Latest / Speaking of Supply Chain / The Power of Digital Twins

Transcript

The Power of Digital Twins · Speaking of Supply Chain

0:06
You're listening to Speaking of Supply Chain, a meboch podcast. This is a
0:10
show for logistics professionals looking to learn more about the latest innovations in supply chain.
0:16
Each episode will feature a conversation on topics such as mitigating supply chain disruption
0:20
and reducing risk, current automation trends, sustainability initiatives, and more. Let's
0:26
dive right in. Hello, and welcome to Speaking of Supply Chain, where
0:33
we explore trends, current events, and innovations impacting the logistics and supply chain
0:38
industries. I'm your host, Ellen Wood. Today we are going to be
0:42
talking about digital twin technology and some different things that it's able to accomplish for
0:47
end users in terms of productivity, in terms of sustainability metrics, in terms
0:54
of resiliency factors. Digital twins are such a powerful tool. And joining us
1:00
today we have Marco Moore joining us from Sony Aroco. Am I pronouncing that
1:06
correctly, yes, Saco and also Savier Farrees and he is with me about
1:11
consulting and we're going to talk about this project that we have done for son
1:15
Rocco. And welcome both of you. Thank you, thank you very much
1:19
for this opportunity of sharing our project. It was a very interesting project that
1:23
we did with meback and with Chavvy, and we are happy to share with
1:30
other people what we did. Okay, So yeah, I'm feel that we
1:34
have Marco here. So with Marco, we have worked together for quite a
1:38
few years, so we have done quite a few projects. So it's a
1:42
pleasure always working with Marco. So it's a great opportunity and I feel very
1:47
happy to have Marco here and tell us you know how it is to work
1:49
with me back, and also all the things that he's doing and all the
1:53
interesting projects his his carry on. Yes, So Digital Twins are such an
2:00
exciting development in the supply chain technology area, and I know that our listeners
2:05
are really excited to hear about this particular project. So Marco, give us
2:10
a little bit of background into your career in supply chain, a little bit
2:16
about the company. Just give us a little background so that we can understand
2:22
how Digital Twin came into the conversation. Well, I am working in supply
2:28
chain since many years and also with automatic warehouses and AGVs, and I started
2:32
more from the outsourcing and purchasing than I advanced it to planning and logistics and
2:39
all of that. And in Sunyer Alco, we had this project where we
2:45
had several let's say issues in our flow. It is not a easy flow
2:50
in the plant of material flow, and it is difficult to understand where the
2:53
problems start, you know, depending on what we are producing. In a
2:58
certain moment, you have a problem, but then you change product that you
3:00
start having a problem in another place. So then came the idea of asking
3:06
NVAK to help us to create this digital twin to understand all the issues we
3:10
were having in our material slope. Okay, so what materials do you make?
3:15
We produced mainly melamines and what we call NFC, but also we also
3:21
produced the rollboards that then not surface it to create these decreative panels. Our
3:27
main objective is to increase the production that you surface it, so we upgrading
3:32
the panels. We also sell some products rowboards Plects said, we call it,
3:38
but our main objective is to go to the high value products. So
3:43
all of these these wood panels and there it's not just the raw wood boards.
3:47
It is composite panels that are assembled and produced and built together. So
3:53
what were some of the bottlenecks that you were facing. That led you to
3:58
think digital twin is going to be an effective tool to resolve these issues.
4:04
Yes, our process, actually we don't have a saw meal or something.
4:08
Right, the saw meals for us is like a provider because we use sawdust,
4:12
we use recycles. Would we have also our own recycle centers to receive
4:17
the raw materials. So in our process, the part of the production of
4:24
the robots is like say, a continuous process, so that is not very
4:29
difficult. Let's say the process itself. But when we start upgrading the boards,
4:34
with the surfacing and with the different finishes that we can have a different
4:39
flows inside of the plants, if the panels are catted or not, and
4:44
the kind of papers we are using over these rowboards, and then the positions
4:50
that we have inside of our cranes and things like that. This is where
4:54
things start to be more complex and the difficulty comes when we have, for
5:00
example, a different number of boards to produce package in a different number of
5:06
boards per pack also and things like that. So in certain moments, maybe
5:11
your bottle, your bottleneck starts to be in the packaging line, but in
5:15
other moments can be for example in the crane because the number of positions and
5:18
the number of aspus you are moving at that specific moment are higher. So
5:25
we usually what we do is we try to assimilate it with Excel files,
5:30
but we cannot have the same level of complexity at the same time. So
5:32
the digital quin will bring this ability to test different scenarios where we can plan
5:40
in different ways, and this is one some of the things we did with
5:43
digital QWIN is to test these different strategies strategy where we produce with more frequency,
5:49
with different plot sizes and using different flows in the plant. And also
5:56
another topic that came later is we wanted also to what kind of investments we
6:01
will do to overcome these battle x and then when we start having ideas of
6:08
the new investments that we will need, these investments usually are very high CAPEX
6:14
investments and it's important to test it well before we start really doing it.
