The value of digital twins in space, manufacturing and healthcare

Helena Jochberger:

Dear guests, we are about to talk today about the broad topic of digital twins. We start by setting the scene around the current application of digital twins in the industry. In general, the digital twin concepts are not new, but let us elaborate a little bit on how they are currently being applied to the respective industries. Maybe, Harjit, you would like to kick it off. What is a digital twin in the space industry, and what is it used for?

Harjit Sheera:
Absolutely. Digital twins are basically a virtual replica of a physical system. And certainly within the space industry, they're increasingly being used to improve things like mission design, operational efficiency and system reliability. 

For the space industry, it's particularly relevant because space is a harsh environment, and the systems we developed to get to space are complex and they're expensive. Within the space industry, it allows us to mimic these environments and see how our systems respond before we spend all the money and all the complex design effort to develop those systems. But also, it allows us to adapt those systems while we're on the ground, when they're easily reachable as well. Looking at some of the applications in things like satellite design and testing, simulation of space environments, digital twins are used to really simulate the harsh conditions of space. 

These are things like the radiation and extreme temperature fluctuations that you get and the effects of microgravity. We know how it obviously affects us here on Earth, but to see how it affects the products that we produce and the systems that we create in situ is a very important thing, so that we can develop those products and change how we manufacture products as well. And then there are things like structural and thermal analysis and other areas like mission planning and operations. 

As I said already, it's very expensive to put together a mission, and it's very complex. The more we can simulate these environments and adapt them and really refine how we're producing these environments, that is also a good thing for us to do so that we can, again, reduce those costs and improve the way that we're producing our products.

Helena Jochberger:
Very interesting. And I think especially in the area of simulation, as you’ve said, you simulate something that later on will be applied. 

How is that, Ben, in your industry, in health and life sciences? Is there any simulation aspect to it where are you using digital twins? 

Ben Goldberg:
Yeah, there's absolutely simulation. It's always lovely to hear about our endeavors in space. They’re such grandiose engagements that we're doing, and I'll bring things a little bit closer to Earth; but it's fantastic, and there are obviously a lot of similarities. 

The work that we do with digital twins and healthcare can really be anything from that physical infrastructure that you relate to, be it a piece of hospital equipment, to leveraging IoT sensors and such to capture data or, even broader, at a hospital physical building level, making sure that certain systems are being checked in a digital twin, replicating that physical entity that is put in place. That can even extend to an entire hospital system, for that matter. 

There's been a lot of very interesting work that's happened specifically around that, where you're able to capture patient flow within a hospital as well. Leveraging that digital twin to capture elements throughout the patient journey within the four walls and within the healthcare system, and then, to an even more micro extent, but also very compelling and pertinent, is to have a digital twin of the patient themselves. To be able to monitor actual elements that are associated with that individual and try to be more proactive in healthcare delivery. 

Helena Jochberger:
Very interesting. Having the human in the loop or in the middle, of that specific persona, is a great criterion that defines your industry. 

You mentioned, for example, the IoT devices and obviously when I'm listening to you, Ben, your industry and my manufacturing industry have a lot in common. That would be, I would say, my next question. Are there any lessons from your specific industries that could be applied to others, in the best sense of learning from them? 

Ben Goldberg:
Yeah, maybe I'll start with this one. I think healthcare has a lot to learn from other industries. Healthcare in general tends to be fairly risk averse in how we implement technologies because in the end it really impacts patients. So, we'll do a lot of work that's a little bit more on the administrative side and test the boundaries of emerging technologies. But a lot of the lessons learned through other industries are really what compels the health and life sciences industry to be able to implement a lot of the solutions that we do. 

To your point, Helena, about life sciences and the manufacturing associated with pharmaceuticals, that is a huge industry that we can leverage to really inform a lot of our current-state processes and evolve the model. But let’s turn to Harjit for some comments as well.

Harjit Sheera:
Yeah, absolutely. For us, it's more to do with the performance and the manufacturing side of things and the tolerance of the way that your systems work. It's more of a manufacturing aspect. Looking at the lessons that we can learn from space, it's using things for real-time performance and monitoring and for predictive maintenance. So, those kinds of things can be applied certainly to our industry, but it can stretch across to industries like the manufacturing industry and the energy industry as well, where you're optimizing a lot of factory equipment, but also power grids and oil rigs.

