Ralph Wagner: You want to share data between devices, and you would like to do this already on board. So you do not want to get data from different vendors in different formats and try to combine them days or months later.

Stefanie Schuster: On the one hand, you need a CSMIM broker, which is the centralized component somebody needs to own. And on the other hand, you have ideally all components and devices being compatible with CSMIM.

Ralph Wagner: This is worth airlines looking at, right? How do airlines get the data of many different equipment types they have on board? How do they do this today?

Stefanie Schuster: The airline is, of course, the entity who has the biggest interest in seeing the cabin as one entity and having insights into whatever is going on there.

Ralph Wagner: Starting to think along that line. And also from a cost side, I would really ask airlines to look at the cost side of the past years — what does it actually cost you to connect services, or which data you do not have because you did not want to invest, because you saw the number and you thought, oh, maybe the data is not that interesting.

Stefanie Schuster: Because airlines, they have to adapt it in order to make this standard work.

Stefanie Schuster: Hello. Welcome to another episode of our podcast, LOADED: Talking Software for Connected Aircraft. And today's topic is why should airlines add CSMIM, a new ARINC standard, to their RFPs? Before we go into what this standard is about, let's talk about how we see the cabin of the future. Which use cases do we see there? What do we talk about?

Ralph Wagner: Let's start with some marketing claims. You want to know what happens in the cabin. You want to know which devices are functioning. You want to share data between devices, and you would like to do this already on board. So you do not want to get data from different vendors in different formats — sometimes you don't even get them at all, and you try to combine them days or months later. You want to do this in real time, on board, with as many devices as possible. So when we talk about devices, it can be the light, it can be the smart oven, it can be the seatback screen, it can be your avionics data that you use to, for example, predict when you will land. Whatever you're looking at in the cabin that produces digital data can be used by other systems and brings value to other systems — so that the crew knows when the oven is not functioning, that the essential entity knows how many screens are functioning in the cabin and which screens are not functioning, when you want to understand which lights are operating right now as a crew member, when you want to distribute avionics data not to every single entity — every single entity needs to understand how they get the avionics data — when you would like to make the avionics data centralized and available, then we are talking about a CSMIM. It's an ARINC 853 standard, and it is about connecting devices. Yes. And the messaging — it basically orchestrates the communication. This is how I would describe it.

Stefanie Schuster: And the centralised data, of course, also means that you connect between — like cross-functional — between different systems, between different vendors. Because you could still say, okay, in one solution you know which screens are on and off and you know which lights are on and off. But here it's about centralizing that and making this data available between systems. So showing the lavatory state on a seatback screen.

Ralph Wagner: Yes. And especially to do this with minimum effort and cost, because for sure you can connect two systems with each other. But it is needed that both parties are sitting together. Both parties are aligning which data should be exchanged. Both parties need to create the handshake. Both parties need to talk about the security. Both parties need to talk about what kind of infrastructure, cables, Wi-Fi they need. And one party needs to pay for that. So all of this is something you can reduce massively when you, as an airline, have a look at what you actually require from the participants in the network in the cabin, so that you have the most benefit out of it from an operational perspective and from a cost perspective. I think that is the main goal.

Stefanie Schuster: Let's go very shortly into a little bit more of a technical view. What are the main components we talk about here? Because we'll also talk about who should own it, who should pay for it, who should operate it. So what do we talk about here?

Ralph Wagner: Maybe let's start with the standard itself. So, initiated in 2019, and I would say the important participants — Boeing and Airbus — were driving this, with other industry members joining the group. And it is about how to think about the cabin network with a centralized architecture — how to exchange data. What you need for that is a box, only for the broker. It would be a minimum, small server, like a really little thing. An MQTT broker can run on a Raspberry Pi. So you need a server, you need a broker to run, and then you need infrastructure around it. You need to be able to connect the different devices to this box — either you have an ethernet connection to the devices, or, more and more, what you have is Wi-Fi. So you have a Wi-Fi connection from different devices, from different entities in the cabin, to this server. And this server is then basically orchestrating all the communication between all the devices. It's like a modern, cloud-like architecture set up.

