#105 – GSM-R Is Dying. Are Railways Ready for FRMCS?

GSM-R is on its way out. FRMCS — the 5G-based successor for railway radio communications — is on its way in. But the migration from one to the other is anything but straightforward.

In this episode, Doc Frank talks to Rob van Looveren from the Netherlands, a railway consultant with hands-on experience in FRMCS specification work for freight trains through the European-funded FENIX project. Together they unpack what the migration from GSM-R to FRMCS actually involves: the technical challenges, the planning complexity, the freight operator funding dilemma, and why a bistandard onboard approach is likely the only migration strategy that really works.

The Netherlands context provides a useful European lens — and the comparison with Australia's situation turns out to be more interesting than you might expect.

If you find this episode useful, the best thing you can do is subscribe, share it with colleagues, and leave a comment. It costs nothing and makes a real difference.

When it comes to advancing your understanding of modern train control technologies, nothing beats world-class training. I've partnered with Informa Connect Academy to create what I believe is the most comprehensive portfolio of online training courses in advanced signalling available anywhere. Find out more here: https://informaconnect.com/academy/rail/

Transcript

0:00

My conversation today is with Rob van Loferen from the Netherlands, Arabic consultant, and we're talking about the situation in the Netherlands around ERTMS, ETCS, and especially the migration from the GSMR radio system, the original one for ETCS, to the new upcoming radio standard FRMCS. Please like, share, and subscribe to this channel and this episode. Please enjoy my conversation with Rob van Loferen. Rob, welcome to the show. Thank you for having me in this podcast. Shall I introduce myself? Yes, please. I was going to say it's the first time that you are on the podcast, so please let us know who you are, what you do, and what your relation is with ERTMS and ETCS and radio and so on. Yes, so my name is Rob van Loferen. My background is similar to yours, I think. I studied on the Technical University in Delft in the 70s, so to say. After my studies and after military service, I went to work with KLM, our Dutch air company, and there I got a bit in automation. And after a few years, I left there and I went to work for PTT Telecom, which was a governmental organization at that time. And later it became KPN Telecom. It was privatized. They didn't do it in the best way, but anyhow. And so I worked there for 23 years. And then they let me go because they had to reduce the company size, please. And I never got a paid job after that. Because it was in 2008, there was the financial crisis. And then I thought, yeah, what to do? And then I was always interested in railways. And that was, say, 2012. In 2012, I started to reading subset 026. And it took me two months to fully understand what they were saying. So that was my start with ERDMS. And then the year later, I said to my friend Peter Wankers, okay, I have a nice idea, we have to focus on freight trains. And to see if we could fill in something for level three train integrity. But that's still not yet filled in that part. But we have developed that our smart freight train ID to put all sensors on wagons and so on. But that shall be waiting for FRMCS to be feasible to create that in Europe. I think it's not an issue, maybe for Australia. Okay, so that's a bit my career up to now. So I am also involved in European funded projects, which was all about transport and logistics. And I'm still involved in the because after that project, we raised an association without profit. And I'm a member of that association. And we prepare also new calls for all these funding projects in Europe of the European Committee. So that's now a new challenge. But that's just part time. And that's now going on. And before this meeting with you, you said the one you want to talk about FRMCS. So with the help of AI, I made some surveys on where we are now with FRMCS. And ERTMS, ATCS also. So that's it. Yeah, so let's talk about FRMCS. You mentioned this. Can you explain to the listeners what this is and why it is so important? Yes, we have ERTMS operational in Europe, but it's all based for the train track site communication on GSMR, the second generation mobile network. And since the provider said, Oh, this technology, I stopped producing that. So we have to go to change to another network. And the 5G, 6G technology is then the choice, the new choice. But the big difference between GSMR and 5G is that GSMR, you have to connect to track site to the radio base station via the telephone network and GPS facilities. So that's with the new network technology 5G, 6G and so on. We change to a technology which is internet to the edge. So you can then skip this telephony protocols and so on, the technology and directly built on the internet protocols. But that's quite a change, because we have a lot of trains already running with the old equipment. And that transition to FRMCS, because that is based on 5G, of course. That facilitates a lot of things for the railways. But