Another major factor for us is looking at risk reduction through simulation. Again, this is something that echoes in Ben’s industry as well. Here we're looking at simulating environments in test scenarios virtually so that we can reduce risks and then, again, optimize these outcomes. As I already said, healthcare is another area this stretches to but also automotive as well. 

Then, we're looking at things like the optimization of complex systems. As I mentioned earlier, a lot of our satellite systems are very complex systems—trying to stretch this to optimize the performance of our interconnected systems as well. And this stretches across to supply chain management as well, where we're looking at modeling supply chain networks and improving resiliency and efficiency. Those are just some aspects. 

Of course, there are many more, including collaboration and cost savings, and anomaly detection. There are various aspects that we can continue to talk about, but I'll stop there. 

Helena Jochberger:
Very interesting aspect. You were mentioning maintenance, and maybe I'll give it a minute or so to describe it a little bit. You deal with complex systems, Harjit, and so do we in manufacturing, especially when we are manufacturing ships or aircrafts, and obviously, you might have a digital twin along the lifecycle. 

So, you are in R&D and engineering, and you might have a digital twin as planned or as designed. Eventually you built the thing where there might be some deviations from the planning. Then, once you have it up and running, and you're in the mode of servicing or maintaining it, you have a digital twin, as maintained for all the complex systems that have a very long lifecycle. For an aircraft, for example, after 30 years, every single part within the cabin has been exchanged one or multiple times. I think this is where a digital twin can help us. And then also looping it back to the engineering to say where can we do better in terms of design?

Helena Jochberger:
So, interesting aspects. Let me move also to the next topic. When we look to CGI’s Voice of Our Clients analysis, we certainly know that sustainability is a key topic for a lot of our global industries. Maybe there are also some common drivers across your industries. How can digital twins contribute to achieving sustainability goals? Maybe, Ben, let us start with healthcare and life sciences.

Ben Goldberg:
Not long ago, I was on our colleague Peter Warren's podcast as well, and Pete is our Global Industry Lead for Energy and Utilities. There’s a really nice complement between energy and utilities in the health sector in terms of some of the drivers, notably around sustainability. One of the best examples that comes to mind is an organization that we've worked with that has a digital twin solution that they implement within hospitals. Some of the benefits that they can ascertain are, again, following the patient through their journey within the four walls of the hospital. They can track where they are at any given time. 

Hypothetically, if you've got a patient who is enroute from triage into a patient room, that room can be prepped and ready for them. The lighting can be set, and the temperature can be set, and then they'd be entering the room. At the same time, once their stay at the hospital is done and they've left the room, triggers can be put into place to lower the blinds, turn the temperature down, turn off the lighting, and such. There are immediate sustainability implementations that can be done throughout that whole journey. It's really interesting. Some of these micro changes can make a big macro effect as well. That's some of the stuff that I've seen within the industry. 

Helena Jochberger:
Very interesting. For sure, when we look towards space, Harjit, I know that with space data, we can enable so many sustainability activities. Would you like to state some of them? 

Harjit Sheera:
Yeah, absolutely. Sustainability is huge for everybody, especially when we look at the amount of Earth observation data we can gather. Looking at the way that we can simulate what's happening on Earth with, not just our food systems, but with the climate and the environmental challenges that we have, how they're all affecting Earth. So, things like digital twins of the Earth, where we can simulate what happens when, for example, the global temperature increases by a degree, half a degree, one and a half degrees, we can see how that changes all of our environmental conditions—how waves are changing, how the tidal motions will change. 

Also, one of the things that they're doing in the space industry is to try and make these aspects more real to people because it's all very well looking at lots of data on a screen and moving it on your screen, but it's very difficult to really understand what those impacts are going to be. 

Certainly, within the space industry we've been using some of the techniques that we have in the gaming industry. Using gaming devices but using real life data that we've captured to be able to really show people through virtual headsets what's happening with the data. What does it really mean when our temperatures rise, when our wave heights increase? What sort of effects will that have? How does it feel to be in the middle of a hurricane that has occurred because of our temperature changes? 