6:18
So in the past, I imagine this happens also in other companies. We
6:23
made some mistakes, okay, because we evaluated the situation not with the digitalt
6:29
in with Excel file and in average scenes that everything is going well, but
6:33
then we start seeing or that we had too much capacity in one line and
6:38
maybe not enough in the packaging line or whatever, and then we need to
6:42
make the adjustments, but sometimes themselves costing a lot of money. So making
6:47
mistakes in these kind of investments is also very expensive. So the digitalt WEN
6:54
also helped us to mitigate a lot of a lot of these risks of investments.
7:00
Let me put a little bit of background about what Mark is telling me
7:02
that the panel would. Of course, this is a high volume product,
7:06
and if you go to a site of Sonar alco of your industry, is
7:11
a huge site where a bunch of trucks are coming in bringing some material raw
7:16
materials, and a lot of trugs are coming out with a finished product,
7:20
and everything is huge. So you have a lot of volumes, huge pieces
7:26
of equipment which are doing different parts of the process, and you have teams
7:30
making decisions from one part of the equipment doing the planning, and one kilometer
7:33
away you have another bunch of people making decisions about another part of the equipment,
7:39
and it's everything related. The decisions that one team is making, it's
7:42
going to affect to the next one and for them, it's very difficult to
7:46
have this visibility because it's very complex. Each one is focusing one part of
7:50
the process and to bring all the complexity into one unique view, you need
7:57
something else that what is Marco is is explaining it's not possible to do within
8:01
an Excel value need this Digital Twin technology so you can really simulate and assess
8:07
that all the decision make one who's going to walk through all along the process.
8:11
Yeah, I was going to say, Marco, you used the word
8:13
mistake or and I don't think it's a mistake because at the time and historically
8:18
Excel was what we had. Now when Digital Twin came into the picture,
8:24
then we were able to look at things differently, We were able to be
8:28
a little bit more proactive and to simulate those those scenarios. So it's not
8:33
that you were making a mistake with investments previously. It's that was all the
8:39
visibility that we had in supply chain. So it's not a mistake, it's
8:43
just that's the way it was. And now we've improved upon that process.
8:48
We have more capabilities. Now we're able to look at the data differently and
8:52
we're able to streamline those processes and really drill down and understand where some of
8:58
those bottlenecks are taking place and figure out what's going to work instead of guessing.
9:03
And I think that's the key that we're not guessing anymore with an excel.
9:09
You're you're limited on the amount of data that you're going to get,
9:13
but you don't have to take a guess anymore. You can put that scenario
9:16
into the Digital Twin and test it and see, okay, if we if
9:20
we do X, then Y, if we do Z then A. So
9:26
that I think is huge for companies, especially like yours, when the process
9:31
is so complicated. We're talking about boards that are manufactured from raw products that
9:37
are that are not simple. We call the decorative finalsies melamine finals yet yes,
9:43
the melamine panels. So you know, all of that is a long,
9:48
complicated, detailed process that if something should happen at any point in that
9:52
process, that's a bottleneck. That's a big deal in terms of production and
9:58
keeping things on schedules. So this is an excellent place where Digital Twin is
10:03
able to keep you from having to make those guesses like we historically have had
10:07
to do. So from that operational perspective, we understand a little bit more
10:13
about maybe not the entire process, but just that it is complicated and what
10:18
were some of those efficiencies that you were able to uncover during this project.
10:24
Were you able to find something in terms of the production, was it something
10:28
in terms of customer service times? What were you able to accomplish? Yes,
10:33
and this project had several points that we could improve. The main point
10:39
was service. We were trying to be more reliable to our customers. We
10:43
also improved productivity because then we were planning it with more certain about exactly what
10:50
we could do, so we ad less rework what also was making us loose
10:54
time, and of course we also had less emissions and all of that because
11:00
sequently if you start to be more efficient in the process, you also have
11:03
less consumption of energy and raw materials and everything. So yeah, it was
11:07
a very successful project for us and with very good impacts in the group,
11:13
so obviously the project was successful. Did you learn anything interesting throughout the process?
11:20
Did you find any surprises as you were going through the data as you
11:24
were running scenarios. Yes, we when we started running the first scenarios,
11:31
we found certain things that we didn't model in our in our digital twin,
11:37
because our knowledge about the process is what we put it in the digital twin.