the transition phase, we have a lot of trains running on GSMR. And on board, you have your radio part of it to communicate. But that shall be replaced by the FRMCS 5G equipment. And also the radio base station needs facilities to communicate with the 5G network. So that's quite a challenge, that transition in Europe. Yes, and there's only a certain window of time to do this, right? I'm not quite sure what the timelines are in the Netherlands, but there is an end date for GSMR support that I have heard of. And it will also take a bit more time until FRMCS is fully specified. I think they're currently running tests. But tell us more about this. And maybe you know best about the situation in the Netherlands. So what are the timelines in your country? So when I was in the Phoenix project, I and Pieter Verkuts helped UIC with making these functional requirements specifications. Because they only set it up for passenger trains. And we told them, oh, we know all about these freight trains. So we can help you writing the specifications for the freight trains. And so we did during the timeline of that Phoenix project. But the further development is still going on. But next year, they will start really a few trials. And so I don't know the exact flow of all those activities. But also the specification for baseline four will come. That will integrate also FRMCS. And then, yeah, I think from next year on, transition phase will start. But that will run up to 2035. So that's almost 10 years transition phase. Okay, so 2035 is the end date for GSMR? Yeah, but the transition is quite complicated, because you have already equipped trains. And when you're not following a specific plan that you have the right trains for your corridors, trains can't run. So that's very precise planning of also getting the equipment, the trains ready for that. Yes, yeah, but it's important to know what the timelines are. So if there is a hard end date by which all the GSMR must have been taken off and switched off and decommissioned, that's an important milestone to know. So that you know how much time you've got available for your migration, right? That's of course, yeah. Yeah, but reality tells us also that there will also be complications. So yeah, we'll see where we are in 2035. And maybe on some parts, there will still GSMR be operational. They just keep it running. They can use old equipment, which comes free to replace broken parts or whatever. That's also for telecommunications possibility to keep it running. Yeah, I agree. I think it will be a slow and painful death to GSMR in some areas. And I could imagine that drivers that just can't get the migration done in time. Yeah, they will try to couple something together and keep it alive somehow. Without industry support, I don't really know how that works. But maybe they have some in-house resources that are close to retirement, but still can be convinced to do a couple of years more. And yeah, but I mean, eventually they will have to go to the new technology. That's of course, yes. But they cannibalize just all the equipment to keep the operational parts running. Yeah. But in Europe, the main focus is on our 10D corridors, the main corridors for the railways. We started with nine corridors, but now we are maybe at 12th or more corridors defined for the mainstreams of especially the goods transport. Because passenger trains, they run normally only in national. And there are just not so many who are running international in Europe. Because due to all extra equipment to install for the different countries with different signaling systems. That's the current situation. And we will to go to one railway area in Europe, this ERTMS as the security system, the signal system. Yes, yes, that's the objective. But it's not going as fast as the Europeans were hoping for, right? Oh, no. No. Then, you know, we have the European Railway Agency. Yes. And they were moved in a position that they will control all developments and introductions in 2016. But then already there were ERTMS equipment installed on locomotives. And the national authorities then said, oh, we are responsible for the safety. So could you adjust that and that and that? So the Elfstone and Siemens and who delivered the equipment, they made changes. And with those changes, the system became not interoperability anymore. And that's the era, as we call the European Railway Agency started in a position that they were challenged to harmonize it in all of Europe. So that was not so much ideas about how you do version control of all these specifications. But there are different suppliers for the equipment on board. And I've seen the first built ERTMS equipment for the trains. So it's quite a cabinet. A lot of electronics nowadays is getting much smaller due to, thanks to ASML, who made the possibility to make these very tiny chips. So it's created all kind of a legacy for that starting situation. And that hinders, of course. Yeah. You mentioned you wrote some specifications for freight trains. What specifications were those? What was specified there? That was focused on the use of the capacity of the FRMCS network. So they could build it for what needed. But for the passenger trains, it is much