Episode Transcript
[00:00:10] Speaker A: Welcome to the EU Energy Projects Podcast, a podcast series from Enlida and France focusing on the clean energy transition for the European Union and the EU Commission funded energy projects that will help us achieve it.
In this episode of the EU Energy Projects Podcast, I'm joined by Maria Janes representing the Orite project, and Andreas Villareal from Pyramid, both part of the Zabala Cluster. Together we explore how digital approaches are advancing the sustainability of Europe's aviation with new materials and fuels.
Thank you both for being here with me today. And to begin, could you please briefly introduce your project and explain how they fit within the broader goals of the Zabala Cluster? Maria, ladies first.
[00:01:03] Speaker B: Okay. Thank you for having me.
R and D project consultant at Zabala Innovation. And I'm supporting the coordination and management of the all right project. The name stands for Airport Level Demonstration of Ground refueling of Liquid Hydrogen for aviation. And that's exactly what the project is about.
Is coordinated by the Italian national agency ENEA and brings together 20 partners for seven European and associated countries. With almost 10 million of European funding.
We are now halfway through the project. In simple terms, we are helping to bridge the gap between research and real deployment. And it's not about developing technology in laboratory, it is about understanding how the airports actually introduce hydrogen into their daily operations safely and efficiently.
[00:01:53] Speaker A: Excellent. Thank you, Maria. Andreas, would you like to tell us a bit about Pyramid?
[00:01:56] Speaker C: Certainly. And thank you so much for having me on the podcast.
So Pyramid is also a project that has started relatively recently. We had our kickoff in 2024 and the project itself stands for AI based testing pyramid towards Virtual Certification of Next Gen Composite Area Structures. Now the idea is to solve one of the issues, one of the biggest bottlenecks in the development of better materials and better technology, which is the so called testing Pyramid. So certification of these components and certification of their manufacture relies on a sequence of physical tests. So from material to coupon to element to area structure, which is costly, which is slow and which is rigid, and so changes at one level of the production line and what level of the research often requires, at least in the status quo, repeating much of that certification project from scratch and with little insight so far as to how they will affect overall performances. And so the idea is to replace this with what we call high fidelity virtual tests that are going to allow using AI to link the various levels of this certification pyramid in order to have much better insights, much better simulations to cut the computational time and to speed up decision making. So we're going to do this using four use cases for demonstrators of Technologies where we are going to essentially teach the this AI tool, the iTool, as we call it, the virtual testing platform.
How these feedback loops can help improve the testing pyramid and render production more efficient, developing new technologies and materials along the way.
[00:03:57] Speaker A: Maria, can you please tell me what are the challenges of introducing hydrogen in aviation?
[00:04:03] Speaker B: So aviation is one of the hardest sectors to decarbonize. Land cars, aircraft need a huge amount of energy while keeping weight as low as possible. That's why hydrogen is attracting so much attention. But it contains also, I mean, three times more energy per kilowatt in the combustion of jet fuel. But the challenge is that hydrogen behaves very differently.
If you want enough energy on board an aircraft, you need immediately liquid hydrogen, which has to be stored at cryogenic temperatures. So that's completely changed again. Suddenly airports need new storage system, new refueling requirements, equipment safety procedures, trained person, and new regulation.
One of the things we have learned through Altright is that the different countries apply different approval procedures, different safe requirements, and different permitting processes for those demonstrations.
So one airport could take just few months to approve a hydrogen demonstration, while another could take even years. So that's one of the reasons our project is looking beyond technology and trying to help to create a more homogenized European approach.
[00:05:22] Speaker A: Exactly. I think the problem of homogeneity, or whatever the word is, if I'm pronouncing it wrongly, is one of the biggest challenges in EU funded projects. I've heard it many times before now, staying in the challenges, let's say, topic. Andreas, I'm sure there are challenges around current aviation materials. So can you tell me what AI can bring to material testing, for example?
[00:05:50] Speaker C: Certainly. So in terms of challenges, modern aircraft, they already rely heavily on composite materials, which essentially integrate several different materials and process structures in order to obtain a high hybrid results greater than the sum of its parts.
But these composite materials are very complex in their manufacture. So we need to study their behavior, which depends not just on the material, but on the manufacturing process itself. So, example, if the fibers are oriented in an incorrect way, if the resin that is used to essentially coat these fibers to create the material structure doesn't flow properly, if the curing conditions isn't properly monitored, even microscopic defects, air bubbles, things like that, that can have a massive influence on the final performance.
And understanding all of these parameters, understanding the relationship between all these parameters, it just takes, at least in the status quo, it takes an enormous number of physical texts and an enormous amount of data.
And so the changes that we are looking at with AI is that we are looking at essentially a new hybrid framework linking all these levels of the testing pyramids together.
And we are looking at ways to also predict the mechanical behavior without having that full physical testing.
