WEBVTT
Language: en-US

1
00:00:09.340 --> 00:00:10.890
<v Speaker 1>Welcome back to the Future Ready podcast.

2
00:00:11.029 --> 00:00:14.750
<v Speaker 1>I'm your host, Connor Pike, and joining me
again today are two of our AI experts

3
00:00:14.750 --> 00:00:17.319
<v Speaker 1>from Siemens, Samir Desai and Mathias
Lasko.

4
00:00:18.110 --> 00:00:21.719
<v Speaker 1>Last time we left off talking about
Siemens' broader AI approach, and when it comes

5
00:00:21.719 --> 00:00:25.430
<v Speaker 1>to AI, the importance of having the right
tool for the right job rather than a

6
00:00:25.430 --> 00:00:26.389
<v Speaker 1>blanket approach.

7
00:00:26.750 --> 00:00:28.190
<v Speaker 1>Samir, can we pick things up from there?

8
00:00:28.510 --> 00:00:31.709
<v Speaker 2>Yeah, and, and you brought up a very good
point, Connor, there, the right tool,

9
00:00:31.819 --> 00:00:31.989
<v Speaker 2>right?

10
00:00:32.229 --> 00:00:36.790
<v Speaker 2>And, and that is something that we call as
part of our three-pronged strategy.

11
00:00:36.869 --> 00:00:38.419
<v Speaker 2>One of the prongs is ancient.

12
00:00:38.909 --> 00:00:45.270
<v Speaker 2>is what we call engineering AI or in-app
AI, which is the users are using the

13
00:00:45.270 --> 00:00:46.310
<v Speaker 2>software products.

14
00:00:47.240 --> 00:00:48.139
<v Speaker 2>For years.

15
00:00:48.639 --> 00:00:53.279
<v Speaker 2>Each one of these products is designed to
do something special, right?

16
00:00:53.529 --> 00:00:59.520
<v Speaker 2>And as the users are performing their task
in that particular tool, in the

17
00:00:59.520 --> 00:01:04.080
<v Speaker 2>particular context of everything that is
around it, including the data, that is

18
00:01:04.080 --> 00:01:08.500
<v Speaker 2>where we interject AI and say, we
understand what you're working on.

19
00:01:09.000 --> 00:01:13.050
<v Speaker 2>Do you have a 3D geometry or you have a
simulation, or do you have a change, or do

20
00:01:13.050 --> 00:01:15.559
<v Speaker 2>you have a manufacturing process plan that
you're working on?

21
00:01:16.029 --> 00:01:22.230
<v Speaker 2>Our tools understand that, and then we
bring in AI to assist the user in doing the

22
00:01:22.230 --> 00:01:27.190
<v Speaker 2>next step better, in predicting the next
step that they're to perform, in predicting

23
00:01:27.190 --> 00:01:30.470
<v Speaker 2>how the geometry is going to evolve, how
the change needs to happen.

24
00:01:30.550 --> 00:01:34.910
<v Speaker 2>So, it's really bringing the AI to the
user rather than saying the user, you go do

25
00:01:34.910 --> 00:01:37.769
<v Speaker 2>over here and go figure out how you're
gonna do AI, right?

26
00:01:37.870 --> 00:01:39.610
<v Speaker 2>So that's, that's a big thing that we do.

27
00:01:41.370 --> 00:01:46.440
<v Speaker 1>So then thinking about, um, it's a really,
I think that's an excellent kind of look

28
00:01:46.440 --> 00:01:50.639
<v Speaker 1>into how the AI is influencing the
engineering process, right, from, you know,

29
00:01:50.889 --> 00:01:55.650
<v Speaker 1>moving from design through to um Um, as
we're approaching more automated production

30
00:01:55.650 --> 00:02:01.029
<v Speaker 1>as well, um, AI managing the factory floor
perhaps a bit more, so.

31
00:02:02.300 --> 00:02:06.010
<v Speaker 1>Matthias, how do you think, or how do you
see, perhaps is a better, a better

32
00:02:06.010 --> 00:02:09.779
<v Speaker 1>question, how do you see companies
actually deploying AI on factory floors?

33
00:02:09.899 --> 00:02:15.940
<v Speaker 1>Um, what sorts of, uh, systems are they
putting in place to, to automate their

34
00:02:15.940 --> 00:02:17.210
<v Speaker 1>production a bit, a bit more?

