MICHAEL BIRD

Hello and welcome back to Technology Now; a weekly show from Hewlett Packard Enterprise where we take what’s happening in the world and explore how it’s changing the way organisations are using technology. We’re your hosts Michael Bird...

AUBREY LOVELL

...and Aubrey Lovell.

And this week we are delving into the fascinating world of the digital twin.

We’ll be asking the obvious question: what is it?

We’ll be looking into how digital twins are made and the technology needed to do so.

And we’ll be answering the question: why would someone want one of these in the first place!

MICHAEL BIRD

That is right, so if you’re the kind of person who needs to know *why* what’s going on in the world, matters to your organisation, this podcast is for you.

(oh) And if you haven’t yet, subscribe to your podcast app of choice so you don’t miss out.

Right, Aubrey, let’s get into it!

MICHAEL BIRD

So Aubrey, I'm going to put you on the spot here. How do you present data?

AUBREY LOVELL

Well, I mean there's obviously a lot more options these days. There's all different types of systems that we use and metrics that we track, but I'm curious, how do you present data?

MICHAEL BIRD

I sort of organise my life in spreadsheets with the occasional graph. Anyway, I think for many organisations it's a big question and it is an incredibly important one to answer and it's all well and good to have a big table full of numbers, don't I know? And maybe it's showing readouts from machines or sensors, but it is really difficult to visualise what these can mean in practice when you're overwhelmed by the quantity of data in front of you.

Now, one way of showcasing this data is using models. These can be models on a screen or even 3D models using virtual and mixed reality to help somebody understand what the data in front of them is actually saying.

AUBREY LOVELL

But if you want to demonstrate how something works, think about a factory floor or an aircraft. You need that model or twin to match the object in real life exactly. But to make the model identical to real life, you need the data and lots of it. If you want your model to be changing in real time to reflect a factory floor, you need not just the infrastructure to measure what's going on, but also to process the data and present it in a digestible way.

So how do you go about building and using one of these twins? Well, while he was at Hanover Messe, Michael spoke to Sebastian Schwarz, General Manager at NetAllied, to find out more about these digital and 3D twins.

MICHAEL BIRD

Okay, so my understanding of digital twins that they aren’t very visual... they tend to be more on paper, but it's the sort of stuff that you're creating like 3D versions of digital twins.

SEBASTIAN SCHWARZ

a digital twin itself has nothing to do with 3D at the beginning. Effectively the foundation of each and every digital twin is data. Without data you're lost. If you want to build something like a digital twin and digital twin is quite a complex term.

So what we have typically before we have digital twins is we have digital shadows and before we have a digital shadow we have a digital representation of something and all of that has nothing to do with 3D. So the solution really is having a representation, a digital representation of your asset, a product, a factory, whatever, then have it as a digital shadow that is reflecting really what is happening outside there

As always, 3D has nothing to do with it. It's just an additional layer on top of that, an additional attribute to your assets. But it's quite valuable and if you want to take a look what your factory might look in the future, a 3D digital twin might be an interesting thing.

MICHAEL BIRD

So you’re creating essentially a digital version of a physical object, a building, a space, on top of a digial twin... It sounds pretty sci-fi.

SEBASTIAN SCHWARZ

Yes, sound sci-fi, it's not so much sci-fi for us. We are doing exactly the same thing. It was claimed differently the past 20 years ago, but it somehow sci-fi. Let's see it like that. Why? Because if you really want to have a living good and great and working digital, it involves so much work. You have digitalization strategy, you have high quality of your data, you need to address your data, you need to align your data, you have a consistent data flow, it's quite useless to have a digital twin that you build up once and that is not living. So if you change something in your production pipeline, for example, you really need to make sure that your digital twin is living.

And then on top, yes, sure that 3D component is something that you see from film industry. So there is a bridge between that sci-fi videos and sci-fi films and a digital twin from a technical point of view, yes.

MICHAEL BIRD

OK, so What is the purpose of a 3D version, a 3D visualization of a digital twin? Because presumably you have all the data, you can understand what's going on.

SEBASTIAN SCHWARZ

So in contrast to the typical digital twin where you only know the feedback from your sensors is that additional layer includes better optimization. For example, if you have the 3D model attached to your digital twin, you can do collision detection. You can have clearance, you can have something like optimizing the path that your workers are taking better logistics just due to the fact that you have with this additional layer of complexity but also on information, the 3D layer, you can start training your people far beyond the product has been manufactured.