And so what that allows. This is a dedicated architecture which is being led in its development by Iker lan, which is also spearheading coordination for them project, which essentially serves as a foundation for the AI tool suite that we're building.
And basically we're managing data from material selection through manufacturing trials, all coordinated and fed into this AI.
And this is essentially not a replacement for testing, but it's also. But it's a way of learning much, much more from every test in a much more efficient way.
And one of the important factors there is also that it's also explainable AI. But what we mean by this is we don't want essentially to feed all this data into a black box, get an answer out of it, and then be having the engineers not even understand the process of reflection. So engineers need to understand why the recommendation has been made. And so the system is to be able to explain which of the variables influence the result and what changes might improve the performance.
And also what we're also doing is federated learning, which allows different organizations, different partners, to share their data to improve those AI models without necessarily exchanging and disclosing commercially sensitive proprietary data. So we get the benefits of having this AI tool learning and feeding and optimizing the development, preserving the independence and the commercial security of everyone involved, and having clear, intelligible answers that are going to be useful for the engineers and for the manufacturers and developers out of the way.
[00:09:14] Speaker A: Andreas, correct me if I'm wrong, but this AI tool will be used or is used in various pilots, right, in different countries? I would assume so, because that's the case in most projects. Would then the challenge that Marius project is facing with the different rules and regulations regarding digital tools, et cetera, apply to you too?
[00:09:39] Speaker C: I mean, this is a Horizon EU project, so the regulations when it comes to data security are rather strict. And we of course also have a dedicated team on data management and in trying to conform to essentially a shared data architecture framework which allows all of the partners to manage and exchange large volumes of performance data in an efficient way. Because we're talking about a lot of data.
We have a KPI handbook which standardizes all of the benchmarking across, like all the mechanical, the functional, the manufacturing metrics and. But we, and we are also, we also have a dedicated exploitation team looking at the security of sensitive data, commercial exploitation, et CETERA so that all of this information is handled in a responsible way.
[00:10:34] Speaker A: Okay, that's really good to hear.
Circling back to hydrogen, Maria, and the introduction in aviation, what are the elements involved in advancing hydrogen refueling?
[00:10:45] Speaker B: Well, so people could imagine that refueling is just connecting a hose to an aircraft, but it in reality is much more than that. You need to think about the entire hydrogen supply chain, where it's hydrogen produced, how it is transported to the airport, where it is stored, who is responsible for each step, and how the emergency services respond if something goes wrong. So then there is the refoiling equipment itself and the consortia are developing the whole refoiling process in all right. We are working on technologies to make leuked hydrogen refoiling much faster than currently is possible to move towards the refueling speed that are compatible with the normal airport turnaround times.
We are building a new high flow centrifugal pump by Linde, an intermediate station that conditions the liquid hydrogen before refueling and the fully instrumented hydrogen tank installed inside a real aircraft fuselage.
So at the same time we are looking at the hydrogen powered ground vehicles in the airport normally fill it with gaseous hydrogen because they allow airports to gain operational experience before hydrogen aircraft that are arrived. So really it's a combination of technologies, logistics, operations, safety and regulation all working together.
[00:12:13] Speaker A: Does the digital twin look like and how is it helping advance hydrogen refueling?
[00:12:19] Speaker B: So the digital twin is being developed by the centers, the Austrian center, cellcare and ait. It's really a powerful engineering tool and it's a virtual model of the complete refueling system. It simulates what is happening inside the pipes, the pumps, the hydrogen tanks. During refueling, different chemical, I mean parameters like pressure, temperature, hydrogen flow, and even how the liquid behaves inside the tank can be analyzed virtually.
So that means that engineering can test different operating conditions without always having to perform this expensive physical experiment.
So yeah, it's not replacing the real demonstration, it's more like doing them smarter.
[00:13:06] Speaker A: Yeah, and just assisting it or being next to it, let's say. Right.
Andreas, back to AI. What approaches are being investigated and utilized?
[00:13:18] Speaker C: Well, as mentioned earlier, one of the things we're looking for is to have a explainable AI. So the idea that we want the information to be generated to be, to be not just data goes in, something happens and then data goes out. But to have an AI that is able to inform us as to why it is doing the recommendations that it is doing and what are the processes that led it to this conclusion.
Specifically we are looking at what we call iTool. This is going to be the project's virtual testing platforms. And, and what it's going to do is it's going to bring together these three connected modules. So iTest, which is supposed to focus on understanding the material's behavior under stresses, or so mechanical stresses, heat, thermal stresses, electrical stresses. I process, which is looking at modeling manufacturing processes such as liquid resin infusion, fused deposition, modeling.
And essentially this is the part that is going to help engineers understand how these parameters are going to influence the final component. And then Imodel is going to use that information to try and predict how the completed structures are going to perform during testing.