35
00:02:19.029 --> 00:02:21.699
<v Speaker 3>Yeah, systems based on all the different
areas that we touched on, so a lot of

36
00:02:21.699 --> 00:02:24.350
<v Speaker 3>predictive AI but also generative er
approaches.

37
00:02:24.500 --> 00:02:30.369
<v Speaker 3>The first ones even workflows now, um, and
very different areas of applications.

38
00:02:30.500 --> 00:02:35.820
<v Speaker 3>So Very famous and popular one in in
industries is obviously visual quality

39
00:02:35.820 --> 00:02:41.660
<v Speaker 3>inspections, or using machine vision um
applications in industries to solve the

40
00:02:41.660 --> 00:02:46.380
<v Speaker 3>problem of inspecting products or
intermediate products if there are defects, simply

41
00:02:46.380 --> 00:02:46.699
<v Speaker 3>speaking.

42
00:02:46.899 --> 00:02:50.250
<v Speaker 3>So usually you have a camera and some
products running over a conveyor belt, for

43
00:02:50.250 --> 00:02:53.139
<v Speaker 3>example, and you want to automatically
inspect those.

44
00:02:53.520 --> 00:02:56.679
<v Speaker 3>That's still a manual task in many
applications today.

45
00:02:56.800 --> 00:03:01.360
<v Speaker 3>So you have human operators standing there
all day long, inspecting the products on

46
00:03:01.360 --> 00:03:05.750
<v Speaker 3>the conveyor belt and picking the ones
that are not OK, so to say.

47
00:03:06.759 --> 00:03:10.270
<v Speaker 3>And that's what we can clearly automate to
a large extent today with AI

48
00:03:10.270 --> 00:03:13.720
<v Speaker 3>technologies, and we do that with
different products actually that we have.

49
00:03:13.800 --> 00:03:17.660
<v Speaker 3>We have like Inspector, which is an out of
the box quality inspection system.

50
00:03:18.399 --> 00:03:24.919
<v Speaker 3>For defection, detection of defects,
surface defects, for example, or um, detection

51
00:03:24.919 --> 00:03:27.669
<v Speaker 3>of foreign objects or missing objects.

52
00:03:28.080 --> 00:03:33.399
<v Speaker 3>So the more simple use cases works fine
for metals, electronics, plastics, for

53
00:03:33.399 --> 00:03:38.440
<v Speaker 3>example, and there are extremely complex
quality inspection use cases in consumer

54
00:03:38.440 --> 00:03:42.279
<v Speaker 3>packaged goods or food and beverage, for
example, where you have thousands of pieces

55
00:03:42.279 --> 00:03:46.759
<v Speaker 3>running over the conveyor extremely fast
and you still want to inspect those.

56
00:03:47.720 --> 00:03:51.440
<v Speaker 3>And there we have another product which is
called visual inspection cockpit.

57
00:03:51.770 --> 00:03:54.449
<v Speaker 3>So this is for, we called it closed loop
AI.

58
00:03:54.529 --> 00:03:59.789
<v Speaker 3>So it inspects the product, it detects a
defect, and it sends a signal to our PLCs.

59
00:04:00.289 --> 00:04:04.410
<v Speaker 3>So the control of the machine of the
plant, and then this one gives a trigger to

60
00:04:04.410 --> 00:04:06.669
<v Speaker 3>reject the products automatically.

61
00:04:07.330 --> 00:04:10.800
<v Speaker 3>And, and all of this in real time, so to
say, and this is really cool.

62
00:04:10.850 --> 00:04:13.210
<v Speaker 3>This is highly embedded into the
production line.

63
00:04:14.199 --> 00:04:18.799
<v Speaker 3>So quality inspection, besides that
predictive maintenance, obviously, so

64
00:04:18.799 --> 00:04:23.299
<v Speaker 3>forecasting if there is a defect of a
machine, if a breakdown could happen soon.

65
00:04:23.970 --> 00:04:28.679
<v Speaker 3>Um, actually avoiding that breakdown,
sending a maintenance engineer in time, that's

66
00:04:28.679 --> 00:04:31.720
<v Speaker 3>predictive maintenance where we have
SenseI as our product.

67
00:04:32.369 --> 00:04:37.500
<v Speaker 3>And if you look at process industries, um,
like chemical production, for example,

68
00:04:38.000 --> 00:04:39.559
<v Speaker 3>uh, we have advanced process control.