Also looking at sustainability, this is helping us to really look at the way that our companies are operating, the way our supply chains are operating and how that echoes all the way through the business that we do with all the suppliers that we use as well. How can we all improve the way that we operate, and how does that affect our overall impact on Earth as well?

Helena Jochberger:
Yeah, absolutely. That's very interesting. What we also see within the sustainability goals, and you were mentioning previously, Ben, also Pete Warren, our Global Industry Lead for Energy and Utilities, and he's heavily involved, of course, in the energy transition and in all things hydrogen. 

What we also see here is that, at the moment, the industry or market lines continue to blur. What I mean by that is that when an energy company starts to produce hydrogen and gets a byproduct, fertilizer, it really moves out of its own industry swim lane and into becoming a manufacturer. 

I think an interesting question in that context would also be how can we create across the industries shared data ecosystems, or digital twins, that respect the unique requirements of different industries? Because we all know that we are heavily regulated. Every one of us is, but in a different way. Is there any idea from your side how we could tackle such an endeavor in creating shared data ecosystems? 

Ben Goldberg:
It's a great question, Helena. I think even within the industries there's enough to tackle to have a lot of that interoperability happen, to have a lot of that governance put into place, to have the right standardization put into place. You know, I mentioned architecture, but even with data governance, there's a whole slew of different elements that really need to be put into place within each industry.  

It's reminiscent of a post that one of our strategic business unit leaders, Tara McGeehan, who leads our UK and Australia operations, put up on LinkedIn the other day talking about what CGI’s approach is from a fulsome view of patient care. And she refers to a comprehensive health record that actually ties into a variety of different feeds, so outside of just your electronic health record, but also including genomics information, wearable tech, lifestyle data, different treatment options, and disease management. That’s just within the industry. To paint the picture of the data that we would be trying to fit in to have that be true, comprehensive view of an individual, let alone all the other factors across industries, is a very good initiative, but there's a lot that needs to be factored in as you go. 

Harjit Sheera:
Yeah, I would echo the same as Ben. Standardization and regulation are huge aspects for us. If you just look at the amount of space data you get, the standardization across all the data is slowly improving. But there are so many formats of data and so many degrees of metadata that we have. And each format is slightly different and must be manipulated slightly differently, which takes a lot of processing power and storage power to really capture all this data. Because, of course, we know the amount of data we can retrieve and collect is only going to increase as the years go on. 

So, trying to make all of this data usable to the people who are creating those applications, and then to the users of the data themselves is something that a lot of people have been working on to improve. There are lots of new processing techniques that have been involved, like a lot of the platforms that we create. But this is really, again, to try and make that data more standardized, but also regulate how it's being gathered and manipulated to enable more usability. 

We have aspects where people like the European Space Agency have created platforms where they're gathering lots of data in an attempt to really try and create platforms to standardize what's happening and to make sure that the data that we have, because we have collected so much, that we can actually use it. So, it's not just sitting there, it's being used. We can use it properly, and a lot of the constraints that we have are not really due to the amount of data we have. A lot of the time, it's trying to create those applications that can really use that data to its full potential, so that we can really make the most of the information that we have.

Helena Jochberger:
Let us think a little bit bigger for a moment and imagine a future where digital twins across all our industries could be interconnected, creating a global ecosystem that would amplify their impact on society. The key word of today is, for sure, resiliency in all sorts of aspects because we are living in tense geopolitical times. So, what could there be in the future with having interconnected digital twins coming from the industries and fostering such a global ecosystem? Ben, any ideas if we think into the future? 

Ben Goldberg:
I love it. I love it because it's so aspirational. It's a compelling vision, but honestly, I think it's within reach. I mean, it comes back to the points around collaboration and the whole concept of wanting to be collaborative and problem solving, and where there's a will, there's a way. If I think about even within the industry, the opportunities that exist, when all of a sudden you start to tap into the massive data repositories that are being created on a daily basis are really fascinating. 

There's a whole concept within life sciences when it comes to reevaluating some drugs that are on the market right now, and what type of drug repurposing can happen where you hadn't realized that a pharmaceutical can actually address a different disease that it hadn't been created for. And that's something that's quite common. And think of that from a global perspective—the more data that you have, imagine what you could find out and the linkages that happen across ecosystems that perhaps hadn't been as evident when you're more siloed in your approach. So, I'm a big fan of the vision. I think it's a lot of what we do. It is sort of an underlying theme in the way that we address big data. We address collaboration in architecting these solutions for our clientele, and for the world at large as well.