And this is going to happen in three different operating modes. Like we have a training mode where it's going to learn from all of the physics based simulations that we've made.
It's going to have a thorough mode where AI and simulations are going to work together with simulations that are going to be refining those predictions.
And then in autonomous mode, this is when the AI has essentially learned enough from previous data to provide all of those predictions, all of those rapid predictions for the engineering problems and providing that improvement to the testing pyramid.
So basically the aim is to really have that improvement in the virtual testing environment where AI serves to create linkages much faster and to understand the knock on effects of certain changes in parameters over the whole pyramid, over the whole development process, which is where the project got its name. Like Pyramid AI Pyramid makes total sense
[00:15:48] Speaker A: and what sorts of new materials could emerge.
[00:15:51] Speaker C: So this is where it gets actually really fun for material sciences, because this is an aeronautics project, this is an AI technology project. But it is also very much a project in relationship and in conversation with material sciences. So today most modern planes are using composite materials already. So we've moved away from aluminum in most cases, because also composite materials are more ecological. They provide a lot of strength, exceptional strength really. But they're also much lighter, typically than aluminum and other such materials.
But what we're really interested in is what we call functionalized composites. So where we see if those materials, already lightweight and mechanically strong, can have additional functions built into the structure. So for example, can we improve electrical conductivity of these materials to provide better lightning protection? Can we build into it electromagnetic shielding? Can we support radar transparency? Can we add integrated de icing systems? Can we have a material that has baked in health monitoring, so to speak?
But also another key aspect in aviation and in general in R and D is going to be sustainability.
So can we look at some materials that are going to be kinder on the environment and kinder on the resources throughout the entire production process. We do a lifecycle analysis that includes things like bio based thermoset resins, recyclable thermoplastic composites. And also then what we need to factor in is things like toxicity recyclability, like what's the environmental performance of it.
And that we can look at what goes into the new composite materials and we can also look at the manufacturing process itself.
So looking at more efficient ways to manufacture large complex parts.
So improvements on what we call fused deposition modeling, which is essentially you could compare that a little bit to 3D printing, like at an industrial scale. It's not really 3D printing and my, my colleagues in material sciences are going to kill me to compare to that, but it's close enough for this conversation.
And the other one is liquid resin infusion, which is where liquid resin is drawn through and infusing these fibers that serve as a reinforcement to create those structures that have a lot more solidity and a lot more rigidity through this. And so yeah, we are using, we're developing these technologies for the new materials with four industrial demonstrators. So we have, we're working on a vertical sterilizer fairing aircraft door structures and also on wing leading, leading edge concepts for the wings.
So we are going to have all of these separate elements evaluated as well to see how they behave under aerospace applications conditions.
[00:19:15] Speaker A: Sounds quite interesting, Andreas, but I would like to go back a little bit to the part of the data and data management. We touched upon it already with you telling us how Hyramid is handling data. But I would like to ask you again both and maybe start with you, Maria. How are you handling data management?
[00:19:33] Speaker B: In our right, we have engineering data coming from the sensors installed in the hydrogen tanks during the demonstrations. We will collect information related to the behavior of the system during refueling, essential for the validation of the digital twin. And of course all this information is managed carefully. Partners share that data through secure platforms. And we have the data management plan that defines who can access different data sets, how they will be stored, which results can eventually be made openly available to the community and at the same time valuable know how, remain protected. So that's the way we proceed.
[00:20:21] Speaker A: Excellent. Andreas, I don't know if you want to add something.
[00:20:25] Speaker C: Well, it's very much the same reasoning that we have where data management is one of the foundations of the, of the project and obviously a very important parameter.
So we've developed this KPI handbook on the common benchmark for the evaluation of all of those parameters.
But we also are working, as mentioned, with industrial partners who need to protect commercially sensitive information. So we also have a data management plan.
We are also therefore very glad to have the federated learnings approach to AI, where partners can, independently from each other, pool their resources, even sensitive resources, and still have the added benefit of the AI learning from it and kind of improving on the models without any sensitive information being divulged. But also, whenever appropriate, we do follow fair data principles as well. So findable data, is it easy to access, accessible data, is it interoperable? Which means, can the formats of one data set be used by another way? Are there any compatibility issues that can be solved? And of course reusable? Is the data information used for one particular case scenario, is it going to be made available for other projects, for other research, etc. So this is something that has been of interest to the EU and to EU scientific research for quite a long time. And whenever appropriate, this is also something towards which Pyramid is contributing.
[00:22:03] Speaker A: I would like to ask you both if there has been a response to your project's developments from the industry already and if. Yes, what is that? Maria, would you like to go first?
[00:22:14] Speaker B: Yes. So, in the case of all right, airports, aircraft manufacturers, Harrigan suppliers and public authorities all recognized the Harrian could play an important role in the future aviation. But everyone also agrees that we need to start learning now. That's why demonstration projects are so valuable.