69
00:04:39.640 --> 00:04:44.679
<v Speaker 3>So this really goes down into the
parameter optimization of your control program,

70
00:04:44.839 --> 00:04:49.839
<v Speaker 3>tuning the parameters on the fly based on
the values of the process.

71
00:04:50.410 --> 00:04:55.269
<v Speaker 3>So this is also deeply embedded into um
the the production process itself.

72
00:04:56.089 --> 00:05:00.649
<v Speaker 3>We call it actually physical AI if we have
this close integration into the

73
00:05:00.649 --> 00:05:05.570
<v Speaker 3>industrial environment, if we actually act
also, we do not only perceive reason, but

74
00:05:05.570 --> 00:05:08.489
<v Speaker 3>we also take action automatically or
autonomously.

75
00:05:08.989 --> 00:05:12.989
<v Speaker 3>This is what we call physical AI, maybe to
add another category to predictive

76
00:05:12.989 --> 00:05:14.369
<v Speaker 3>generative, gentic.

77
00:05:15.149 --> 00:05:20.070
<v Speaker 3>There is also physical AI not only
connected to autonomous robots and and and those

78
00:05:20.070 --> 00:05:25.109
<v Speaker 3>aspects, but also tightly embedded into
control system of a production line.

79
00:05:25.670 --> 00:05:29.299
<v Speaker 3>And actually I would say all of this is
industrial AI, how we understand it.

80
00:05:29.589 --> 00:05:33.859
<v Speaker 3>It doesn't matter which flavor of AI it
is, all of this is under the bracket of

81
00:05:33.859 --> 00:05:34.380
<v Speaker 3>industrial.

82
00:05:36.329 --> 00:05:36.679
<v Speaker 3>Awesome

83
00:05:37.529 --> 00:05:42.940
<v Speaker 1>And, you know, one thing that um
Obviously, we love to talk about technology and,

84
00:05:42.980 --> 00:05:45.670
<v Speaker 1>and, and all the exciting things that we
might be able to do with it.

85
00:05:46.309 --> 00:05:49.970
<v Speaker 1>Um, I think one of the practical realities
maybe for customers is the thought of

86
00:05:50.109 --> 00:05:55.350
<v Speaker 1>adopting a new tool, and, you know, maybe
they already have other tools in place,

87
00:05:55.399 --> 00:05:59.429
<v Speaker 1>and what they don't want to do, presumably
is totally replace their entire

88
00:05:59.429 --> 00:06:00.329
<v Speaker 1>ecosystem.

89
00:06:00.739 --> 00:06:05.670
<v Speaker 1>Um, so how can we work with, with what
customers already have in place and build on

90
00:06:05.670 --> 00:06:06.230
<v Speaker 1>top of that?

91
00:06:07.089 --> 00:06:11.350
<v Speaker 2>It's a wonderful question, because we are
already doing this with our customers,

92
00:06:11.429 --> 00:06:11.640
<v Speaker 2>right?

93
00:06:11.790 --> 00:06:17.500
<v Speaker 2>So, um, I can do 2 or 3 different variety
of examples, but we'll give you the value

94
00:06:17.799 --> 00:06:21.559
<v Speaker 2>of how customers are thinking about this
and how we are partnering with our

95
00:06:21.559 --> 00:06:22.660
<v Speaker 2>customers for this.

96
00:06:23.000 --> 00:06:28.600
<v Speaker 2>So if you take an example of our PLM
system, Team Center, that's the #1 PLM system

97
00:06:28.600 --> 00:06:29.459
<v Speaker 2>in the world today.

98
00:06:29.940 --> 00:06:34.420
<v Speaker 2>We have a customer who's been using it for
over 10 years, right?

99
00:06:34.730 --> 00:06:40.559
<v Speaker 2>And then we added certain elements of AI
in there, and they said, we would like to

100
00:06:40.559 --> 00:06:41.320
<v Speaker 2>try this thing out.

101
00:06:41.440 --> 00:06:42.299
<v Speaker 2>So what did we do?

102
00:06:43.320 --> 00:06:47.549
<v Speaker 2>We provided this capability to them in the
Teamcenter product, but we partnered, we

103
00:06:47.549 --> 00:06:50.100
<v Speaker 2>ran a set of proof of concepts with them.

104
00:06:50.549 --> 00:06:51.000
<v Speaker 2>They say, Mr.

105
00:06:51.010 --> 00:06:53.859
<v Speaker 2>customer, what are your top 3 use cases?

