Do you desire a more joy-filled, deeply-enduring sense of accomplishment and success? Live your business the way you want to live with the BUSINESS BEATITUDES...The Bridge connecting sacrifice to success. YOU NEED THE BUSINESS BEATITUDES!
TAP INTO YOUR INDUSTRIAL SOUL, RESERVE YOUR COPY NOW! BE BOLD. BE BRAVE. DARE GREATLY AND CHANGE THE WORLD. GET THE BUSINESS BEATITUDES!
Industrial Talk is brought to you by Arduino the original all in one IoT platform. That's right, go out to arduino.cc. And you will find documentation you will find devices, powerful boards to help you with your digital transformation journey. It's all there, go out to arduino.cc. Find out more see how you can connect with these professionals to help you along with your digital transformation journey. arduino.cc. Also Industry IoT Consortium. At industrial talk, we always talk about education, we always talk about collaboration, we are always talking about innovation. And if you're a business that has any desire to be resilient to the future, you need to be able to educate, collaborate, as well as innovate with other industry professionals. That's a must. Industry IoT Consortium brings that all together, you need to be a part of this community, you need to be connected with these leaders that are all apart the Industry IoT Consortium, go out to iiconsortium.org. Find out more again, you will not be disappointed, you're just going to be happy.
Welcome to the industrial talk podcast with Scott Mackenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots, and let's go
Yes. Great. I'm digging it.
We're gonna go to a bar tonight. Just a bar. We're gonna go to a bar tonight. Of course. Why wouldn't we have a couple of gin and tonics? Yes. So it's a great, that's a great option. Sure. All right, for the listeners out there and talk to us a little bit about your background. Michel, and then we're gonna get into COSMOTE So what that's all about and then we're going to talk about what? What solutions are being provided.
Well, my background in in a nutshell, I used to be for many years now, in the past the university professor, I was a specialist of complex systems math, computer science. Hello,
are you a doctor? Doctor, I see have to see this card. If you're out there. This is the card he has right there. Right there. It's a sticky note. And there's no doctor. So you're a doctor good. I
used to be a processor. So you know, and, and so and then I developed, you know, Technology, my lab and I decided to bring it from academia to the real world. And that was what we call today's 360 degree simulation.
Cosmo Tech, you rattled off something that's an acronym, an acronym for something, what does it mean?
Sure. Oh, Cosmo means complex systems modeling. Say,
there it is. It's pretty good. Yeah. And you got the URL Cosmo Tech, because nobody had that. That's, that's a good one. So with that, give us a little sort of a definition of digital twin. And then let's get into the simulation component. If that's good, give us a what is a digital twin?
Sure. But digital digital twin is a digital replica of a system. Originally, it was a digital replica of physical objects. But now we are doing digital twins of systems of organizations. And by the way, that's why we use the word 360 degrees to say that we take, you know, and we present the systems, and we connect to the data in real time. And oh,
so, you know, everybody uses digital twin, I have a digital twin of my building, I have a digital twin of my substation, and so on and so forth. But what you're doing is you're sort of holistically looking at the whole, you know, whatever the business itself, it says,
exactly, we don't only look at the at the physical system, the physical parts, but also to all the interactions, the processes, the people who work with the system, what they do, and how they interact with the system.
So take us through what that looks like. I'm a company. I've got the I'm, I'm interested in digital twin, or I'm, I'm using a digital twin. Take us through that process.
Well assume that you have a supply chain to manage. And you know that there is a lot of uncertainty. And now more and more complexity in your supply chain. And you want to make that's
that's an understatement. And in fact, this pandemic, everybody's a supply chain expert.
Oh, that's right. And you have the problem is what you have to do with the supply chain, you have to reconcile things that seem to be completely impossible to reconcile. So you want to increase the profit, you want to be resilient, at the same time you want to reduce cost and reduce the co2 emissions. That's your problem, right? And all that changes all the time. So what do you need for being able to manage this and make the best possible decisions? You need to have three things? The first thing is what happened in the past? What can I learn from data from the past? The second thing is what is currently happening, you're looking at data from the present. But the most important thing you need is to have access to data from the future, what is going to happen? If I do this? What if what is going to happen if I do that, and what we do with digital twin, we create the replica, we link it to you know the data, and we have the ability and that's the simulation plant to simulate it to see what is going to happen in the future so that you can test 100,000 futures to find the one you want and the resulting one.
Do you want that future? Look, you have to load the data, some data, you have to load some historical data in some way to paint a possible picture in the future. And then do you? Does the simulation allow you to learn or what help us?
That's a very that's a very good question. Because we can use data from the past and we do. But what is very important in what we do is that what we put in the digital twin is the knowledge of the different experts. We don't only rely on data from the past, but we are in a system industry or system. You have experts who know very well what is happening on your live production line. What is happening with the robots, what is happening with logistics, what is happening with when people come to work, are they in three eight shifts or do the work? So this knowledge, we put the knowledge and we make the links with the knowledge and because we have all these? Once we have that we just have to connect to the state the current state of the system, and what do we do? We simulate So based on the knowledge, we can add knowledge that we have learned from data from the past, right? But there is the knowledge that comes from the brain of people. And it's
so deep. So this is cool. This is cool. And as you run multiple simulations, you're putting in different different type of data, different type of parameters, whatever it might be. And then you're is there. Is there a time frame of into the future? Like, can can the future be a year? Can the future be two years? What is sort of that ideal sort of future state?