11:41
And then it was curious I think remembers. We discover that were some
11:46
people doing certain things in the execution that no one really knew about. Some
11:54
people knew about it, but you know, the small touches you have to
11:58
do in the process to make it slow, but that actually there is no
12:03
plan about it because no one really knows that someone is doing. So we
12:07
needed to collect this because also we're not like a kind of best practices.
12:13
But it was interesting because we were seeing that the model was like showing a
12:16
blockage and actually in real it was not happening. So we discover what is
12:22
happening here, So it was interesting, and then we collected this problem because
12:26
actually it was something we didn't want to do, but someone was doing and
12:31
saying like, well if I don't do that, this doesn't work, and
12:33
we discovered and then we took actions around that. So also an important point.
12:39
It was also very interesting some things that came out of this project that
12:43
we didn't expect. It is the simulation itself, and to be also so
12:48
visual. It was interesting also to show the issues that we had to our
12:54
colleagues or even to the board members, because they don't really need to understand
13:00
spend everything about supply chain, but they could see how the material flows were
13:05
like blocking in certain points of the flow, and it was very visual for
13:09
them to understand the need of improvement or the need for our new investments.
13:13
So it was also one of the ways we use the digital twin needs to
13:18
show with more visual to other stakeholders why we are doing these investments, why
13:26
we need to change something. So I remember, on top of the of
13:31
managing the data and understanding the problem, I think also the digital twinners help
13:35
us all the change management and Marcos showing the different stakeholders that you can do
13:41
things differently. Probably you can explain better. But yes, one of the
13:45
things that were worked very well is at the beginning we were using real data,
13:50
real orders from a certain period that was representative, but we were not
13:56
touching the data, so we were using what it was there. Later on
14:01
when we were showing it to our colleagues in martin In Sales, we are
14:05
showing the issues that were being created in the flow and the need for investments
14:09
and the changes we needed to do. And then we also decided to make
14:15
some changes even in the kind of products we were setting. This was a
14:18
little bit on top of the project and we were showing this way it flows
14:24
if we don't produce these setain products. So they understood, and it was
14:30
also easier to explain them why in one situation things were not flowing as well
14:37
as when we don't produce certain products because our setup, our industrial setup was
14:41
not really prepared to do these kind of products. So it is not that
14:45
we cannot produce it, but if we can avoid to produce it for us
14:50
with the current setup, it is better. And everybody understood it, and
14:54
then people realize like, well, maybe I don't really need to sell it.
15:00
ABDC is just something that sometimes I can accept, but it's not really
15:03
that's important for our offer. So there were some changes on that side too
15:09
that made a lot of impact. So what I'm hearing is that, first
15:13
of all, when it came to the processes in the facility, and I'm
15:16
sure this is the case for a number of companies out there, that the
15:22
people on the floor are doing something deviating from the standard procedure, what the
15:26
procedure is meant to be. And in your case, it was good.
15:30
They were trying to eliminate problems and being very proactive about it. The problem
15:35
is that it wasn't documented and that causes problems a long way from that actual
15:43
situation, and these ripples have an effect. And this is a prime example
15:48
of that where when you're trying to do an overall efficiency you know, revamp
15:56
of what's going on and really see the process and drill down into those those
16:00
individual areas, and it's not happening the way that you intended it too.
16:03
So I think that's fantastic that you called that out to say, look,
16:07
yes, we're glad that you're you're proactive about doing this, but we really
16:11
need to know what is actually happening on the floor. But also the stakeholders.
16:17
This digital twin gives those stakeholders of you that they might not normally get
16:22
into the details of what's happening during that operational process. So that's that's really
16:30
a great great way to sell a digital twin to your your stakeholders, to
16:34
say, look, we need we need this visibility, we need to understand
16:37
what's happening, and we need to understand what's possible. I think that's the
16:42
bigger thing, is to see what's possible and have a better understanding of what
16:48
the results of each possibility are going to be. You know, when we're
16:51
looking at different scenarios, when we're saying we could do A, B or
16:53
C, what does it mean if we do A and having tangible results to
16:59
really explain it. And Savia, as you mentioned, the change management,
17:03
so not just to those stakeholders in the c suite or in the boardroom,
17:08
but also on the shop floor, those individuals who are changing things on their
17:12
own, trying to make things better and say, look, this is what's
17:17
going to happen if we do it. It's going to make your job easier,
17:18
it's going to make your job more efficient, it's going to improve so
17:22
many things within this operation. Let's do it. So that gets that buy
17:27
in for the change management. I think these are some great points in favor
17:33
of digital twins as as an option when you're looking to find these efficiencies within
17:37
your operations. Do you agree, yes for sure? And it is it
17:44
was something we learn because now we want to do more digital twins and we
17:48
are also doing already some projects. Me back and the Ecoloa tension that we
17:53
have to be very careful that that is nothing outside of the model that they
17:59
are doing and that it will affect. But also the opposite, right,
18:03
we can test something showing that maybe certain process don't need to work in the
18:07
way they think in the shop floor or in their own processes. So this
18:15
is really powerful when we test it and we can test different alternatives and checking
18:22
if the flows should be in that way or not. But with real data,
18:25
it is feel powerful, all right, Savvy. Any closing remarks from
18:29
you on the effectiveness or the value that a digital twin can bring to a
18:33
production client, you know, any any other types of industries out there.
