WELCOME to Freemium Subscription
exclusively by New In Signal...

ASK THE DOC | Monthly articles by Doc Frank Heibel

HIGH PERFORMANCE SIGNALLING | Doc Frank Heibel

Does ETCS need a supply strategy?

[restricted_content]

First of all, welcome dear reader to my column about “High Performance Signalling”. I hope you find these articles insightful and instructive and will do my best to give you premium “food for thought” as well as practical and applicable advice. Yet it is prudent to add this…

Disclaimer: Any individual or business decision you may make based on the content of this article is solely and fully your own responsibility, and you cannot under any circumstance hold me, my business, or the publisher of this newsletter responsible for any consequences of such decisions.

Coming to today’s topic, I am specialised in two technologies for High Performance Signalling: CBTC, the Communications-Based Train Control system, and ETCS, the European Train Control System. I found that when it comes to performance-enhancing varieties of ETCS, they become more and more similar to CBTC. It is therefore that when I study something for one of the two technologies I also check the relevance to the other.

I first came up with the concept of a supply strategy for CBTC. The reason why I believe a supply strategy is needed before starting the introduction and later rollout of CBTC is the lack of interoperability between CBTC products from different ‘mainstream’ suppliers such as Alstom, Hitachi, Siemens, or Thales. That means a train fitted with onboard CBTC from one supplier cannot operate on CBTC wayside infrastructure delivered by another supplier. (There are exceptions, but this might be for a future article.) This lack of interoperability has consequences for planning the deployment of CBTC, and that planning needs to be done before starting implementation. That’s where the supply strategy fits in.

To read more about this article and to have full access to premium content SUBSCRIBE to Basic Subscription Plan or Premium Subscription Plan.

[/restricted_content]

[membership_content level=”basic”]

Now to ETCS. I can already hear some of you think, well if a supply strategy is needed due to lack of interoperability, that should be a non-issue for ETCS. Because as everyone knows, ETCS is interoperable by default, right?

Sadly, not necessarily. I have seen many examples where railways assumed just that, “automatic” interoperability of ETCS subsystems from different suppliers, and then had a rude awakening during commissioning, when those diverse subsystems did not interoperate as expected. There are several reasons for that, which I might explain in detail in another, future article.

So, if ETCS is not automatically interoperable, how does a railway ensure that it is? That is where a supply strategy can help greatly. The necessary steps are a bit different from what makes for a CBTC supply strategy, but the first fundamental question is the same:

How many suppliers (here: of ETCS trackside subsystems) are desired for the network-wide rollout?

Let’s unpack this to highlight the finer details of this question so that they don’t get overlooked. The railway needs to start with their considerations for the ETCS trackside subsystem, because this is the subsystem that defines the specific ETCS application of that railway, including the signalling principles. Those principles may slightly change to accommodate ETCS better, but nonetheless they will be incorporated in the trackside subsystem of ETCS, whereas the onboard subsystem should be “generic” (at least that’s the theory for ETCS).

The next important point is that the railway needs to look beyond the first project introducing ETCS, across the entire network-wide rollout, even though that rollout may take decades to complete.

And lastly, the word “desired” in the question above indicates a preference of the railway.

Let’s look at the consequences of the possible answers. There are only two scenarios with practical relevance at this stage. Either the railway can live with (or even prefers) a single ETCS trackside supplier, or they think they want more than one. The preference of a single supplier is usually only good for rather small networks. A good example in Europe would be Luxemburg where the entire network was equipped with ETCS Level 1 from a single supplier (Thales, in case you wonder).

Most national rail networks will be much larger though, and therefore the most common answer to that first question is that more than one ETCS trackside supplier will be needed in the long run. At this point it does not matter so much whether the number of suppliers will be two, three or more. As long as it’s more than one, the recommendation for further proceeding will be the same.