106
00:06:54.230 --> 00:07:00.019
<v Speaker 2>Let us run through those with you, with
you, not handed over to you, but with you,

107
00:07:00.519 --> 00:07:06.480
<v Speaker 2>and show to you how this is gonna add
value and bring more importance to your end

108
00:07:06.480 --> 00:07:07.079
<v Speaker 2>users.

109
00:07:08.070 --> 00:07:13.510
<v Speaker 2>And they loved it, and at the end of it,
they bought 2500 seats of that software

110
00:07:13.510 --> 00:07:14.899
<v Speaker 2>offering from us, right?

111
00:07:15.109 --> 00:07:16.140
<v Speaker 2>So, so that's one.

112
00:07:16.510 --> 00:07:21.769
<v Speaker 2>The other example is look at completely
different industry, chip design.

113
00:07:23.149 --> 00:07:29.910
<v Speaker 2>ARM, which is a chip design company, um,
they have to work under 6 Sigma quality.

114
00:07:30.730 --> 00:07:33.940
<v Speaker 2>Which means that when they design chips
and when they go to manufacturing, the

115
00:07:33.940 --> 00:07:36.980
<v Speaker 2>manufacturers, only 1 in a billion chips
can fail.

116
00:07:37.929 --> 00:07:39.769
<v Speaker 2>1 in a billion, 6 sigma.

117
00:07:40.209 --> 00:07:44.910
<v Speaker 2>To achieve that, they have to run a lot of
simulations, and they were running about

118
00:07:44.910 --> 00:07:46.410
<v Speaker 2>5 billion simulations.

119
00:07:47.709 --> 00:07:49.649
<v Speaker 2>Over a period of months.

120
00:07:50.940 --> 00:07:56.820
<v Speaker 2>What we did is with our one of our IC chip
design verification software, Solio,

121
00:07:57.350 --> 00:08:01.989
<v Speaker 2>we introduced a generative or a predictive
model in there.

122
00:08:02.700 --> 00:08:04.899
<v Speaker 2>And they started using this predictive
model.

123
00:08:05.649 --> 00:08:11.190
<v Speaker 2>And instead of having to do 5 billion
simulations, they're down to 2000.

124
00:08:11.730 --> 00:08:16.529
<v Speaker 2>That's, that's a magnitude of 2500 x 2.

125
00:08:16.690 --> 00:08:17.760
<v Speaker 2>actually more than that.

126
00:08:18.250 --> 00:08:23.070
<v Speaker 2>So right, it's 2.5 million x improvement
that they've gotten out of it.

127
00:08:23.579 --> 00:08:27.369
<v Speaker 2>Um, and instead of months, they are able
to do that in hours.

128
00:08:27.679 --> 00:08:32.130
<v Speaker 2>So these are like some very real examples
of what we have done.

129
00:08:32.559 --> 00:08:33.429
<v Speaker 2>With our customers.

130
00:08:33.520 --> 00:08:34.979
<v Speaker 2>The last one with PepsiCo.

131
00:08:35.559 --> 00:08:41.549
<v Speaker 2>Industrial metaverse, we've partnered with
PepsiCo, as well as Nvidia, where we are

132
00:08:41.549 --> 00:08:45.200
<v Speaker 2>using Nvidia GPUs to do what we call as
digital twin composer.

133
00:08:45.929 --> 00:08:48.489
<v Speaker 2>And we built this together with PepsiCo.

134
00:08:48.570 --> 00:08:49.559
<v Speaker 2>So, what do you want to do?

135
00:08:49.650 --> 00:08:50.489
<v Speaker 2>What is your scenario?

136
00:08:50.599 --> 00:08:51.849
<v Speaker 2>What's your factory layout?

137
00:08:52.090 --> 00:08:55.409
<v Speaker 2>How, how are we gonna do this
visualization, and we're with you for that.

138
00:08:56.340 --> 00:08:58.340
<v Speaker 2>At the end of the project, they have it
deployed.

139
00:08:58.469 --> 00:09:00.030
<v Speaker 2>They're down here today.

140
00:09:01.239 --> 00:09:06.849
<v Speaker 2>And they have achieved up to 15% reduction
in their Capital expenditure.

141
00:09:07.130 --> 00:09:13.909
<v Speaker 2>So, real AI offerings with partners,
keeping it real, avoiding the hype,

142
00:09:14.609 --> 00:09:16.510
<v Speaker 2>and bringing it to the, to them.

143
00:09:16.650 --> 00:09:21.409
<v Speaker 2>So, that's how we work and take AI in with
our customers and with our partners.

