by, by using energy data, you've
found a massive hole in, in what you're

510
00:35:22.639 --> 00:35:27.399
<v Nick Leeder>doing, you're running the machines much
too higher power, um, or they're running

511
00:35:27.399 --> 00:35:29.100
<v Nick Leeder>idle for way too long.

512
00:35:29.600 --> 00:35:34.120
<v Nick Leeder>You've found a new case, a new benefit,
which is what not necessarily what you

513
00:35:34.120 --> 00:35:36.179
<v Nick Leeder>started to, to have in the first place.

514
00:35:37.919 --> 00:35:41.719
<v Niall Sullivan>That's and I can relate it back to Sensai
actually, because I spoke to one of our

515
00:35:41.719 --> 00:35:46.399
<v Niall Sullivan>delivery team a few months ago, who, who,
um, told me a little story about a

516
00:35:46.399 --> 00:35:51.439
<v Niall Sullivan>customer who was almost reluctant to
provide negative feedback about the app that it

517
00:35:51.439 --> 00:35:55.560
<v Niall Sullivan>missed something, or I can't remember the
exact details, but in fact, he was sort of

518
00:35:55.560 --> 00:35:59.199
<v Niall Sullivan>saying, no, no, no, you should, you should
be telling, telling us this because like

519
00:35:59.199 --> 00:36:03.399
<v Niall Sullivan>I said, it, it almost creates, if you spin
it positively, it creates an opportunity

520
00:36:03.399 --> 00:36:07.699
<v Niall Sullivan>for something else, a new, a new place to
focus on, a new.

521
00:36:08.209 --> 00:36:13.060
<v Niall Sullivan>Um, just new ways to improve, so it
doesn't always have a negative that can be

522
00:36:13.060 --> 00:36:14.459
<v Niall Sullivan>turned into a positive quite easily.

523
00:36:14.800 --> 00:36:14.810
<v Niall Sullivan>I,

524
00:36:15.949 --> 00:36:19.580
<v Nick Leeder>I, I worked with a brilliant guy called
Pete Morris, and a big shout out to Pete on

525
00:36:19.580 --> 00:36:23.129
<v Nick Leeder>this, but with Pete, he, he, I gave him
some feedback recently.

526
00:36:23.179 --> 00:36:25.689
<v Nick Leeder>I said, look, sorry mate, I've got this
idea for you.

527
00:36:26.060 --> 00:36:29.570
<v Nick Leeder>Here's some feedback, very well
constructed, at least I think I did it.

528
00:36:29.620 --> 00:36:30.899
<v Nick Leeder>I think I did it constructively.

529
00:36:31.320 --> 00:36:36.459
<v Nick Leeder>And he went, no worries, feedback is the
food of champions, you know, it's like,

530
00:36:36.959 --> 00:36:41.439
<v Nick Leeder>it's like great, none of us can get better
if we don't get feedback on what's going

531
00:36:41.439 --> 00:36:41.780
<v Nick Leeder>on.

532
00:36:41.959 --> 00:36:44.179
<v Nick Leeder>And that applies to the projects you're
running as well.

533
00:36:44.689 --> 00:36:45.040
<v Niall Sullivan>Yeah.

534
00:36:45.510 --> 00:36:47.040
<v Niall Sullivan>Applies to every walk of life actually.

535
00:36:47.080 --> 00:36:51.239
<v Niall Sullivan>I'll probably um broaden it out, but
speaking from personal experience, um.

536
00:36:51.850 --> 00:36:53.340
<v Niall Sullivan>But, uh, but no, that's fantastic.

537
00:36:53.459 --> 00:36:57.419
<v Niall Sullivan>I think, um, we, we're coming towards the
end and as, you know, regular listeners

538
00:36:57.419 --> 00:37:00.979
<v Niall Sullivan>will know, we always, and you would know,
as a regular, um, guest as well, we sort

539
00:37:00.979 --> 00:37:05.699
<v Niall Sullivan>of try and end with, there's been a lot of
practical sort of takeaways today anyway,

540
00:37:05.770 --> 00:37:09.219
<v Niall Sullivan>but if there's sort of one or two things
that people sort of take away and start

541
00:37:09.219 --> 00:37:12.979
<v Niall Sullivan>thinking about when, when thinking of
running these AI projects, what, what would

542
00:37:12.979 --> 00:37:13.360
<v Niall Sullivan>that be?

543
00:37:14.620 --> 00:37:20.860
<v Nick Leeder>I'm So I, I think all my advice is
probably gonna be centered around.

544
00:37:21.850 --> 00:37:28.090
<v Nick Leeder>Proving the impact, cos again it's not
that the technology does not work, it works,

545
00:37:28.169 --> 00:37:28.520
<v Nick Leeder>yeah.

546
00:37:28.850 --> 00:37:33.169
<v Nick Leeder>In fact, as I said earlier on, the
capability of these technologies is outstripping

547
00:37:33.169 --> 00:37:36.639
<v Nick Leeder>our ability to use them and create value
from them.

548
00:37:36.929 --> 00:37:38.669
<v Nick Leeder>And if you have that in the back of your
mind.

549
00:37:39.620 --> 00:37:41.340
<v Nick Leeder>All the time, you know.