Let’s take a step back and think about how the first introduction of ETCS will mostly look like. The railway organisation identifies one pilot project and procures one ETCS supplier for the trackside fitment of that project. I believe that is a common mistake which will likely lead to interoperability problems during the future stages of the ETCS rollout. This is because the specification of the ETCS trackside application in that scenario will only be agreed between two parties: the railway client and the selected supplier. One of those parties understands the ETCS standards and their finer details much better than the other. Take a guess.

Yes, of course, the supplier has practical experience with implementing ETCS standards over twenty years or more. The best the railway may have to offer is participation in the ERTMS Users Group, which is much less practical experience when it comes to detailing the system requirements of that ETCS application. Now what would you think is the interest of an ETCS supplier who is first in a new ETCS market? That’s right, market domination, or less dramatically, maximization of future market share. One way of getting there is making it more difficult for other suppliers to enter this market in the future. How? By minute tweaks to the first ETCS specification that fit perfectly to the first supplier’s solution but not that well to any other supplier.

Ah, you think that is a conspiracy theory and would not happen in practice? That reminds me of someone’s quote that the difference between a “conspiracy theory” and proven reality is about eighteen months. No, really, I have seen that happening before, even though most railways would not care to admit that things went out of their control because they understood the ETCS standards much less than their supplier.

So, what happens in this scenario during the later procurement process for the second ETCS project is that one bidder appears to understand the railway’s requirements much better than all other bidders. You may have guessed who that informed bidder is, of course the supplier of the previous first project. Quite often that supplier can also offer a more attractive price, so there is a big temptation for the railway to just give that second project to the same supplier that already delivered the first project. And with that decision it will get more and more difficult to bring other suppliers on board in the future. And the railway might wonder how that could happen, given that ETCS should be interoperable and therefore allow for a highly competitive “open” market.

Now that we have seen how things could turn out not so good when going with a single supplier at the start of an ETCS rollout, let’s have a look at how things could go better when working with two suppliers at the same time. In this scenario of joint development of the initial ETCS specification by two suppliers they will keep each other honest. That means if one supplier suggests a feature that may lead to a proprietary advantage, the other supplier will cry foul and that issue will be resolved right away, before the specification gets locked in.

You may wonder whether that joint development process may work even better with more than two suppliers. In theory yes, but in practice it will be much more difficult for the railway to deal with more than two suppliers at the same time. So you could say that two suppliers is the “sweet spot” for initial development of ETCS specifications.

What are the consequences of wanting two suppliers for the initial specification development? They have to be properly motivated, and paying one of them just for helping with spec writing while only the other supplier gets to implement ETCS is not good enough a motivation. So, what is needed here are two pilot projects, one for each of the two suppliers. That way both suppliers have real “skin in the game” and will do a good job of specifying the solution they will have to implement afterwards. You can now see why this is called “supply strategy”. It requires foresight of long-term supply requirements and respective planning to get there.

By the way, if you really have only one project for initial ETCS fitment, say one line of your network, what you can do it cut it in half and give one portion each to your two selected suppliers. That creates an immediate interface where interoperability is needed, and your suppliers must solve that problem right from the start. Perfect.

The jointly developed ETCS trackside specification will be the railway’s template for the entire ETCS rollout, and it will also be the benchmark for any supplier of ETCS onboard subsystems to comply with. “Swearing in” ETCS onboard suppliers to that trackside specification is another critical step for achieving ETCS interoperability in practice. And the sequence of completing the trackside specification before starting procurement for onboard ETCS is another key aspect of that supply strategy.

So much for today. If you have any questions on this topic, contact the editor and you may get this answered in another column of mine, “Ask the Doc”. Thank you for reading.

[/membership_content]

HIGH PERFORMANCE SIGNALLING | Doc Frank Heibel

What is High Performance Signalling?

[restricted_content]

My previous articles in this newsletter column were all related to ETCS, mainly because it made the publisher happy. Happy customers are important, so this is your first takeaway right there in the first paragraph. Hope that makes you happy too, because (you guessed it) happy readers are important to a writer, to me in this case.