Stefanie Schuster: And it's also a little bit similar to cloud architecture. It's about how devices can register, how you handle permissions, who is allowed to do what, and how you standardize the different data types — and how this data is then also locked down, like standardizing the entire infrastructure.

Ralph Wagner: I would say what is thought through is the idea that you have your different entities — your IFE system, or your screens, or your oven, your smart trolley, or your cabin crew application, or your galley, lavatory, etc. So you have all of these, and they decide what data they would like to make available and what interaction they make available. With this, they basically participate in this network. And what is also defined are roles. So an airline can think about — okay, these are the roles — or a trusted party can think about what roles this network should operate with. And then you decide, on your side, which role you give, which kind of access. You can also extend the role, but this is kind of the main thinking. And then, on the centralized hub, you need the certificates of the participants and the assignments of the roles. So that is basically defining what data I get, what data I can influence, what I can interact with. From the integration perspective, if you add a smart oven to such an infrastructure, and this smart oven understands how to connect to the network, and it receives the right to participate in the network, it can immediately receive, for example, the data of whether you are in flight or not. So this is something that immediately happens — you do not need two parties to sit together and discuss this. Everything that is already running on the server, everything that is already connected, is potentially available if you are granted the rights to participate in this network. So this is basically the technical basis of it, which makes this integration —

Stefanie Schuster: It's the ARINC 853 specification. Specification means it's not an implementation, right?

Ralph Wagner: It's not code. It's something — it's a standard that basically describes how you should build it. Yes, technically, everybody can build such a broker. Everybody can build such an entity. Everybody can own it. Everybody can put it into a vessel and agree with the participants to hook up on it, and with this, basically control the communication in the cabin. There is nobody enforcing bringing the hardware and bringing the broker on board — no, it's not like that. You can decide which hardware and which broker you put on it. And it's also not specific hardware, it's standard hardware. You don't need specific hardware — you can take hardware from many different parties and let such a broker system run on it. And such a broker system is something that I think airlines need, and maybe airlines would like to operate, right?

Stefanie Schuster: Yes. Let's distinguish the two topics. On the one hand, you need a CSMIM broker, which is the centralized component somebody needs to own — we can talk about that as well — and somebody needs to implement, so that needs to run on board the aircraft. And on the other hand, you have ideally all components and devices being compatible with CSMIM. So that's what airlines can't do all themselves, because there are typically some devices, some components, which are coming from vendors.

Ralph Wagner: Many of them.

Stefanie Schuster: Most of them, yes.

Ralph Wagner: Usually they come from vendors.

Stefanie Schuster: Seatback screens, whatever devices you have on board. So in such a case, you need to insist on CSMIM compatibility, right — to use your CSMIM broker in the best possible way.

Ralph Wagner: Absolutely. I think that this is in the interest of airlines. I think it's also in the interest of the participants of the network, and the suppliers to some extent — especially when they make use of it. But ultimately, it's in the interest of an airline. Because we didn't talk about the ground, right — what an airline would need on the ground to basically understand this infrastructure. And the data itself is centrally collected. It's centralized in every aircraft — you can collect it, you can collect telemetry data. So this broker can also be used not only for message exchange, it can also be used for message recording. Which means an airline has a perfect overview of what data is exchanged and what happens on the aircraft. An oven that does not work and sends a message to a crew panel can send the same message, can collect a message, can create log files, and this message can be brought to the ground. And on the ground, an airline can analyze the data, can basically combine the data of different aircraft, can even combine data of different — let's say ovens, if you wish — on board, if they connect to the same network and have a similar structure of data. At least the technical way you connect the data is the same. So I think this is a big advantage — to get a fast overview and to really do real-time maintenance, for example. So I think this is worth airlines looking at. How do airlines do this today? How do airlines get the data of many different equipment types they have on board? How do they do this today?