focused on providing internet access to the passengers. But for freight trains, you don't need that, of course. No, but you need the safety critical stuff, right? The ETCS. Yes, that of course. But we also specified all the operational processes for freight train operations. And we wrote this also with the IDs of our smart freight train ID that all wagons are also equipped with censoring stuff to see if there's something going wrong with brakes or boxes, all that kind of stuff. And that can then be communicated to the train driver. But that's just our idea of the smart freight train. That's an extra. Okay. So you said that the migration may already start next year. So to my knowledge, they are still working on the final specifications of FRMCS. And they're currently doing tests in different countries to inform any necessary changes to go into the next version of the specifications for FRMCS, hopefully the final version. Yes. So what makes you hopeful that the migration can already start next year? Because obviously, in order to be able to do the migration or to start the migration, you need to have products. You need to have products for the radio system itself. But you also need for the ETCS track site system, the radio block centers to connect to FRMCS. And you also need, importantly, you need equipment on board of the trains to interface with FRMCS. So what is the status there? Do you know? I asked to JetCPT if they already have the radio equipment for FRMCS for onboard the trains. And I got an answer. No, it's not yet there. Yeah. Yeah. But these are not the big questions. The industry can really fast produce that equipment. That's not the main problem. But when you make a transition, you have to make the trains ready for FRMCS. But you also need to make the radio base station, the RBC, ready for FRMCS. Yes. And that complicates the transition because the radio base station, okay, that is along the long lines. You have different of those stations. That's not so big number. But the trains running on that corridor, they should all be prepared. And they are owned by different companies. So that can be tricky. And in the worst case, after the transition, some companies couldn't run on that corridor then. So that's a planning issue or an alignment issue and so on. Do you know how the migration strategy in the Netherlands looks like? What the sequence is of fitting vehicles, maybe dual fitting vehicles? That I don't know in detail, because it changed in various times. Now they started just in the north of the Netherlands. This, I think the line between, not just a small line, they will start near Delftsheil, I think. So that's very isolated in the north. Not so much to do with other rail lines and so on. Because when you start in the center of the Netherlands, in Utrecht, for instance, yeah, they're all, most all trains in the Netherlands pass there. So in a corner of the Netherlands, where not so much trains running, it's safe to start, I think. So then you can see what troubles it gives after you've become operational. Yeah, I mean, there are probably multiple possible strategies for migrating to FRMCS. I think the most likely will be that the trackside equipment will be transitioned gradually from GSM-R to FRMCS, which means that you will have a mix of GSM-R fitted lines and FRMCS fitted lines. And that means that you need trains that can work both on GSM-R and on FRMCS. So that's at least how I would foresee it, which means you need an onboard system, an onboard radio interface that can interface both with GSM-R and with FRMCS, and can switch dynamically when there's a transition from a GSM-R area into an FRMCS area. Is this how you see it as well? Or do you think that there's a different strategy in the Netherlands? I looked up things and that is just a complicated transition, because you should go, you prepare everything and then you make the transition and then you're in a new situation. That's what I got as an idea from what I read. But yeah, making the trains ready at a time and it's because parallel, there is a kind of solution, but I don't have it in my head now what that precisely looks like. Because with the RBC you can't run them with both networks, I think. There's also a problem. When it comes to your training on advanced signaling and train control, nothing beats a good quality. That's why I partnered with the Informa Connect Academy to create seven fantastic online trainings that I believe are the best and most comprehensive portfolio of online trainings in this space in the entire world. You see the seven trainings that I'm offering through Informa in the boxes here. And you also see at the very top a web link where you can find information on all of these training courses. If the education and your personal development in the field of advanced train control is of any importance to you, I recommend highly that you check it out. And I would look forward very much to welcome you in one of these training courses. All of them have interactive elements, so we can actually meet face-to-face in those trainings, even though they are online courses. It needs to be seen to be believed. And now back to the conversation. So I think the TrackSight transition will