Wizata
Adrian, thank you for joining me for today's episode of Espresso 4.0. Welcome. Thank you. Thank you for having me. I'm Adrian, born and raised originally in Germany, and I moved to the UK when I was 16 or so.
I went to school there, did my degree up in Scotland in the granite town of Aberdeen, then moved to Austria to do some research in the field of facility management and real estate evaluation, and in, I think it was 2009, I came back to Germany, joined the family business.
The family business is basically been doing CAT software since the early 80s, when the dad started it, then moved on to facility management. More the operational processes of technical facilities and installations, any sort of technical business process basically.
In 2009, I started our spinoff opening to energy management, bringing me more into the manufacturing area. Coming from economic management, basically. And in 2019, we started our new exhibition framing, and we now look into the digital twin field from our group of companies.
And now that you've mentioned digital twins, I gotta say that's a pretty hot topic in industry 4.0. Everybody's talking about it, but there seems to be some disagreement about what digital twin means.
Who best to ask than the man himself? Why don't you tell us and our viewers specifically what a digital twin is? Let's start with that as a foundation.
As you mentioned, that's a very good question, especially because everybody defines it differently. A digital twin is basically a digital representation or a copy of a real-world object.
The object being, of course, either process, a real-time asset, an entire factory, a small production line, or a business process you're running.
So, in the end, it's basically nothing more than a digital representation of something real in real life. And, of course, the idea behind the twin is to represent that asset. It is to run tests on and optimize the real-life asset.
It is a digital representation that can be used for numerous different processes, and that's where all the different mindsets really come from. For example, when you have to talk to somebody who does more of the process twin of machinery, they are looking mainly at data.
Operational data, IoT information, data flying in. How much pressure did I need to produce that piece of whatever I was producing?
Then you have a more environmental Digital twin that looks more at the machine, producing a certain object, all have different sets of information.
Ultimately, the topic of the digital twin is huge, and it depends on what you want to do. So it's huge in the end, and everybody still makes out of a twin what they think a twin is.
So why don't you tell us? You've mentioned a little bit. It helps optimize processes and test certain things, right? As opposed to testing them in real life or how does it work?
What are the benefits?
It depends on the use case. When looking from our point of view, looking at the factory or technical installation as a whole, as a physical asset, I think one of the core benefits is saving time.
What we find on the shop floor, especially in the manufacturing site or utility, you name it; you will have lots and lots of engineers and specialists going back and forth onto the asset and have a look at what actually is going on in real life.
Of course, they have plans, models, and databases with lots of information.
But they will still go back and forth to see what's happening in reality. There's research out there saying we're talking about 20% to 30% of the overall working time of those specialists wasted on going back and forth.
I was on the phone the other day with a customer of ours, he's just back from the US to have a look at their site over there, and his colleague is just about a board plane to fly over as well.
And then, of course, you can use such a twin to optimize certain processes.
Where do I store my material? How does my process run? How does this work? How would that work? Even to the extent of planning purposes like would my machine fit in where I'm planning on having it?
It's a very different approach, and it all comes down to saving time and money basically. Saving your scarce resources, optimizing your engineers and specialists' use, and making them more efficient.
I'm assuming you're using the digital twin. I suppose many use cases can be augmented or supplemented with AR, right?
Yeah, again, depending on the use case. We initially thought the idea was to have a twin in your browser when we started. Have it in your web browser, have it as a platform, and then depending on whether you are working remotely in your home office or looking at a remote site on your workstation, on your laptop - you have it on a web browser and use that.
Of course, the idea is when I have such a Twin that contains information on the current status, on whatever the next step is, as you mentioned, to bring that into AR to help the guys on site with that information as well.
So then we're using tablets and mobile phones to superimpose the information we have in our twin onto reality. We're using tablets and mobile phones because we found the technology regarding glasses, and the hardware is simply not there yet.
And we talked about a few different customers, mainly from manufacturing chemical pharmaceuticals - they said someone is walking around with one of those headsets on their shop floor. They're just out of the window instantly because not going to happen, not for health and safety reasons. They can’t do that.
On the other hand, they've got a mobile phone, they've got a tablet, X proof, you name it with them. So they want to use that to superimpose that into reality.
Is this a technology that's reserved only for certain types of producers and manufacturers? Or is it universal? And if it's better for some, who are those and why? No, I think those twins are by no means industry-specific.
And since we've been out there for the past three, coming close to four years, our customers have been from different industries and various business backgrounds. Of course, those twins really come to life when things become complex, when you have distributed size, or when things become larger and larger.
A twin may be nice but not really necessary when you have a small production size or not too complex a production. You can operate that as you do on a day-to-day basis. You will have a couple of people out there doing that.
Yes, there may well be a benefit in terms of capturing information, storing data, and becoming more resilient. Of course, there is. But it really becomes powerful when we're talking about very complex installations, fast-changing installations or remote sites, spread-out installations, having sites across the globe, the country, and the region.