550
00:37:42.209 --> 00:37:46.729
<v Nick Leeder>Don't ask where can we use AI, that's the
wrong question.

551
00:37:47.889 --> 00:37:50.989
<v Nick Leeder>Ask what's one problem we must solve.

552
00:37:51.530 --> 00:37:53.770
<v Nick Leeder>You know, what's the one critical problem?

553
00:37:55.199 --> 00:37:57.600
<v Nick Leeder>And then could AI help us?

554
00:37:58.610 --> 00:38:05.169
<v Nick Leeder>So value comes from solving um for impact,
solving for an

555
00:38:05.169 --> 00:38:05.850
<v Nick Leeder>outcome.

556
00:38:06.739 --> 00:38:11.870
<v Nick Leeder>That gets everybody bought in, other
people so convinced that AI can be used and

557
00:38:11.870 --> 00:38:16.709
<v Nick Leeder>adds value to them because obviously
everybody I guess will start from a

558
00:38:16.709 --> 00:38:20.610
<v Nick Leeder>misconception that AI is a risk to them as
opposed to an opportunity.

559
00:38:21.250 --> 00:38:24.989
<v Nick Leeder>It solves the bit, you know, the financial
question and the CFO question that you

560
00:38:24.989 --> 00:38:25.610
<v Nick Leeder>need to answer.

561
00:38:26.500 --> 00:38:32.179
<v Nick Leeder>And you move from actually solving or you
solve a real problem and not showcasing

562
00:38:32.179 --> 00:38:33.639
<v Nick Leeder>innovation and technology.

563
00:38:34.719 --> 00:38:41.479
<v Nick Leeder>Um And make sure that that real problem is
not a point and

564
00:38:41.479 --> 00:38:42.560
<v Nick Leeder>isolated.

565
00:38:43.449 --> 00:38:49.790
<v Nick Leeder>Thing, it's something that is replicatable
across the facility, you know, you've got

566
00:38:49.969 --> 00:38:54.409
<v Nick Leeder>5 of the machines of the same type that
occasionally have those issues, or you've

567
00:38:54.409 --> 00:38:57.010
<v Nick Leeder>got 3 other factories that have the same
equipment.

568
00:38:58.010 --> 00:39:04.129
<v Nick Leeder>So start where the pain is, but make sure
that that pain is felt in many places

569
00:39:04.129 --> 00:39:08.929
<v Nick Leeder>because if you don't have that, then your
pilot or your proof point, however

570
00:39:08.929 --> 00:39:13.370
<v Nick Leeder>valuable, is only ever gonna be applicable
there and will never scale and the impact

571
00:39:13.370 --> 00:39:14.149
<v Nick Leeder>will never increase.

572
00:39:17.350 --> 00:39:18.030
<v Niall Sullivan>Absolutely.

573
00:39:18.229 --> 00:39:21.120
<v Niall Sullivan>And scale is where the big success is,
right?

574
00:39:21.270 --> 00:39:24.429
<v Niall Sullivan>So that should be the, the aspiration, of
course, um.

575
00:39:25.340 --> 00:39:29.479
<v Niall Sullivan>But yeah, no, thank you so much, Nick, um,
really useful advice.

576
00:39:29.820 --> 00:39:33.659
<v Niall Sullivan>Um, we're gonna, um, I think we usually do
after your episodes, we'll link to your

577
00:39:33.659 --> 00:39:37.860
<v Niall Sullivan>website, um, and LinkedIn so people can
connect with you afterwards if they want to

578
00:39:37.860 --> 00:39:39.560
<v Niall Sullivan>discuss any of these topics further.

579
00:39:39.820 --> 00:39:43.540
<v Niall Sullivan>Um, I guess that covers everything from
your side in terms of where people can learn

580
00:39:43.540 --> 00:39:44.080
<v Niall Sullivan>about you.

581
00:39:45.300 --> 00:39:45.909
<v Nick Leeder>Absolutely.

582
00:39:46.370 --> 00:39:50.010
<v Nick Leeder>I think LinkedIn is the most things and
then obviously on my website we have to be

583
00:39:50.010 --> 00:39:54.570
<v Nick Leeder>on the Buzzwords blog which we probably
post once every 4 to 6 weeks, which

584
00:39:54.570 --> 00:39:58.649
<v Nick Leeder>generally covers um a key, some of these
topics and some of these points we talked

585
00:39:58.649 --> 00:39:59.310
<v Nick Leeder>about today.

586
00:39:59.929 --> 00:40:00.409
<v Niall Sullivan>Excellent.

587
00:40:00.600 --> 00:40:06.010
<v Niall Sullivan>So yeah, we'll we'll link to those, but
um, but for today thank you again, Nick for

588
00:40:06.010 --> 00:40:11.209
<v Niall Sullivan>joining us and sharing your expertise, um,
and thank you everyone for listening to

589
00:40:11.209 --> 00:40:12.989
<v Niall Sullivan>this episode and we'll see you on the next
one.

590
00:40:13.169 --> 00:40:14.110
<v Niall Sullivan>Thank you very much.

591
00:40:16.149 --> 00:40:21.080
<v Speaker 2>Please like and subscribe on all major
podcast channels and find out more about

592
00:40:21.080 --> 00:40:24.679
<v Speaker 2>Sensei predictive maintenance at
Siemens.com.