be in line with the RBC control areas, I think. It has to be. But that again leads to a situation where you will have a patchwork between GSM-R fitted areas and FRMCS fitted areas. And the way to deal with that is to have trains on board equipment that can deal with either of the two radio systems. Because the other alternative would be a dual fitment of the TrackSight, and I don't think that's a very feasible idea. Then you would need to have an RBC that can switch between... No, that doesn't work. That doesn't work at all. When I think about it, that doesn't work, because then you have to stay safe. Yeah, well, and you would have an RBC area controlling different trains, and some trains run on GSM-R and some trains on FRMCS. So the RBC couldn't switch the radio systems back and forth depending on which train they're talking to. That wouldn't work. So I think we're back to the other idea of migrating that you transition a TrackSight area from GSM-R to FRMCS and that's it. And then you do the same thing section by section by section in accordance with the RBC control area. So I think the RBC areas will be the cells, if you will, for which the radio system needs to be migrated. And then you need capable trains that can do both. And this will be necessary anyway, because if you think about freight trains running international, even if you say the entire part of the freight train journey in the Netherlands, it's all migrated to FRMCS. And then the trains goes across the border into Belgium, and they haven't migrated yet. And then again, you need the train to be able to deal with both FRMCS and GSM-R. So I'm here in Australia, I'm a very strong proponent for having such a bystander approach on board, that the trains must be able to deal with both radio systems and can basically switch between the one and the other. Because I believe that's the most feasible migration strategy, maybe even the only that really works. Yeah, I already discovered that you have your communication boxes in onboard trains, which can switch between satellite and the network and so various possibilities, which is not common in Europe. But the Netherlands, we have a very dense railway network for passenger transport. So for us, it's a horror that migration plan, because you encounter a point where just a line can't be run by all trains for a time, I think. But I don't know the details how they will do it. So we'll see. There are no technical details, which will block something and so on. Yeah, I think it will go back to the old problem that ERTMS had all the time. For the migration strategy to work, ideally, you would have to fit the entire train fleet with a bystander onboard system before you start the trackside migration. So that means that the onboard equipment needs to invest first. And then you have the problem, as you mentioned, you've got many different operators, some of them quite small, just with just a few locomotives, and they now need to touch their ETCS onboard again and need to upgrade it to be able to deal with FRMCS as well. And they will have financial struggle to do so. They will say, well, I understand that this may be necessary from a national point of view, but who is going to pay for it? I can't afford it. Yeah. And that's the same struggle that they had previously with fitting ETCS in the first place. Who pays for it? And if they have to pay for it themselves, where is the return on their investment? Yeah, retrofitting a locomotive is very costly. And yeah, so that's a major problem. But for we have in the Netherlands, for passengers, we have the Nederlandse Spoorwegen, NS. They run the main lines and a few sidelines. There's another train operator. So the major part is only of NS and sidelines is now also for other companies to run. But for the freight trains, we don't have our own freight train operator. It's all German or from abroad. And those trains mostly run from the ports directly to Germany or to the south, because we are a small country. Yeah. It's just 200 kilometers from the port of Rotterdam to the border with Germany. Yes. So that's nothing compared to Australia. Yes, it's much shorter distances, but still, migration is quite complicated. I think in Australia, we may be in an even better position than Europe because we haven't got this much of an installed base for GSMR. So we've got some railways that are just starting out and they can start out straight with FRMCS. So they don't need this migration. They just go to FRMCS and then they're fine. I think there's another difference between Europe and Australia. So in 1989, the commission has said, okay, we will, after the fall of the wall in Berlin, maybe you know that, then they have said, we make one railway area in Europe. All the national networks are connected with each other. And the situation in Australia is different because you have near around Perus, an area where part of the population lives, then you go to the southeast, Sydney, Melbourne there. Those are the main areas for Australia. But in between, you have also the plains, the outback, and that separates these areas, in fact, from each other. Yeah, but we do have a railway there. So it's not completely separated. Yeah, but in the European