Framing the Disruption Landscape

Transcript

Rich Maslen:So what we heard there was about operational disruption, delays. I've seen some data that says that last year about 1 in 4 flights was delayed or people faced disruption. I'm not too sure about those figures because it seems like every flight I go on at the moment gets delayed. So maybe there's somebody out there that actually has a perfect fly-in and has no problems at all, and I'm, I'm taking one for that. Yeah, I think that's a good point. Them. I'm delighted to be joined up here on stage by Marc Lumpé from Air Transat and Jitendra Sindhwani from IBS Software. So what we're going to talk about over the next 15 minutes is about that disruption and what is happening, how it's affecting airlines, but also how they're adapting now to find solutions to better position themselves moving forward with it. So some real-world examples, Marc, from you, you know, in terms of disruption. How significant has the last couple of years been in terms of impacts from, say, weather, from geopolitical issues on the operations of Air Transat?

Marc Lumpé:Well, first I would like to say maybe you should think about your choice of airlines. We are running consistently way above 80% OTP. But I think you have to take a step back because the last couple of years have of course been challenging, but I think it is more about how we're driving this to the future. And the future I think is mainly going to be around technology and how we use that, because IOPS management historically has always been a triangle of tension between passengers, aircraft, and crew, and you need to find an optimal solution to that. while considering factors such as cost, passenger satisfaction, etc., etc. So the problem is that you're facing here is that you are basically moving in an n-dimensional space because the potential options are huge, and humans then have to find somehow the optimum solution, which they can't. It's just too complicated. You can put certain measures around this, you can put certain organizational points around this. This is why we are running this high OTP. But going forward, and hopefully we're going to have a little bit of time, but this is really short here to go into this. This is a prime example of a perfect environment in which you want to integrate quickly, and we are deploying this as we speak at the moment, true artificial intelligence solutions with humans. So, and this is really where this whole thing is going to, and this is going to be a real game changer to have more resilience in the system from the airline side.

Rich Maslen:And are you seeing true value to go into agentic AI and finding these solutions? Are they delivering real-world benefits to you?

Marc Lumpé:Yes, and this is the path we're going down. So currently we are developing because none of this, and I'm sorry, but none of this exists in the market. So, and I don't see this coming at least my timeframe coming into the market. So we're developing this with technology partners and we're deploying this right now. So, and the advantages, we can already see this, we already have Agentic Solutions deployed in the operations center. The overall platform, I hope that we're going to be able to deploy a first minimum viable product towards the end of the year. That is going to already help with full implementation sometime beginning of next year, hopefully. But this is going to be a game changer because you will be able for the first time to find truly optimal solutions. And if you think about the costs that delays actually produce, in our case, of course, there is a lot because we're flying a lot to Europe. So we are always subject to the European legislation. I mean, just one widebody long delay that you're avoiding, you're at around $300,000. So we're talking about millions there, and this is going to help leverage that, especially as running an airline at 85% OTP comes at a cost because you have to do certain things to protect that. And you can lower this cost to arrive at the same OTP using that technology.

Rich Maslen:So, Jitendra, from your perspective, you know, you're providing solutions to this. You know, what are you seeing from airlines? Obviously, we're finding a lot of airlines are more willing to collaborate with partners outside now. What has been their response to it? And, you know, where do you see this balance going moving forward in terms of technical solutions versus that whole human interface that we have?

Jitendra Sindhwani:Yeah, I think we are seeing a very diverse experience across the airlines, you know, depending on which airlines you talk about. I'm glad to hear you have an OTP of 80% plus. But there are a lot of airlines which are struggling with that a lot. So we see serving the global customer base that we today have, the diversity across the performance varies, you know, it's a lot. Clearly, I think the legacy platforms and the legacy technology that has been supporting these disruptions has created the biggest problems for them. I know you spoke about not having the solution. In fact, we just recently launched, and in fact, that's going to be announced formally now. We launched the first AI-first platform completely dedicated at managing disruptions. We believe the technology today can enable airlines to manage that part very well. You know, and I completely agree with Marc. This is not just solving the aircraft or the crew problem, it's solving the passenger problem also. And that's why when we look at what we have been able to develop now, we are bringing all the components of the airlines together, giving one view to the, to the manager who's supposed to take the decisions, actually give him the decisions rather, and help him save his time not looking at multiple screens or making multiple calls. But actually making the decision in a quick time. So clearly, the industry has evolved a lot. There is a complete desire to make it happen. Our solution can absolutely do that. And I'm probably trying to make a sales pitch here, but I should do that, part of my job. But definitely the solution exists now. I think the biggest challenge is the willingness to make it happen. There are structural architectural changes that organizations have to make on their side. And the sooner they do that, the faster they can reach the solution.