MICHAEL BIRD

Yeah, because I guess we as humans are quite visual people and I think visualizations is quite helpful to picture, you know, a factory floor or a product line. You can sort of go, oh, we do that and then we do that. Oh, that makes sense.

SEBASTIAN SCHWARZ

So I'm coming from the engineering perspective. Our digital twins do not really look that fancy. Why? Because it's engineering data being visualized. So it happens that the building itself is green and the floor is red and steel is yellow. the engineers plan it like that because they want to have a quick visual representation. This is one aspect of the 3D digital twin. The other one really is then if I have the materials, if I have textures, stuff like that, then I can prepare great renderings. I can even for people that are not yet into that engineering business can present the details. What will your factory look like in the near future?

MICHAEL BIRD

So what is the process of building a 3D twin? Is it a lot of fancy 3D cameras, LiDAR sensors, that sort of thing?

SEBASTIAN SCHWARZ

I know everybody's claiming LiDAR scans as a digital twin. The problem is this is only the last step that you ever should do. If you're in LiDAR scans, you did something wrong in the beginning. it all depends on the data, the quality of the data, the confidence that you have in your data, the actuality of the data, stuff like that. And all of that is killed if you're taking a LiDAR scan. Because a LiDAR scan, you take that picture, it takes, let's say, an hour to scan your factory, and then you get out of your factory, and then it looks the next day completely different.

It is a good approach to start if you have nothing. If your factory is 100 years old, well, it might be that you do not have 3D data for your digital twin. But effectively, you should keep your data clean. You should keep consistent with the data. If you modify something like a conveyor, something, a moved robot, something like that, then please keep in mind to update your planning systems, to update your billing information management systems, stuff like that. If you do it correctly, you will hopefully never again have to do a second LiDAR scan again.

MICHAEL BIRD

So you’re saying, good data in good data out. So rather than doing a 3D lidar scan, you're taking existing data from those organizations, whether that is, don't know, CAD drawings. I'm probably oversimplifying it, but that sort of thing.

SEBASTIAN SCHWARZ

Yes, exactly. You need to be as close as possible with the engineers, with the designers. So this is, you have that digital representation, you have that digital shadow that is coupled with the real data, you have the digital twin that is feeding the information back, but all of that depends on a consistent data flow from the very beginning, from your engineering phase up to then the 3D digital twin.

AUBREY LOVELL

Thanks, Sebastian.

It really is unbelievably Sci-Fi and I do have this vision in my head straight out of a film with the floating models in front of me.

Really looking forward to finding out more later in the episode.

MICHAEL BIRD

Ok, well now it’s time for “Today I learned”, the part of the show where we take a look at something happening in the world that we think you should know about. Aubrey, what have you got for us today?

AUBREY LOVELL

Alrighty then Michael, a couple of weeks ago, you did a Today I Learned all about the breaking news of a new dinosaur. Well, if you thought that a new dinosaur was a surprise, then I've got an even bigger one for you. Are you ready for it?

MICHAEL BIRD

I'm ready. What is it? What is it? What is it?

AUBREY LOVELL

Okay, scientists claim to have discovered a new color. Can you believe it? How do you even do that? I mean, now I'm kind of triggered because with the whole planet situation, it's like, this is a planet. No, it's not. Yes, it is. And now we have a new color.i

MICHAEL BIRD

... explain this new color thing.

AUBREY LOVELL

Well this apparent new colour, named olo, was seen by five people after scientists in California fired lasers into their eyes to manipulate their retinas. This is getting a little crazy, guys. The colour itself cannot be printed or shown on a screen but was described in the paper announcing its discovery as a “blue-green of unprecedented saturation”

MICHAEL BIRD

Okay, and it can't be printed or shown on screen, but our producer has put an approximation in front of us, Aubrey.

AUBREY LOVELL

It's definitely giving teal or like a seafoam green if you're into interior color design.

MICHAEL BIRD

Yes, I think seafoam green or teal is how I'd describe it.

AUBREY LOVELL

Now I'm even more perplexed because how can you even consider it a color if you have to fire a laser into your eyeball

MICHAEL BIRD

It feels like a very exclusive colour...