There is no ideal we go from 15 minutes to 40 years, depending on the use case. 15 minutes? Yeah, we're doing definitely are sure. We Yeah, for example, assume that you have just a line, a production line. And you you're connected to the data, and you know that you, you're going to make a decision. And you have to go as fast as possible to make this decision. Because you see, for example, that a machine, the product productivity of a machine is dropping. So what is going to be the impact of this drop on your production, right? So you very rapidly simulate, just to know what's going to happen. And once you know that, what you do, you say, here is what I can do here is the different view of the different options. And now I'm going to simulate all these options to make the best possible decision. We do that timescale. And we do that up to 40 years, when we manage very, very big assets and equipments, for example, for maintenance and renewal.
Yeah, exactly. And for example, that's what we do. Mr. Big electric networks. Yes. Electric net? Yes. That's a great example. Yeah. And we are, for example, we have full digital twin of all the high voltage, I mean, transmission, electric utility of France, the whole country, 1 million assets, 4 million interactions. And we have all the teams with where they are their skills, their cost. And every year, the company of T spends 800 million for maintenance. And when you go down that road, yeah, you're absolutely right. And now they're running simulations all the time to find the best decision, because there are a lot of cascading effects. If you delay an investment, this is going to you're going to have more outages on this, then you're going to run the teams there. But the team will not be available to do something else. And maybe you can have a cascading effect. So yeah, do you do you
got a couple of questions? Do you simulate the need for the workforce to be educated? Like there's a challenge today that there is a future skill requirement? And if I don't do it here, I'm not gonna get it there. Sure. And by
the way, we are doing that very, very concrete example. The same the same problem than the one I described on the electric grid. Yeah, we you have that in factories, because you have robots, you have fleets of robots, and they're obsolescence. So they, they lost, and they they age, and at the moment, they don't work anymore. So you have to decide to replace them by new ones. But then I have to train the people to be able to manage these years, when am I going to train the people when I trade them, they are not productive. So it has an impact. So you have to optimize all the things. And that's what we do with the simulation digital twin, so different times, but then
again, at 360, that's where you have to a lot of people are sort of siloed in their digital twin approach or look, do you also in that utility analogy, upload historical maintenance records. And so you'll know that this transformer had the oil changed out and whatever it is, or it's X amount of years old, and that type
of an exec. So we have that for each equipment, depending on the utility, some have all the data, some other loans, if they have we can from that learn from this data to create a model of each equipment when they don't. We have general equations that are known from this business, and that we can use and then we do that for 1 million asset. And we put all that together, and then we can run these simulations and create value from it.
So the question I have is, it's like a, again, it's a tsunami of of simulated data. Now how do I, how do I sit there and sort of swim through that. And, and be efficient at that. And now, my, you know, assessment,
two answers to that. One, the first one is when you create and when you deploy the simulation digital twin, you start by the problem you want to solve. And that is going to tell you what you're going to look at in the simulation. So we provide a huge amount of data to have from the future. That's data from the future. And that's not only one future, it can be hundreds of future depending on what you do and the way you manage it. And this, the first way is you start by knowing what are the KPIs I'm interested in? And these are the one with the platform we're going to show you that's the first answer. But the second answer I would say, is, probably is very, very interesting. Because all these data are what we call synthetic data, there are data generated by the by the twin, right by the simulation. And you know that one, what you can do with this data, you can you can use AI algorithms to be trained with these data, right. And then you can learn from the data you produce about the system, you have. And you can use these tools to find the best course of action when it's too complex. So that we have, I would say AI at the beginning, when you say for example, to get the data for a given asset, and being to learn when this asset with AI or you know, neural networks can upstream us that you put that in the model, you put the knowledge of experts, then you run simulations, and you use AI to learn from the results of the simulation.
I am a big fan of simulations. I can't put my arm around the level of complexity that you're talking about. I mean, that's if you're looking at just utilities alone. I mean, that's just as it's mind boggling.
Sure. And by the way, remember the it's complex systems modeling very, and the very important thing is complexity is you, you always want to get rid of complexity, because you say, Oh, if I touch here, it's going to have an impact there. But now assume that you want to have an impact there. And that you have the possibility to know that you just have to touch here, by cascading effect to obtain at very cheap cost, the impact you want. That's the future. See? And that's the President
says, Well, I gotta tell you, man, I've enjoyed this conversation. It I hate to call it short, but it was too short. How do people get a hold of your Michel? If they want to get a hold of you? How did they get? How do they
contact you? Oh, they go to our website? Because we'll take that. Very, very simple.
There it is. Yeah. Thank you. Thank you for your patience. By the way, I really enjoyed the fact that you waited around. That was a good time.
It was great time. And now the tapas and the gears right.
genotoxic. Okay. I'll amp it up. Alright, listeners, thank you very much for joining industrial talk by once again, we are broadcasting from the IoT solutions, World Congress, Barcelona is the town. And as you can tell, it's getting close to the evening. And that means it is Janet Donek time. But this has been a great event. Michel, thank you. Thank you so much. All right, listeners, we're gonna wrap it up on the other side. So do not go away. We will be right back.
You're listening to the industrial talk Podcast Network.