Rich Maslen:Marc, I want to come back to you with this because, you know, it's all too easy just to say technology solves everything. And you've, you've touched on the fact of having that human side as well. But, you know, where is there within the disruption management side that we need to keep that human touch?

Marc Lumpé:When you want to find a true optimal solution, you need to, you need to, you need to allow the system, and that has been something that was impossible until the advent of AI, break the rules. For example, you may have an MCT of 45 minutes, minimum connecting time of 45 minutes. You may have a solution that requires you to do this in 35 minutes, so you will have to break that rule. Allowing the system to break these rules would be dangerous because this is usually part of a broader negotiation between human actors within different parts of the organisation, organisations, because You may even have to check in with the airport for that. So the integration between humans and technology is very important there. That, as you said, that there are solutions being proposed that may even require breakage of certain rules that can be broken. But then it has to be done by humans to decide, do we want to do this? Can we do it? And how do we do this? And you could, of course, argue at some point of time, these agentic solutions can also negotiate among themselves, which to some extent they can. But you still will need humans to be in the loop to judge if this is something that is actually desired. So the true integration between human agents and artificial agents, as by the way anywhere when you talk about agentic AI, the integration between machines and humans is going to be the key to make this work. Which, and just as a side note, will require huge investments into training and cultural change programs because we have to take the humans along with us on this. Because I'm drilling this message quite a bit and you probably have heard me saying that already. It's for me, all of this is, although we're talking always about technology, it's 30% technology, 70% humans. We don't take our teams with us on this journey, we are going to fail. And that's really, I think, the key message here.

Rich Maslen:I think that's something that we have to spread out everywhere, the fear of technology taking us over, the important role that humans will, you know, keep within that. Yes. You know, data has been such an important thing, and airlines hold so much data on travelers and everything, and it's become more important how they use that. So, Jitendra, from your perspective, you have all this information, but a lot of it is still siloed within the airlines. You know, how important is data orchestration in creating a genuinely intelligent system moving forward?

Jitendra Sindhwani:That's going to be the start. But I will also say that, you know, that everybody knows. And as Marc said, there is an investment that has to be made to bring the data together. But I think the important part is that where does AI really play a role? Is it playing on the edges? Is it enabling? Or is it actually driving the whole, you know, operations platform itself? In my opinion, the AI has to play right from the beginning, which means that it has to be a learning curve. There has to be an ability for it to orchestrate the data right at the beginning itself and make the decision. Yeah. Bringing AI-first platform truly is how it's going to be solving from here. Just reacting to one point that you made about 30-70 ratio, I think during the change part, the technology will be 30, humans will be 70. As it moves forward, the reverse will happen with 70% of the decisions or even 80% of the decisions will be made by the machines. The humans will be there to make sure that it's actually going in the right direction. There's a big part of training of humans that will be needed here in terms of how do you live with the machine. Data is needed, systems are going to be there, but this change will have to be very carefully managed.

Marc Lumpé:If I may, this wasn't about the quantity of decision taken. This is about the effort that an organization has to put in because deploying technology usually is the easy part. You have to make sure that the humans that are working around the technology actually embrace it, buy into it, understand it, can use it. And this is why when it comes to an overall investment point of view, that is where the 70% come in, that you have to give also your management time to the human component because this is going to make the difference between success and failure.

Jitendra Sindhwani:I completely agree with that. That's why I said that's the transition phase where that'll happen. But the goal should be that the machine should make more decisions than the humans.

Rich Maslen:Obviously, the process is changing very quickly and the adoption of technology is moving really quickly. You know, Marc, 5 years from now is a long time in terms of the development, and we're framing this in talking about disruption, you know, disruption management. You know, do you believe what areas of this will become completely automated within the next 5 years? But where do you think airlines still need to make that significant manual intervention even at that time?