AUBREY LOVELL

And I've assumed that you've heard of rods and cones in the eye, right Michael?

MICHAEL BIRD

I have heard of rods and cones...

AUBREY LOVELL

These are the cells in the eye which detect light. And in this case, cones are the ones we care about as they are responsible for our color vision. So we have kind of three different types of cones which respond to short, medium, and long wavelengths of light.

And in real life, there is no color which activates the medium type cones without either the long or short cones. However, this study used lasers to bypass the long and short cones and excite the medium cones exclusively. Hence, probably the laser.

MICHAEL BIRD

I see. I see, I see. So it was triggering a part of the eye that isn't normally triggered or can't be triggered. I see, okay.

AUBREY LOVELL

That’s correct. And now, obviously, as you can see where this is going, the work has had mixed responses within the scientific community. However, the researchers hope that it will help us understand more about color blindness or even how the brain creates our visual perception of the world from the information it gets from our eyes. Huh.

MICHAEL BIRD

I think that... I can imagine the conversation of those scientists asking test participants or other scientists to take part in this experiment. You can imagine it was something like, “so Dave, I've got something I'd like you to take part in.” “Yeah, yeah, what is it?” “Involves lasers.” “Okay. Yeah, I'm keen. Keep talking”. “I'm going to shine them in your eyes.” “Okay. I'm out. You can find somebody else.”

AUBREY LOVELL

Can you imagine the paperwork that you needed to sign to get that rolling? Goodness gracious

AUBREY LOVELL

All right, well, let's return to Michael's interview with Sebastian Schwarz, recorded at Hanover Messe.

MICHAEL BIRD

so I feel like we've got absolutely ahead of ourselves, because what I haven't actually asked you is about, who are you building these models for?

SEBASTIAN SCHWARZ

So our customers include the biggest Aircraft manufacturing companies and what they have is they're not so much interested in the in the factory. They are more interested in the product the product in this case is the aircraft and each aircraft is a complete different one from the previous aircraft. So they are really highly individualized for the specific airlines for example and they need a digital twin. And the data here is enormous. So we specialized on handling these huge amounts of data

MICHAEL BIRD

And from a data perspective, in terms of the sheer amount of data required to build a 3D visualization to begin with, this is not something that's static. You're continually feeding data into it.

SEBASTIAN SCHWARZ

So just some rough numbers to give you an idea. A final assembly plant of an automotive manufacturer. This is where they are out of the paint shop and now they are integrating the different stuffs.

There are so many jobs for one specific vehicle that can be customized and you have a hundred thousand tasks. You have 80 million different nodes. So it's the leaves of your digital twin. One leaf would be one single screw. One single screw and one station and one vehicle. This is one element. We have 80 million out of that.

So we just provide the technical infrastructure to handle even that well that 1 % 3D digital twin use cases.

MICHAEL BIRD

And you're continuously feeding data using sensors on the floor, sensors in the factory, or sensors on the aircraft, or sensors. When we say sensors, what do we really mean by that?

SEBASTIAN SCHWARZ

Well, a sensor effectively is something that is giving me data from real time infrastructure. A sensor itself can be actively measuring the distance of a gripper to the body and wide in automotive production. A sensor can identify what is the current position of my robot. Everything that is giving us data, real time data, and they need to be accumulated and handled in a fancy way that makes sense. And they're very important not to forget is if you want to use your digital twin data throughout the whole production pipeline and then feed stuff back again to the beginning to the engineers, to the designers, you really need to take care about not only having the data available but also make it uniquely identifiable.

it is a complex task and it's an ongoing task. So at the end it's something probably nobody wants to hear because it's so boring. But from my perspective, it's not boring. It's digitalization.

MICHAEL BIRD

So what are some of the big challenges? I mean, we sort of alluded to one, but what are some of the big challenges that come with creating a 3D twin of, a factory?

SEBASTIAN SCHWARZ

Several of that problems have already been mentioned: the amount of data, the actuality of data, the confidence that we have in the data, the effort it takes to migrate the next iteration of my digital twin to keep it consistent with the reality, all of that. It requires often that you have a top-down approach where really different silos and different departments are working closely together and be aware of there is something before me, I am consuming my engineers’ data, I'm doing the simulation and now somebody is doing something with my simulation. I want to provide consistent access to the data and access that is easy to be reused for non-specialists that follow my specific domain.