They allow everyone to gain practical experience before commercial hydrogen aircraft enter service. So through Allwright, we close work with other European projects and organize workshops bringing together airports from different countries to share their experience.
Instead of everyone solving the same problems independently, we are building on a common low lead space.
I think also it's important that airports are no longer seeing themselves only at transport hubs. Many are starting to look at becoming future energy hubs, integrating renewable energy, hydrogen production, storage and different users on the airport side. So that's a really sieve of mindset.
[00:23:28] Speaker A: Andreas, your thoughts?
[00:23:29] Speaker C: Well, the project is still in progress, so to speak, so it is a little early to have a full view of the industry's reaction and certainly not too early to talk about industrial adoption. But already at this level, we are involving quite a few industry partners in the project. So the consortium, which is bringing together 13 different organizations across seven European countries, it brings together research organizations, AI specialists, but also aerostructure suppliers. So Turkish Aerospace, Collins Aerospace, Potez Aeronautique, Sophitec, Aereo, and they're all helping to shape the research.
And what that means is that all of those technologies, they are being continuously evaluated by those people who are involved in the real engineering challenges of industry.
And it is already very encouraging to see that the idea of faster decision making, reduced physical testing, optimize resource digitization while maintaining our safety standards, while maintaining the quality standards and indeed improving on them, improving on those performance standards. This is something that we can already see the industry and our industrial partners are very, very interested in. And of course we are continuing to cultivate engagement with the wider community, such as with participation in events like EASA AI Days where we were highlighting the increasing interest in AI enabled engineering tools, but also where we are having the opportunity to network and exchange with other partners involved maybe in different projects, complementary projects.
And so I think we can already say that the industry is aware of the added value that this type of AI assisted engineering can provide and also of the benefits that could be provided by this new generation of materials in aerospace. Plus, of course the application goes beyond aerospace itself. Like many of the technologies that are being developed for this specific scenario, they have a way of finding themselves in a number of different fields that have applicability far, far beyond what they were originally designed for.
[00:25:55] Speaker A: Well, that makes total sense. I mean at ENLIT Europe, at the EU project zone, we have space engineering, we have space representatives with new, let's say advanced technologies that are very, very useful for, for the energy sector. So yeah, that makes total sense and it's very, very welcome. As we are approaching the end of our discussion now, I would like to just ask for the ending as a last question to both of you. If you had a wish or a message that you would like to send to the European Commission or to, to regulators both in Brussels and in the various countries, that would make your lives easier. And by that I mean easier the way the projects, the projects you participate in progress. What would that be? Maria, I will start with you again.
[00:26:47] Speaker B: Our message is probably that Europe doesn't need to start from zero anymore. Many airports have already completed successful hydrogen demonstrations. Now it's time to build up the experience. And the European Union Aviation Safety Agency EASA has a key role in play here. Projects like CO Wright are generated value technical and operational evidence to see a more modernized approval process across Europe, clearer guidance for temporary demonstrations and stronger collaboration between aviation authorities, environmental agencies and emergency services which are also key for this further deployment of the technology.
[00:27:31] Speaker A: Excellent, thank you Maria.
[00:27:32] Speaker C: Andreas, your thoughts very much on similar lines, especially on the issue of regulator industry collaboration. So the conversation between those that are developing new technologies and the ones that are regulating it, having that conversation ongoing, having that conversation be active, that's essential so that the rules, understandable, necessary rules, evolve alongside the technology rather than serving as an artificial bottleneck because policymakers are playing catch up because they don't yet understand how the game has changed with new technology. So that's one thing.
Specifically, the certification frameworks, they should be at least more flexible to enable the adoption of those new advanced materials and those new developments without compromising all of the safety rigors that we are still striving for.
But also, when it comes to digital and virtual certification, they are going to be key if we are going to meet the EUST 2050 Climate Neutral Aviation goals.
So having those lighter, more sustainable aircraft designs that we are going to need if we're going to achieve climate neutral aviation, that requires the smarter ways, the smarter tools to design and certify them.
And so this is where we hope that Pyramid and similar projects can play their part.
[00:29:02] Speaker A: Excellent. Thank you both very much. I'm sure that the people from the commission and from regulators are listening to our podcast, so they will hear you.
And yes, as I said, thank you both very much for this very interesting discussion.
[00:29:16] Speaker C: Thank you for having us.
[00:29:17] Speaker B: Thank you.
[00:29:20] Speaker A: You've been listening to the EU Energy Project's podcast, a podcast brought to you by Enlit and Friends. You can find us on Spotify, Apple and the Enlit World website.
Just hit subscribe and you can access our other episodes too. I'm Areddida Radimo. Thank you for joining.