view, they have no idea of these rural areas without any telecommunications except satellite available. Yeah. I mean, maybe the interoperability topic is a bit easier here in that you can't have that many railway jurisdictions that are running their trains back and forth. So for instance, if you are in a country like, say, Switzerland, somewhere in the center, or take Germany, for example, they could get trains from the Netherlands, from France, from Denmark, from Austria, from Switzerland, from Belgium, and then trains coming from even further countries, from Eastern Europe, from Belarus, or from the Latvian countries, also the Baltic countries. So in Australia, the main topic for interoperability are really the freight locomotives. So the network connecting the cities is basically just operated with locomotives. So once you fit those locomotives, then the interoperability challenge is if those locomotives go into cities and operate on parts of the city networks. That's where we need interoperability. And then, of course, for the interstate network itself within the interstate network, yes, we need interoperability as well. But that should be an easier task compared to Europe, where you simply have more players involved, more different railways involved with different histories and different signaling systems historically and so on. So it's less complex here, but still not easy. No, no, but that's still a big difference with Europe. We are all connected within all European countries. And what separates in Europe is the Alps, that separates Italy from France and from Switzerland. Switzerland is a bit apart, because they are not a member of the European Union. So it has a kind of an island, but for railways, that's not a problem. And they facilitate much by boring these tunnels through the Alps to go to Italy, because the big economic parts of Europe is the northern part of Italy, and then Germany, Netherlands, that there is the Ruhrgebiet. You may know it in the west of Germany. Those are hotspots for economy. So there, the first 10 T corridor was also projected from Rotterdam to Genova in Italy. Yeah. Yes. You mentioned earlier, the smart freight train concept that you were working on. Can you tell us a bit more about that? Yeah, that we started, it was triggered the idea, because it is, yes, the train integrity was mentioned, but they didn't solve that part. And it's still kept outside. But we did, yeah, we talked to people, how do we connect all these freight wagons with the control equipment of the train driver, and all local solutions for onboard the trains were not feasible to cover the whole train, because technology didn't accommodate that, because the train is quite a length, and we run them trains up to 750 meters, but you have much longer trains, you can make a network on board all these wagons. So then the idea of FRM-CS came up, and my colleague said, he is more the specialist for the radio, he said, that's the solution. We can easily, because the internet is to the edge, so you have internet access direct, you should create it directly on board your wagon. And then you can easily make your applications with the technology. And then you can, via FRM-CS, you can easily connect with the unit in onboard the locomotive, the head of the train. But it's still an idea on paper. Okay, okay, okay. So you would probably, the issue was train integrity for trains that consist of wagons and can vary in length and so on, that has always been the big problem. And the trick would be, you would need some kind of equipment on the last wagon, that basically tells the locomotive, I'm the last wagon and I'm still behind you, I'm still there, you haven't lost me. But that would mean that you would have to put this equipment on different wagons, depending which wagon is the last one. And then the locomotive need to have the other part of the communication receiving signals from, and then you go through a curve and then the back of the train is closer to the front because of the curvature and so on. So there are, yeah, it's not a trivial issue, that's for sure. Now, so then we have to go to the United States where everything started. There, at first, long ago, trains with a caboose at the end of the train where crew was on board. But then they could get rid of this caboose by creating the end of train device. So you mount that on the last wagon, connected to the brake pipe, and you have a radio connection with the locomotive. But when distance, no, okay, this radio, you can direct radio, you can cover up to 10, 20 kilometers at the max, but that's okay, that you can do. Now, I think, and that end of train device is much used in the United States, but in Europe, maybe few, but not that many. But you don't have equipment on all the wagons in between. So with the idea of detecting those anomalies occurring on railway wagons, you should install sensors on each wagon. And then you have the problem of communications to the train driver. That and creating a network which goes from train to train with the communication, yeah, that we found out that was not feasible, that they couldn't accommodate so many hops or so and then it will get blocked to communication. So it didn't work. So with