MICHAEL BIRD

So what's the future for 3D Digital Twins? More VR, AR, better renderings…

SEBASTIAN SCHWARZ

So the more we have a 3D digital twin and the more people see the benefits of the 3D digital twin, they communicate to their predecessors, please raise the quality, please help me to get better data from you and I will provide better data again to you. So we have that closed loop where I'm generating data for my predecessors and my successors. This is a natural way that 3D digital twins will evolve and will get better each and every iteration obviously. it will be the foundation for good and trustworthy AI agents training. so, future is bright for 3D digital twins.

AUBREY LOVELL

Thank you so much Sebastian.

It does feel like a big step into the future using these 3D Twins and I am kind of disappointed that I’ve never had the chance to use one myself!

AUBREY LOVELL

Okay, well, we're getting towards the end of the show, which means it's time for This Week in History, a look at monumental events in the world of business and technology, which has changed our lives. Okay, Michael, you gave us the clue last week.

MICHAEL BIRD

I absolutely did. Last week’s clue was of course:

It’s 1840 and this little picture of the Queen is going to take you places….

What did you think it could be Aubrey?

AUBREY LOVELL

I can't remember what my response was, but hearing it back, was thinking maybe it's some sort of ship, but I don't think that's correct.

MICHAEL

Well, I knew it because I'm a bit of a nerd. I think I said stamps. And it is, of course, the release of the first adhesive postage stamp called the Penny Black...ii

ow, these days we have email and people do use the postal service much less than they used to. However, back in the 1800s, the idea of weight-based prepaid postage was completely novel.

The first design was rejected and ridiculed into submission. However, the second design, which was based on a sketch of Queen Victoria, went down much better. This version of the then Queen would stay on the stamps for the remaining years of her 63 year reign. The stamp allowed letters of up to half an ounce, we'd call that 14 grams here in the UK now, to be delivered for one penny, regardless of the distance. Now within the first year of what became known as the Penny Post,

The number of letters sent had more than doubled, with the number doubling again within the decade. Now obviously with the advent of email, we send far fewer letters nowadays. Imagine if the 100 emails that the average worker got a day came as letters. However, I think I would say that the Penny Black is probably one the most important inventions in history. Aubrey, have you sent a letter recently?.iii

AUBREY LOVELL

I actually have, believe it or not. Yeah, I still believe in it. My handwriting is not so great, but I think still sending letters is wonderful. It's a wonderful thing to have.

MICHAEL BIRD

My young nephew and I became pen-pals briefly when he discovered he could send letters. So I got a lot of enjoyment out of sending letters.

Right, Aubrey. You’ve got next week’s clue. What have you got for us?

AUBREY LOVELL

Okay, so this one is a huge moment in history, and we're bumping it up a little bit. It's 1980, and this microscopic killer is gone for good.

MICHAEL BIRD

Ooh, okay, 1980s. goodness me. It's the eradication of something, isn't it? It's some vaccination program. Or what would that be? Something like measles? No, not measles. would it be? Polio, polio, polio. Maybe that's too late. I don't know. Okay. I think it's going to be the...

uh yeah I think it's going to be some sort of vaccination eradicating some sort of uh parasite virus bacterial thing that's what I'm going to go with.

AUBREY LOVELL

I think that covers all bases, so I think we're going in the right direction for sure... A diplomatic answer.

AUBREY LOVELL

That brings us to the end of Technology Now for this week.

Thank you to our guest, Sebastian,

And of course to you too...

Thank you so much for joining us.

If you’ve enjoyed this episode, make sure to leave us a review or a comment wherever you listen to your episodes!

MICHAEL BIRD

Technology Now is hosted by Aubrey Lovell and myself, Michael Bird

This episode was produced by Harry Lampert and Izzie Clarke with production support from Alysha Kempson-Taylor, Beckie Bird, Alyssa Mitry and Natasha Naik.

AUBREY LOVELL

Our social editorial team is Rebecca Wissinger, Judy-Anne Goldman and Jacqueline Green and our social media designers are Alejandra Garcia, and Ambar Maldonado.

MICHAEL BIRD

Technology Now is a Fresh Air Production for Hewlett Packard Enterprise.

(and) we’ll see you next week. Cheers!