Helena Jochberger:
From your side, Harjit, what would you say, looking into the future, what could be such an across-industry digital twin? 

Harjit Sheera:
Well, we already know that space data can be used in almost any other industry, and there's lots of applications that we use. The way that we really can see the collaboration that we can have with other industries is what Ben already mentioned: an application where you can navigate yourself through healthcare institutional and organizational buildings even. And this is one of the ways that we see digital twins. 

So, the satellite industry or, certainly, the space industry, would provide the data, but then you could cross that data and use that within other industries like transport, manufacturing, and healthcare. For example, the data that we have could help you to create a digital twin of a building, and then you would be able to provide your patients with a navigation route to get to where they're going because, quite often, these places are huge and they're quite complicated to navigate. 

But also, you could apply that to the retail industry. So, again helping people to navigate where they want to go, increasing business for the businesses that are there. 

There are so many applications of what we can do with space data and how we could then create digital twins within other industries. In manufacturing and automotive processes, certainly you know, Helena, about using satellite data in the manufacturing process and how closely they're linked already, and the number of digital twins that are already created to bring those processes together. It's really limitless. And, as Ben said, it's quite exciting. And it's not that far away. 

Helena Jochberger:
Indeed, it is not far away. And you were mentioning especially my area of supply chain or travel transport and logistics, where, of course, satellite data is being used more and more to enhance that real-time transparency. And thus, modeling a digital twin on the supply chain. And the more complex the product is, the more suppliers involved, the more there is the need, of course, for data sharing ecosystems.

Helena Jochberger:
I would like to bring into our discussion a bit of the flavor of AI. We all know that, for the last two years especially, the generative AI topic, and the overall AI topic has really accelerated. When we are including that AI aspect into our discussion of digital twins, CGI speaks of the so-called digital triplets. 

Let me give you a short introduction or definition on how we at CGI understand digital triplets. It’s basically an advanced evolution of the digital twin concept as we know it. While a digital twin typically replicates a real-world entity or a machine or process, in the virtual realm for monitoring, simulation and, more so, optimization, the digital triplet incorporates three interconnected layers. The first layer being the physical layer, obviously, where it represents a real-world object or an environment or a system. Then you have the digital twin layer. This provides a virtual data-driven replica of a physical entity and enables the real-time insights that we just mentioned for the supply chains like analytics or predictive modeling. 

And now comes the third layer, and that is the intelligence layer. It adds a higher-level abstraction with AI, with machine learning or with advanced algorithms to simulate—here again we have the word simulate that we touched upon at the very beginning—interactions across multiple twins and then optimizing interconnected systems and providing prescriptive insights. Building on that definition, my question to both of you would be, how would you see these digital triplets evolving with that integration of AI? Ben what would you say a few words on life sciences?

Ben Goldberg:
Yeah, honestly, I think you're saving the best for last because it's really fascinating stuff that we're seeing evolve. And some of the case studies that we've implemented for clients really are just getting the ball rolling. We’re just starting to see what's in the realm of the possible. I think it's a great introduction on generative AI and how it falls within the scope of the physical entity. 

And, you know, our global AI lead, our colleague, Diane Gutiw, she speaks to it in a very, very clever way and it puts things in perspective for me. What the generative AI layer allows you to do is to interrogate that data. And I think that that's a really cool concept to think about. 

What I also appreciate from it is that you don't have to go through an exercise of ripping and replacing existing data investments. The digital triplet actually extends them and allows that to be, again, queried, interrogated and integrated. What we see is anything from an administrative perspective, querying large data sets to get instant dashboard information about trends within healthcare, be it from a vaccination perspective, a hospital wait time perspective, or an ambulatory care perspective. 

A bunch of those are static pieces of data, but you can get that on the fly and query it and delve a little bit deeper. But then, at a much more clinical level, you can leverage information about a patient's well-being, about certain criteria, around their symptoms that they're presenting, around their demographic. And again, that comprehensive view of them to really get a better sense of how you should be treating them and what some of the prognoses and diagnoses would be because of that kind of interaction. 