144
00:09:21.770 --> 00:09:26.979
<v Speaker 1>Work with the systems they already have in
place in a Uh, production environment as

145
00:09:26.979 --> 00:09:27.090
<v Speaker 1>well.

146
00:09:27.140 --> 00:09:32.219
<v Speaker 1>I mean you have, I'm sure, a large mix of,
of brownfield facilities where customers

147
00:09:32.219 --> 00:09:36.059
<v Speaker 1>have machines of all ages, I imagine, um.

148
00:09:36.940 --> 00:09:39.419
<v Speaker 1>So, yeah, a similar question, I suppose.

149
00:09:39.500 --> 00:09:44.500
<v Speaker 1>How, how do you navigate maybe integrating
AI into these facilities where you might

150
00:09:44.500 --> 00:09:49.059
<v Speaker 1>have machines that are, I don't know,
maybe older than me in some cases.

151
00:09:50.349 --> 00:09:54.719
<v Speaker 3>Actually in a very similar manner, but
with with different technologies, obviously.

152
00:09:54.989 --> 00:09:58.950
<v Speaker 3>So it's, it's all about um data
connectivity in the first place.

153
00:09:59.030 --> 00:10:04.349
<v Speaker 3>So you really need to get the data out of
these um systems and and as you said, it's

154
00:10:04.349 --> 00:10:10.270
<v Speaker 3>usually a legacy system that is their
brownfield applications with um mixed vendors

155
00:10:10.270 --> 00:10:15.070
<v Speaker 3>with different generations of controls,
for example, or of tribes.

156
00:10:15.630 --> 00:10:18.320
<v Speaker 3>And we have to get the connectivity right
in the first place.

157
00:10:18.400 --> 00:10:22.150
<v Speaker 3>So that's the homework that every customer
has to do then, and we also help the

158
00:10:22.150 --> 00:10:26.320
<v Speaker 3>customers on this journey and with some
software components that we can offer to

159
00:10:26.320 --> 00:10:31.400
<v Speaker 3>make connectivity easy, to get the data
out of the systems, and then send this data

160
00:10:31.400 --> 00:10:35.400
<v Speaker 3>to usually a centralized environment like
based on our industrial edge platform, we

161
00:10:35.400 --> 00:10:37.739
<v Speaker 3>can contextualize this data.

162
00:10:38.299 --> 00:10:43.299
<v Speaker 3>Um, even sending it to our, um, Altech
craft studio, we, we will expand on this

163
00:10:43.299 --> 00:10:45.719
<v Speaker 3>later, I hope so I'm talking about the
data fabric.

164
00:10:46.099 --> 00:10:51.049
<v Speaker 3>Um, that's really key to bringing all
these data items together, getting it to a

165
00:10:51.049 --> 00:10:55.299
<v Speaker 3>certain semantic level, contextualizing
the data, and then you can put the AI, the

166
00:10:55.299 --> 00:10:56.599
<v Speaker 3>AI applications on top.

167
00:10:57.429 --> 00:11:03.750
<v Speaker 3>So this is also a journey for customers to
transform their operations into an AI

168
00:11:03.750 --> 00:11:09.869
<v Speaker 3>ready system to really leverage these um
these applications and we also recommend to

169
00:11:09.869 --> 00:11:16.390
<v Speaker 3>go for the the the high value use cases
first, um to to start with some pilots, to

170
00:11:16.390 --> 00:11:20.469
<v Speaker 3>get the breakthrough with those, but from
the beginning, standardizing the

171
00:11:20.469 --> 00:11:21.429
<v Speaker 3>infrastructure.

172
00:11:21.950 --> 00:11:26.710
<v Speaker 3>So not reinventing the connectivity layer
with every AI use case, not reinventing

173
00:11:26.710 --> 00:11:31.349
<v Speaker 3>the AI runtime, so the execution, the
inferencing of the AI models with every use

174
00:11:31.349 --> 00:11:33.419
<v Speaker 3>case, we saw that very often actually
happening.

175
00:11:33.789 --> 00:11:37.750
<v Speaker 3>So we highly recommend to standardize all
this on a joint software-based

176
00:11:37.750 --> 00:11:42.750
<v Speaker 3>infrastructure, and then you can go and
roll out these applications and put further

177
00:11:42.750 --> 00:11:44.450
<v Speaker 3>AI use cases on top of it.