18:37
I know we're talking about you know, production and wood products today, which
18:41
is you know, a particularly complex process, but a digital twin is going
18:47
to help any number of companies and any number of industries. Any other remarks,
18:52
Well, I think Marco has put it very well. It is a
18:56
digital twin. You can't reflect very precise what's going on in reality, but
19:02
also you can start to simulate things that are not happening in reality, so
19:06
you can kind of start to have imagination. Of course, better to have
19:10
a backup detailed imagination, So testing things that you think they are going to
19:15
work, analyzing and together working with customers because in all the process very well
19:19
with potential solutions they might work and then together you can test it again to
19:25
the city and into the digital twin and to see if this potential solution is
19:30
going to improve corrent operations. So I think it's a very complete kind of
19:36
project because you have data analysis, business intelligence, and you have algorithms artificial
19:42
intelligence, then you have a digital twin and on top of that you have
19:45
all the change management with one cold. We have been working in a very
19:49
intense project, very full of rewards at the same time. But I think
19:53
it's a very interesting and very complete project and to the with customers. Marco,
20:00
I know you just said that you were looking at doing more digital twin
20:03
projects with other locations, and is that going to help with just overall supply
20:11
chain resiliency, you know, across your organization because having that that level of
20:18
granularity and that that level of visibility into everything, I would think, you
20:22
know, would enhance that. Yes, that's that's our main idea for the
20:27
future is now we are doing it more plant by plant, trying to optimize
20:33
that plant, also shaking new investments. So MEVAC is helping us already in
20:37
other plants making the twin not only because of to fix the current issues,
20:45
because we always can optimize, but also looking forward to new investments that we
20:51
are doing. But to what I expect in the future is that we can
20:55
connect these digital twins from different plants because we have inten dependencies, like we
21:00
are producing certain boards that are going to other to other plants, and also
21:06
some products can be produced in two different plants. So having this digital to
21:11
could help us to simulate the vast distribution of capacity in certain moments, also
21:18
checking different scenarios when something happens, for example, to check the resiliency of
21:22
our plants. If we have an issue in one place, what will happen
21:26
to the rest of the capacity if we can really adapt or not. So
21:30
in the future I see it, Yes, a complete digital twin that could
21:36
reflect our supply shine. Now we are going plant by plant, but in
21:40
the future I would like to connect these digital twins. Well, this is
21:44
a piece of agreement that sounds very nice. Yeah, it is for the
21:51
future. We have different levels of generality, but we are also simulating already
21:56
with optimizer for example as our snop. Of course, when we go to
22:00
the plant level, we have a less detail than what we have in digital
22:03
twins. So of course, looking to the future, it would be good
22:08
if we could connect it really with our snop our vision of the future when
22:14
we receive a forecast. You know, today we are using data in a
22:18
different level of detail. Fantastic. I love the forward facing approach. Let's
22:23
look to the future and see what we're able to accomplish, because we're able
22:26
to accomplish a lot now, but it's only going to get better exactly.
22:33
Well, thank you so much Savier and Marco for joining me today and telling
22:37
me a little bit more about this story. I'm sure that our listeners have
22:40
enjoyed learning a little bit more about what digital twins are able to accomplish.
22:45
That brings us to the end of our episode. If our listeners out there
22:48
enjoyed it, please rate and subscribe in your podcast app, whatever app that
22:53
you're listening to. If you have a suggestion for a topic or would like
22:57
to be a guest on our show, we'd love to hear from you.
23:00
You can reach me at podcast at mebok dot com. As always, thank
23:04
you for listening to speaking of supply Chain, and be sure and tune in
23:07
next time. Mebok Consulting is one of the largest and most globally recognized supply
23:17
chain consulting, engineering, and advisory firms. For nearly fifty years. We've
23:22
helped clients achieved supply chain excellence and sustainable competitive advantage across the entire spectrum of
23:27
the supply chain by delivering improvements and innovation strategically, tactically, and digitally.
23:34
To learn more, visit meebok dot com. You've been listening to Speaking of
23:38
Supply Chain, a Mebok podcast. Keep connected with us by subscribing to the
23:44
show in your favorite podcast player. If you like what you've heard, please
23:47
rate the show that helps us to keep delivering the latest in supply chain information.
23:52
Thanks for listening. Until next time,