Marc Lumpé:So, I mean, How I see this in 5 years, I'm going to answer this in a second. It's just, I think anybody who's telling you how this is going to look like in 5 years knows is either delusional or doesn't understand the technology. So whatever I'm telling you, we're going to talk again in 5 years. I can guarantee already right now I'm wrong. So here we go. But what I think from a conceptual point of view, what you would see is that the organizations that we're seeing today are going to not just simply not exist anymore. Yeah. We are planning to start taking the first departments fully AI-native next year. And fully AI-native means that you are destroying actually the existing organizations completely and build them anew with AI at their core. And that also means that you will have different touchpoints between machines and humans than you see today, because organizations today have been developed around human limitations, which in the future we don't have anymore in that way. So you will have completely different touchpoints. You will have humans still in the loop, in my opinion. For regulatory reasons, that's still going to be, but that's more of a housekeeping issue. For decisions that involve human-to-human interactions, but also, and I think this is going to become increasingly important, for ethical decisions, because sometimes you may have multiple states that are equally good objectively, but certain are more desirable than others, be it from an organizational cultural point of view, be it from a purely human point of view. And you will have to define these points where you want to put the artificial intelligence into guardrails to ensure that it is aligned to what we actually want. And that is going to be still requiring a lot of human intervention because you cannot leave this on autopilot. You would have to put in guardrails, but where these touchpoints are going to be, quite frankly, let's talk in a year from now. I hope I'm going to have a bit more experience by then, because as I said, we are venturing into uncharted territory here also as an organization. We have a good idea where we want to take this, but we just don't have time to talk about this right now. But as I said, in 5 years, we'll see. Let's see in 1 year.

Rich Maslen:Okay, well, we'll have a repeat debate, no doubt, at our event here in the Americas next year. Jitendra, we're running short of time, so finishing Starting off with you, disruption is now a structural part of travel and we're well used to it now. From the outside point of view, obviously, if I speak to Marc, he's going to say about how great his airline is versus others and what he's doing better. What capabilities do you think airlines will develop that will outperform others, that will make them stand out than being better than one another over the next few years?

Jitendra Sindhwani:I think the ever-going thing of technology transformation has to continue. In fact, it has to speed up. And that's where breaking away from legacy architecture itself, you know, the whole functional architecture or the tech architecture that the clients today have, I think that has to completely change and accelerate. And but I— but in this new age of AI where the change has just started to begin, and I agree with Marc, we don't know exactly where we'll be 5 years down the line. The humans and machines will have to be very carefully You know, you will have to bring the 2 sides carefully together. You cannot tilt on one side completely hoping that the other side is going to be managed on its own. I don't think that's going to happen. So the change part will have to be very, very carefully crafted from here onwards. But the— it has to be done in an accelerated manner. I don't think airlines will have the time to wait for another year and then make a decision on that, because that's going to decide between the winners and the losers very soon.

Marc Lumpé:Maybe it's the last word. I think I fully agree. This is accelerating so much. The stuff that we're deploying today, if you'd asked me a year ago when we were going to deploy the stuff that we're deploying right now, I would have told you it's going to be another 3 to 4 years. And this is just getting faster. What I'm seeing when I'm in the tech scene, the stuff is really accelerating faster and faster. So you want to succeed in the market, that's something I can guarantee. You want to succeed as an airline in 5 years from now in the marketplace, You have to drive this as hard as you possibly can because all the startups that are going to be coming up in the next couple of years, plus the airlines that are driving change right now, they are otherwise going to destroy you. That's for sure.

Rich Maslen:Well, that's a good clear message to end the conversation. Really good point there, Marc. So thank you very much for the short time we've had here. Please, can I have some appreciation for my 2 panelists?

Copyright policy: All transcripts on this site are the copyright of CAPA - Centre for Aviation. Our reproduction policy is as follows: you may quote up to 400 words of any transcript on the condition that you attribute the transcript to CAPA - Centre for Aviation and link to the original video page. All other use is prohibited. While we aim for 100% accuracy in the transcript, there may be some minor transcribing errors.