178
00:11:46.090 --> 00:11:49.890
<v Speaker 1>You know, and you already, you brought up
the, the idea of the, the data fabric, and

179
00:11:49.890 --> 00:11:54.489
<v Speaker 1>I think, um, I, let's, let's jump into
that right away, cause I, I'm very interested

180
00:11:54.489 --> 00:12:01.409
<v Speaker 1>in, in the Yeah, the idea of how do you
connect all these different Uh, maybe silos

181
00:12:01.409 --> 00:12:08.119
<v Speaker 1>of data that exist within those companies,
um, and both how It, it seems like to

182
00:12:08.140 --> 00:12:12.900
<v Speaker 1>leverage AI you need to do that
connection, but perhaps AI can also help in creating

183
00:12:12.900 --> 00:12:15.140
<v Speaker 1>that data fabric and building those
connections as well.

184
00:12:15.299 --> 00:12:16.799
<v Speaker 1>Um, Samir, I don't know if you want to
start.

185
00:12:17.059 --> 00:12:17.169
<v Speaker 1>Yeah,

186
00:12:17.260 --> 00:12:20.780
<v Speaker 2>I, I can start and Matthias, please,
please add on to, on top of that, cause, cause

187
00:12:20.780 --> 00:12:25.020
<v Speaker 2>we both are focusing on that for some of
the, for the key elements of engineering as

188
00:12:25.020 --> 00:12:26.119
<v Speaker 2>well as automation, right?

189
00:12:26.179 --> 00:12:32.140
<v Speaker 2>And so what we typically see, have seen
over again 2 to 3 decades that our customers

190
00:12:32.140 --> 00:12:36.690
<v Speaker 2>have been using, and this is a problem of
what the customers call as Either data

191
00:12:36.690 --> 00:12:39.070
<v Speaker 2>silos or dark data.

192
00:12:39.650 --> 00:12:41.250
<v Speaker 2>What do they mean by that, right?

193
00:12:41.409 --> 00:12:46.940
<v Speaker 2>So, let's take an example that when our
One of our biggest industry is car

194
00:12:46.940 --> 00:12:48.349
<v Speaker 2>manufacturers, right?

195
00:12:49.700 --> 00:12:50.679
<v Speaker 2>They're doing their design.

196
00:12:50.780 --> 00:12:55.700
<v Speaker 2>They have their design information in our
PLM system Team Center, but when they look

197
00:12:55.700 --> 00:13:00.140
<v Speaker 2>at their vendor and sourcing information
that's sitting in ERP, they have their

198
00:13:00.140 --> 00:13:05.020
<v Speaker 2>customer relationship data in CRM, and
then they have a lot of homegrown systems

199
00:13:05.020 --> 00:13:06.400
<v Speaker 2>implementations as well.

200
00:13:06.979 --> 00:13:13.909
<v Speaker 2>Now, when they're building this car, It's
not just the design or the sourcing or

201
00:13:13.909 --> 00:13:17.700
<v Speaker 2>the in-house data that is by itself
important.

202
00:13:19.460 --> 00:13:25.609
<v Speaker 2>The connection of this data is more
important because when they feel, when they

203
00:13:25.609 --> 00:13:30.489
<v Speaker 2>think, hey, I'm gonna make this change in
my design data, it is gonna have an impact

204
00:13:30.489 --> 00:13:31.369
<v Speaker 2>on sourcing.

205
00:13:31.710 --> 00:13:35.390
<v Speaker 2>It is gonna have an impact on
manufacturing, which is where Matthias and and the

206
00:13:35.390 --> 00:13:36.969
<v Speaker 2>team comes in terms of automation.

207
00:13:37.229 --> 00:13:40.429
<v Speaker 2>It's gonna have impact on a multitude of
other things.

208
00:13:40.510 --> 00:13:42.190
<v Speaker 2>And how do you understand the impact?

209
00:13:42.590 --> 00:13:49.010
<v Speaker 2>Traditionally, You would have to get
people from these different disciplines, the

210
00:13:49.010 --> 00:13:52.369
<v Speaker 2>subject matter experts, they're gonna come
together, they're gonna look into their

211
00:13:52.369 --> 00:13:56.179
<v Speaker 2>systems, they're gonna try to figure out,
and it's, it's a long cycle.

212
00:13:56.809 --> 00:14:00.130
<v Speaker 2>Takes days, weeks, months, or sometimes,
right?