Then, those twins really become powerful. Because then once you have these kinds of operations or spread of operations, you have different specialists in different sites who need to help every now and so often.
So that's when those twins really become powerful, when they bring those people together, bring those specialists together to work in the same environment. So I wouldn't really make it depending on the industry. It's more that you need a certain size and complexity for that to become really powerful.
I suppose this is only temporary as technology is getting cheaper because everyone can benefit from it; it's just a matter of whether investing makes sense.
Yeah, basically, well, we're trying to make it as cheap as possible. That's still one of the huge topics of those twins, especially the initial creation of those twins.
And we're looking, of course, at infrastructure, a fast-changing environment. They are key to such a twin, ensuring it stays in sync with a brilliant environment. So to change a pump and change the environment, you need to update the twin, of course, but that is still a topic of cost and feasibility.
When I have two specialists who know the factual neural site inside out, there is no need to provide them with some digital information. Well, they know that. They know their way around that. Of course, those things can come into play when they retire and leave the company.
How do I educate the next person, the next specialist coming in? So that's when that comes into play. But generally speaking, at least at the moment, it's more focused on the larger ones or the complex sites.
And in fact, I think it plays well hand in hand with the aging workforce and the difficulty of finding qualified personnel to work in manufacturing these days. At least at the number that the industry requires.
So with a tool like this, knowledge transfer is very easy. Onboarding is very easy. Managing multiple assets is easier.
When you mentioned a customer from food and beverage in Hungary, where they're exactly that problem of rather high staff fluctuation, they use it to train the new staff in terms of which buttons to press and which not to press. So training them offsite on the job allows for quicker employee training and a knowledge transfer.
It seems that apart from the size and complexity of the process, another prerequisite would be that the process, or rather the production process relies more on the machines than human manual work. Am I wrong? The digital twins can bring more to the production processes that involve a lot of machines working as opposed to, say, an assembly of people assembling it.
Yes and no. That's a tricky one. I wouldn't see that in black and white. The benefit of such a twin - again, what kind of Twin we're talking about - a product or even a process twin because that significantly impacts the actual production.
When you look at the assembly, on the other hand, why not. Assembly has tons of storage. Where do I have enough space? Well, how can I digitize my existing assembly line? How do I recreate? How do I restructure certain things? That's when such a twin can come into play and help you along the lines.
So it's not necessarily black or white. You really just have to think, how am I planning on my Twin helping my day-to-day operations? But having a Twin just for the sake of having a twin will not win anything. It's just a matter of what I want to use my Twin for.
I understand. Okay, fair enough. I know I threw you a little bit of a curveball, but it just came to me as we were looking. But I think you find your way out of it quite well.
How does one start a digital twin project? What are some best practices, and what are some trappings to keep an eye on?
It's an IT project. With any IT project, start small, start easy, and do one step at a time. And that's one of the huge topics we constantly have - Oh, I need my twin off this entire site, ideally the entire company, and I need it tomorrow. Just take a step back and take it slow. That's one of the major shortfalls. They just want too much, too quick.
Best practice, or what we typically do with our customers, is starting small.
Start with one site, one area, with one shop floor. Do the capturing, create the twin, create the geometry, the 3D capacity, and the environment, then bring in your existing information, depending on the process you're looking at.
Am I looking more toward energy management? Am I looking more toward maintenance? Bring that data in and have that thing grow slowly. Grow and grow and grow, but do that slowly and simply.
What information do I need? Try not to overpower too quickly. And then, once a certain process or the defined processes work on a small scale, you can start scaling it internationally, scaling it to the other sites. I think that's one of the key tricks of bringing something like that in.
We reviewed these repetitive themes, which are not foreign to anyone who's watched any episode of Espresso 4.0. Know why you are doing it, identify what you will start with, and do it incrementally.
And, of course, onboard the right people so they can later adopt that solution because there's no point they have an amazing solution in an amazing organization if no one is using it.
And last but not least, what's the coolest thing you recently saw in terms of technology or software development that was stuck up that you want to share with us? It doesn't necessarily have to be a digital twin.
Well, it kind of does, but I think apart from there are lots and lots of cool things coming out. There is so much cool stuff all around, whether cool new drones or new little toys running on.
The coolest one that did stick out for me was Elon Musk's new robot, the Evvr. I knew you were going to say that. It's so good. The reason for that is, well, I think if that thing really comes to life, if that thing really kicks in, that will be a real game changer.
We're playing around with robots as well. We've got a project running with one of those little robot dogs wandering around. The problem is, what we found is that the world we live in today is not made for dogs. It's made for humans.
Any type of environment is made for humans. So if you have a robot dog, you must train that thing to find its way around an environment made for human beings. And if you have a robot that can act like a human being, that would be awesome.
I think it's going a bit more than that, but I think I can imagine where you might think of Elon Musk as a person, whatever you want to, but I think he may be the one who can bring that thing actually to life.
Well, thank you, Adrian. This has been very enlightening. Thank you for being a guest. We love having you.