It's limitless. Once you have all this data and the data, of course, is in a state of being properly managed and properly cleansed and utilized, the world's your oyster. It's really fascinating to see what we can do and what continues to be done. 

Helena Jochberger:
Very interesting. How is it hosted at your end, Harjit? Because I know space has always been a pioneer in the industries. Is that something new to you or are you using that already? 

Harjit Sheera:
It is definitely something that we've been using for many years and, of course, the added layer that we get from digital triplets is that it helps us to evolve significantly in terms of becoming smarter and more dynamic and becoming more autonomous. With all the complex mission tasks that we look to our systems to do, involving AI and having this third layer on top of our digital twins helps us to bring this up to a new level.  AI has enabled us to start to process real-time data and to really learn from it as well. Not just seeing the effects of it, but to be able to adapt it and see how it how it changes and see what the real impacts are in terms of what it does to our systems and how we can see them changing. 

Things like machine learning and those kinds of models really allow us to adjust and to look at the unexpected environmental conditions, like, for example, solar storms and things like that. But adding that additional layer really helps us apply that in a very practical sense and make it more flexible as well. Because we know that a lot of the environments we use within space are unpredictable, we need that level of flexibility and we need that level of adaptability to be able to really understand how the impacts of an environment that is really quite alien to all of us really affect us. It's a really big thing for us. 

Ben Goldberg:
Harjit, I have a question. How often do you use alien when you're talking about space? 

Harjit Sheera:
I try not to use it very often at all because of course I when I talk about it to most people, it makes their ears pick up and they think all of a sudden, I'm going to start talking about lots of alien theories. But yeah, I try not to use that word. 

Ben Goldberg:
Just checking. Just checking.

Helena Jochberger:
Going back to future Mars missions. Maybe we will see triplets simulating these very complex missions that are in the not so far in the future, I guess. 

Harjit Sheera:
Absolutely. I mean, if you look at things like autonomous vehicles, which is a huge thing here, we've been speaking to people about how space data can be used to create systems that are usable on Earth in terms of autonomous vehicles that, not only apply for one person, but how we can create whole transport systems that use autonomous vehicles. Also, again, there are Mars rovers as well, things like that, and you really need to understand how they move to be able to control them better, to be able to allow them to have more functionality. So yeah, absolutely. It's very, very relevant. 

Helena Jochberger:
I see we have still a lot to discuss, but I would say for the moment it would be great to have one key phrase for our audience to take away. Ben if you would have one sentence you would like to leave with the audience, what would that be? 

Ben Goldberg:
I think my thought when it comes to introducing digital twins (if you haven't already, I think many are very far ahead when it comes to that), is that it's about AI because there's so much discussion and hype that goes along with the concept of AI. People are feeling like they need to keep up with the Joneses and if they don't have a solution around it, they should be implementing something quickly. 

There's no rush. I think there's a recognition that it's still nascent for a lot of organizations, and I think it's about introducing it as the tool that it genuinely is, and how it can be leveraged to really inform a lot of critical decision making, be it from a business perspective, be it from a clinical perspective or an infrastructure technology perspective. It really has the capability of doing so much. And it is that new tool in the toolbox that will fundamentally change the way that you can interact with your data. So, I look forward to how that's going to be impacting my industry and, of course, complimentary to that, how it collaborates across other industries too. 

Helena Jochberger:
Thanks for that, Ben. And you, Harjit, what would be your final words when it comes to digital twins, from a space perspective?

Harjit Sheera:
I would look forward to the point where people really understand what satellite data can do for their industries. I think at the moment not everybody understands how they’re using space data today, or what it could do for them and what the potential is. And with the amount of data we have and with technologies and techniques like AI and digital twins, this is only going to help us to widen our experiences and address more ambitious projects, so things like AI lunar bases and things like that. This is really quite exciting. And the use of satellite data in other industries with added techniques like AI and the digital twins is going to bring us closer together to understand the potential of what we can do when we all work together and see perspectives from the other side of a business as well.

Helena Jochberger:
Absolutely. Thank you for these words. And I would say it's worth collaborating also across the industries. Thank you very much for being with us today, Ben and Harjit. It was a pleasure talking to you and thank you to our audience for listening. Bye-bye.