FMCS, you can have communications from each wagon via the FMCS network, the 5G network to the locomotive. That can work. But you need the network. So in the outback, you have no network. So it won't run. Oh, no, we do. We do. We do. We do. You would be surprised. We are not that far behind in Australia. We actually do have a radio communication network along our interstate rail network. And yeah, it's only 4G, but FMCS will work over 4G. So we actually do have a strategy for doing this. So it's not quite as bad. Is it using satellites? No, no, no. It's normal. Just antennas. You have antennas. Yeah, yeah. Normal radio infrastructure. Yeah, yeah. So there was a deal for the interstate network between the network operator and the biggest Australian provider for mobile telecommunications that they can build their infrastructure, their radio infrastructure along the rail corridor. And the railway can use the radio for their applications, for their signalling applications, such as FMCS in the future. So it's quite a good deal for both sides. And that leads a long way towards getting a solution there. So we are quite good in that regard in the meantime. It was different a few years ago, but now we're kind of okay, I think. Okay. But that's quite an investment, I think, because you have to install many antenna locations. Yes, yes, yes. But it's the biggest telecommunication carrier in Australia. So they always had the best coverage, including in areas where other carriers didn't have coverage, didn't afford coverage. And they saw it as an opportunity to become even better, to have an even wider coverage. And if you have something along the rail corridor, then that's obviously good for passengers on the train that run there. There aren't many passenger trains anyway, but the ones that they do, they have now better reception as well. Yeah. And of course, with radio systems, you also have coverage a few kilometers away from the rail corridor. So you can start building on that. Good. So how do you see your activities over the coming years? Do you think you will have any involvement in that FRM CS migration or? No, no, no. I'm more involved in the Phoenix Association, preparing all kinds of proposals for funded projects in transport and logistics. And the project we did run was creating a cyber secure data exchange between data spaces, so to say. And that we will deploy in recent other projects to create telecommunication links, which are standardized, so to say. Because with that, we created the possibility that you can create those trusted data exchange, very cyber secure and so on. But you can then use your own messages to transfer, because the concept issue between these connectors, we call that connectors, you have one standardized message, but you can put any type of message in a payload. And that makes it flexible. But there's still a lot of work to do. Yes, yes, yes. You did mention you were also involved in some European research projects. Do you want to talk about that? Or is that top secret? No, no, I was not involved in research projects. No. Okay. No, this is the only, we did the Phoenix project lasted four years. Yeah. So that was, and I was much focused on interoperability, because the main challenge was to build the IT technology. But yeah, using that IT technology, you should know how it runs in real life. You have to look at how parties work together, collaborate together, communicate together to see what is needed. And then you have to agree on for the supply chains on standardized messages. United Nations did a lot of work in the past already. They started a very long time ago with the, oh, come on, ADFACT. And then in that time, it was a lot of port to port communication. So there were only possibilities of least lines with low transmission speed. So they made a standardization, which you can make your messages as small as possible. So you have really affordable transmission times. But nowadays, that's no problem anymore. Okay. Well, thank you very much for giving us a bit of an insight what's happening in the Netherlands. So let's hope that they sort out the challenges of the FRMCS migration and they can successfully phase out the ancient GSMR technology, right? Yeah, sure. Yeah, that will be still a challenge because in Netherlands, we keep a lot of talking about these things. And we have to make a step to practical approach to really make that transition. Yeah, that's true. Thank you very much for your time, Rob, and all the best. Bye-bye. Thank you for having me. Bye-bye. Bye. Hi, it's Doc Frank again. I hope you enjoyed this conversation. If you like the content, then the best way you make sure that you don't miss out on any of the future episodes, just subscribe to the channel. By doing so, you also get access to all the previous episodes, which I'm pretty sure there will be many that are of interest to you. Another thing that you can do to help not just your own environment, but the entire industry is to share this episode with friends and co-workers because it helps more people in our industry to get educated and an educated rail signaling industry is a better industry. With that said, thank you very much. Hope to see you again soon and bye for now.