213
00:14:00.950 --> 00:14:04.679
<v Speaker 2>Wouldn't it be cool to really provide an
automation of this?

214
00:14:05.489 --> 00:14:10.570
<v Speaker 2>That it understands the relationship of
all this data, and based on that

215
00:14:10.570 --> 00:14:15.960
<v Speaker 2>understanding, you can run AI on top of
it, but you can run analytics on top of it,

216
00:14:15.969 --> 00:14:19.210
<v Speaker 2>and you can perform actions on top of it,
right?

217
00:14:19.570 --> 00:14:23.530
<v Speaker 2>This is where making it smarter becomes
critical.

218
00:14:23.650 --> 00:14:25.929
<v Speaker 2>This is what customers mean by dark data.

219
00:14:25.969 --> 00:14:28.330
<v Speaker 2>It's like, I know I have all this
information.

220
00:14:28.409 --> 00:14:31.809
<v Speaker 2>I know it's somehow connected, but I don't
know what the connection is.

221
00:14:32.219 --> 00:14:36.289
<v Speaker 2>And how could, so we're solving that
through what we call AI fabric, right?

222
00:14:36.380 --> 00:14:42.080
<v Speaker 2>And then, and the primary fundamental
offering there for us is based on our rapid

223
00:14:42.080 --> 00:14:43.099
<v Speaker 2>minor portfolio.

224
00:14:44.090 --> 00:14:48.299
<v Speaker 2>Rapid Minor Graph Studio specifically is
what we are uh using.

225
00:14:48.380 --> 00:14:53.219
<v Speaker 2>It's, it's a great piece of technology
that we have part of our Altair um

226
00:14:53.219 --> 00:14:55.679
<v Speaker 2>acquisition that we did just over a year
ago.

227
00:14:56.140 --> 00:14:59.739
<v Speaker 2>And, and what we are doing there is this
concept, I'm gonna get a little bit

228
00:14:59.739 --> 00:15:06.510
<v Speaker 2>technical here, but you know, uh, is the
concept of ontologies or semantic

229
00:15:06.510 --> 00:15:09.950
<v Speaker 2>relationship of information that is in
these different places, right?

230
00:15:10.020 --> 00:15:12.500
<v Speaker 2>So we are bringing Siemens.

231
00:15:13.260 --> 00:15:19.460
<v Speaker 2>Because of its domain expertise that we
have in all these domain areas, because of

232
00:15:19.460 --> 00:15:24.260
<v Speaker 2>how we understand the data works and it
connects, because of how we know data is

233
00:15:24.260 --> 00:15:28.760
<v Speaker 2>secure and security is more important to
our customers than anything else today,

234
00:15:28.940 --> 00:15:29.179
<v Speaker 2>right?

235
00:15:29.929 --> 00:15:36.289
<v Speaker 2>We bring all of these pieces together into
this ontological view that says, here's

236
00:15:36.289 --> 00:15:42.650
<v Speaker 2>how your design information is connected
to your supplier information, to your

237
00:15:42.650 --> 00:15:46.880
<v Speaker 2>manufacturing information, and oh, now you
can start asking questions.

238
00:15:47.289 --> 00:15:51.200
<v Speaker 2>Not only ask questions, but you can build
agents off of that ontological

239
00:15:51.200 --> 00:15:58.010
<v Speaker 2>representation using either our Mendex
loco tool or the rapid Minor tool.

240
00:15:58.280 --> 00:16:01.049
<v Speaker 2>Um, you can run analytics on top of it.

241
00:16:01.390 --> 00:16:07.409
<v Speaker 2>And once you've figured out what the
connections are, now you can say, hey, my this

242
00:16:07.669 --> 00:16:10.099
<v Speaker 2>piece of information in one system is
changing.

243
00:16:10.469 --> 00:16:13.690
<v Speaker 2>I need to effect a change in another
system.

244
00:16:14.150 --> 00:16:18.669
<v Speaker 2>You can orchestrate that because of the
way of how we've established this whole

245
00:16:18.669 --> 00:16:19.789
<v Speaker 2>ontology, right?

246
00:16:20.109 --> 00:16:22.150
<v Speaker 2>It's a very, very powerful offering.

247
00:16:22.559 --> 00:16:26.380
<v Speaker 2>Uh, it starts, and, and that's why when we
look at, it's not just engineering, it's,

248
00:16:26.409 --> 00:16:30.049
<v Speaker 2>it's in the manufacturing world, it's in
the software world too, and Ila Matthias,

249
00:16:30.090 --> 00:16:34.969
<v Speaker 2>if you, you want to talk about how, how
you're looking at this thing, um, in, in

250
00:16:34.969 --> 00:16:35.530
<v Speaker 2>your domain.