Stefanie Schuster: Well, today they are rather getting it separately from the different vendors — ideally already in a way that they already have their cloud infrastructure in place, and have their own data lakes in place, their own tools to make some sense out of the data. But it's not yet on a level where you can just subscribe to any kind of messaging inside the cabin and decide yourself what to use, and what to look at for interpretations for decision-making. And the airline is, of course, the entity that has the biggest interest in seeing the cabin as one entity and having insights into whatever is going on there, and also being able to use data and messages from every device and think about such cross use cases we talked about in the beginning. Of course, there are also vendors, airframers — there are different parties who have an interest there — but probably the airline has the biggest interest, because they have an interest in connecting all those components in the cabin.

Ralph Wagner: And I also think, from the mindset of wanting to understand what is going on in your aircraft — it's about looking at Tesla's development. You have a lot of different boxes on a car that are basically all doing their own stuff, and what Tesla was driving toward is a software-defined vehicle, and how they did that was by centralizing the data usage. You cannot 100% compare it to aviation — who is Tesla, is it a Boeing or the airline? But the idea, the technical idea, to basically centralize the data, to get the data from everywhere — like you say, what is the effort to do this with all suppliers, collecting the data from all suppliers, and then standardizing it somehow, and then you have even similar equipment. This is the way to go. And I think it's the airlines that would need to make the call and say, that's what we want to have. And to own the broker? Yeah.

Stefanie Schuster: You should own the broker.

Ralph Wagner: So I think a trusted party, or the airline themselves, should look at operating it. I'm quite sure the airline should have a say in it, and should be the one who is also basically granting access — telling, okay, which equipment goes on board, so that the airline has the right to say, okay, you and you can participate in this network, you get these rights for getting data and for interacting with data. I think this is something which would be a very typical job for the airline to do.

Stefanie Schuster: It's right now still rather a new topic. I mean, this ARINC standard, or the specification, is already on the market since some years. Still, such initiatives take time.

Ralph Wagner: We are in aviation.

Stefanie Schuster: The latest specification is now from '24. Yeah, I think it's getting more and more of a topic inside the cabin — probably coming rather from IoT, and then going more into passenger and crew use cases. But it's not the case that you already have a variety and selection of different brokers on the market, right? It's still a situation where it's not yet that established. So it's a good point in time to think about how to set up something like that.

Ralph Wagner: I would say, get knowledgeable about what it actually means. And it will not harm you to have equipment on board that fulfils the standard. It will not harm you to have smart overhead bins, camera systems, IFE systems, screens, display panels that support the standard. It will not harm you — it does not harm. Even if you don't have the broker, you do not lose anything. But if you are implementing such a standard later, if you're implementing the network later, you might not have the network set up — you might wait for getting your LEO connectivity with your Wi-Fi access points, and then you are hooking up your devices to the Wi-Fi and such a server, and then you get the data over that. Or you get another network on board. But to have the suppliers — bringing this standard into the picture, into the cabin — I think that might be a first step an airline could take.

Stefanie Schuster: Because airlines, they will be the enabler — or they have to adapt it in order to make this standard work. Right?

Ralph Wagner: Absolutely.

Stefanie Schuster: Okay. Put it already into the RFPs, even if you don't use it yet across the cabin.

Ralph Wagner: This is, I think, the way to go. So I don't think that it should be forced by whomever — by some airframer saying you must do that, and this is in, and all the equipment must be, or whatever. I think, ultimately, to really get adopted, to get into retrofit programs too, to get into the mindset of how you want the data exchange to happen between the entities — and I'm not talking about all the entities needing to exchange all of the data, it's not about that. It's about what brings the benefit of exchanging data, communicating with systems, having a crew panel that has a real overview of the systems, what they are doing and how they are doing it, and being able to collect this data as an airline and being able to analyze it. Starting to think along that line. And also, from a cost side, I would really ask airlines to look at the cost side of the past years — what does it actually cost you to connect services, or which data do you not have because you did not want to invest, because you saw the number and thought, oh, maybe the data is not that interesting. And I think airlines should look at that.

Stefanie Schuster: Now — we had this topic of who should own the broker, who should own the data. Also this topic of does it have to be one broker? So this might also be a topic to look into, because obviously having one broker owned by the airline again requires some cross-function, or some function inside the airline teams that goes horizontally across all the different functionalities — so looking into IFE, into IoT, into crew enablement, and so on, which might not be possible already. So in such a case, it would also work to have several brokers, right? It's not a go/no-go if you manage to get this one central broker in — and if you don't manage, then leave it.

Ralph Wagner: I think this is even the beauty — the beauty of the standard is that decentralized entity that actually manages all the messaging. It is something that can be two, it can be three, so you can create two or three such networks that are independent from each other, for example. Or you can hook up everything to one system, and then later change the system and look it up to a second one and replace it. So this is the beauty of that. What you're doing is standardizing the way all the equipment, all the entities on board, are available, are accessible. But this centralized system that actually orchestrates — that you could replace. Or you could have two of them running. And this is something very different from an approach where you are, let's say, enforcing on every single supplier and every single entity you have on board a specific technology, and then you want to update that — this is a nightmare when it comes to talking with each other, and they need to align with each other, and one is changing the way how it does that, and everything it's connected to needs to change. So this is something which is also like — oh, you are depending on this broker because it's provided by ABC, and then you are depending on that. But you are actually not, because you can replace it without replacing anything around your equipment. So I think, yes, you can run more, you can do it yourself, you can make use of available technology, but you own the data. So there are different scenarios possible, depending — the airlines, as we know, have different setups, right, and different background knowledge.

Stefanie Schuster: When we talk about standardization and openness — because this is like taking an open approach, right. There are other approaches which are more peer-to-peer, more about hiding and securing the data between different devices. CSMIM is really pushing for standardization and openness across the cabin. Do you think the industry needs an open-source initiative on such a topic?

Ralph Wagner: A tricky question, I'm not sure. It might help to think about that — to think about having somebody who is basically bringing the standard into open-source code, and then letting the industry participate in that. It would reduce the dependency on those who wrote the code. I think the question is whether somebody is really boxing everything so that you cannot get to it. I mean, you can still replace it with another broker — the broker itself and the standard are not that crazy to implement, so it's not something like a Linux system or so. It might be a good idea. It might be a good idea. It's important how the airlines react. So if the airlines start to show interest in it, if the airlines say, I would love to go down this way, I would love to have a centralized broker system that regulates all the messaging that happens in my cabin so that I have access to it, that I collect the data — I think, if it shows that this really is a driver for many participants, at that moment, I think it might make sense.

Stefanie Schuster: Okay, so also in this area, the airline is the driving force — interest from the airlines.

Ralph Wagner: Yes, absolutely. Because otherwise everybody will build in further proprietary stuff. Then you really —

Stefanie Schuster: — need some push on such an initiative.

Ralph Wagner: Or the best you maybe get is a standard on how to agree between two parties to communicate with each other. But then, still, you as an airline are not really into this — there are just the two parties who do this, and that's it. So, yeah. And also to the ground — I think we didn't touch on that. But I think it's important for an airline to have the data on the ground, right, and to hook up to a centralized system. So if you have a centralized system, you can connect to it from the ground. Your decentralized system on board can collect real-time data and message this data to the ground. So that is another big benefit that isn't even really described on the standard side, but this is what we see from a software perspective — that this is a real benefit to have this possibility. So yeah, airlines need to push.

Stefanie Schuster: So, put ARINC 853 as a standard into RFPs. Maybe same as today it's ARINC 429 — having this as a standard line, even put in by procurement, because it can just apply to anything that's in the cabin. Because the airline is the maker or deal-breaker of such a standard.

Ralph Wagner: Absolutely.