251
00:16:35.650 --> 00:16:37.289
<v Speaker 3>Yeah, so I have a very similar view.

252
00:16:37.400 --> 00:16:43.169
<v Speaker 3>So this, this proper, um, data foundation
is really key to scale AI and industries,

253
00:16:43.250 --> 00:16:43.469
<v Speaker 3>right?

254
00:16:43.609 --> 00:16:48.440
<v Speaker 3>And Also from an operations or um running
the factory perspective.

255
00:16:48.979 --> 00:16:52.619
<v Speaker 3>It's the same thing, you still have to
integrate the data across different data

256
00:16:52.619 --> 00:16:54.599
<v Speaker 3>silos that are still there very often.

257
00:16:55.260 --> 00:17:00.979
<v Speaker 3>And it's also um like you can, can also
give a feedback loop to the design and the

258
00:17:00.979 --> 00:17:02.539
<v Speaker 3>planning phase of the production actually.

259
00:17:02.659 --> 00:17:06.540
<v Speaker 3>So just imagine the example that I
mentioned before, the quality inspection.

260
00:17:06.619 --> 00:17:11.859
<v Speaker 3>So you detect a failure or a problem of
the product, and now this occurs over and

261
00:17:11.859 --> 00:17:12.660
<v Speaker 3>over again.

262
00:17:13.168 --> 00:17:18.009
<v Speaker 3>This usually means there's something wrong
in the process of the manufacturing plant

263
00:17:18.009 --> 00:17:20.568
<v Speaker 3>before this, this quality inspection step.

264
00:17:20.649 --> 00:17:26.918
<v Speaker 3>So something is um deteriorating or or
some, some process parameter is

265
00:17:26.918 --> 00:17:27.668
<v Speaker 3>drifting.

266
00:17:28.129 --> 00:17:32.529
<v Speaker 3>So usually this means you have to optimize
something and you have to feedback, um,

267
00:17:32.568 --> 00:17:35.649
<v Speaker 3>you have to give feedback to the design
and the dissimation and the planning part of

268
00:17:35.649 --> 00:17:36.649
<v Speaker 3>the whole production line.

269
00:17:36.968 --> 00:17:40.369
<v Speaker 3>So closing this loop across the whole
value chain.

270
00:17:41.280 --> 00:17:44.229
<v Speaker 3>This is where, where really a data fabric
is needed.

271
00:17:44.900 --> 00:17:47.579
<v Speaker 3>And I think it's a great tooling that we
have in the in the meantime here.

272
00:17:47.670 --> 00:17:52.219
<v Speaker 3>So I was always dreaming about this, um,
doing my PhD in in the field of knowledge

273
00:17:52.219 --> 00:17:55.640
<v Speaker 3>graphs actually, and so this is the
underlying technology that we use here.

274
00:17:56.380 --> 00:18:01.849
<v Speaker 3>So breaking up data silos, having a joint
ontology, we call that, so it's a joint

275
00:18:01.849 --> 00:18:06.000
<v Speaker 3>terminology to have the different items um
on a semantic level.

276
00:18:06.420 --> 00:18:10.420
<v Speaker 3>That's the right step to really break up
data silos and this is needed then for

277
00:18:10.420 --> 00:18:11.479
<v Speaker 3>agentic workflows.

278
00:18:11.839 --> 00:18:16.459
<v Speaker 3>That operate across different phases of
industrial production.

279
00:18:18.579 --> 00:18:23.540
<v Speaker 1>And so, maybe part of building that, that
foundation, it seems to me that the data

280
00:18:23.540 --> 00:18:29.380
<v Speaker 1>fabric in some ways is like providing a, a
really solid base level for AI to operate

281
00:18:29.380 --> 00:18:34.239
<v Speaker 1>from and for you to build, um, you know,
more valuable applications maybe on top of.

282
00:18:34.540 --> 00:18:37.209
<v Speaker 1>But we're gonna have to save the answer to
that question for next time.

283
00:18:37.380 --> 00:18:40.790
<v Speaker 1>Once again, I've been your host, Connor
Pike, joined by Samir Desai and Mathias Law

284
00:18:40.790 --> 00:18:42.160
<v Speaker 1>School on the Future Ready podcast.