Given the name and theme of this column I should really have started off with today’s article. Nothing beats a solid definition at the beginning of writing about something that not everyone may have the same understanding of. So in this article I will explain to you what I mean by ‘High Performance Signalling’, and you are explicitly invited to use this term in your own professional life if you think it has merit.

Besides this overdue definition, I’m also itching to start writing to you about CBTC, the Communications-Based Train Control technology. Why? Because CBTC is arguably the ultimate mainstream technology for High Performance Signalling.

By ‘mainstream’ technology I mean a system that is supported and offered by multiple suppliers, and available globally. This is important for any railway that wants to have a competitive procurement environment with multiple comparable options to choose from. CBTC clearly ticks that mainstream box, and so does of course ETCS.

So what do I mean by High Performance Signalling? And why do I think that CBTC provides that more than ETCS or any other mainstream signalling technology?

I coined that term High Performance Signalling around 2016 because I thought, and still think, that the popular (at least here in Australia) term High Capacity Signalling is unnecessarily narrow. Thinking about the additional benefits besides higher capacity that CBTC offers, I came up with a framework called CARA. A framework is something I use often in my training courses to make it easier (including for myself!) to memorise a list of related things, in this case the main characteristics of High Performance Signalling. CARA is an acronym, hopefully easy to remember as it also a girl’s name, which stands for Capacity Availability Reliability and Automation.

For most railways, higher capacity is their foremost expectation from introducing CBTC, so having the C in first position of the CARA framework makes absolute sense. But there is more to performance than just capacity. The issue here is that if a railway does not focus on those additional benefits they may end up being neglected, with likely inferior project outcomes compared to full focus on all four CARA benefits. (Four letters = four benefits, logical isn’t it?)

The first of the two As in CARA stands for Availability. CBTC should fail less often than any traditional legacy signalling system it replaces. Firstly, because it’s new (and expensive!), and secondly because CBTC is usually more centralised than traditional signalling, and any outage due to a technical failure has much more severe consequences under CBTC. It is therefore very important to design a redundant, fault-tolerant, highly available CBTC system. And once this is done, the higher availability compared to the previous signalling system will be a very tangible and desirable benefit of CBTC.

The R in CARA stands for Reliability. But not technical reliability which is closely related to the availability we already discussed. No, what is meant here is operational reliability, in other words more punctual train services.

CBTC advances this operational reliability in two ways. Firstly, the control of train journeys, including the lengths of station stops, is much more accurate and consistent under CBTC than what is achievable with traditional signalling, so there should be fewer delays and thus better adherence to the timetable to begin with. Secondly, CBTC can also facilitate a faster recovery from traffic disruptions, meaning that trains get back on time quickly after minor delays. This is a real game-changer from traditional signalling where minor delays often grow and accumulate to bigger delays which then may spread out across the network, sometimes causing real chaos for the timetable.

I have heard from several metro railways that talked about their main observations after introducing CBTC that the step change in service punctuality (often called “on time running”) was the most positively surprising and hence most appreciated benefit from CBTC, whereas the higher capacity also gained from CBTC was just “as expected”.

Now to the last letter of CARA and the second A – Automation. CBTC is a technology which provides Automatic Train Control (ATC). The three main functionalities of ATC are Automatic Train Protection (ATP), Automatic Train Operation (ATO) and Automatic Train Supervision (ATS). You can see from all those As that CBTC is literally oozing with automation.

What are the benefits of that? Well, many people think of automation mainly in terms of replacing humans with computers, destroying jobs and livelihoods of railway workers. Unsurprisingly, such prejudice often leads to adversary and opposition to the change induced by CBTC, and often enough to the CBTC technology itself.

I strongly believe that the automation offered by CBTC can be utilised to support human railway staff rather than replacing them. The reduction in human error is a very obvious upside of automation. But wait, there is more.

One of the ‘superpowers’ of CBTC is its capability to regulate train movements, specifically multiple train movements at the same time. This feature often has its own name, Automatic Train Regulation or ATR, but is in fact a function of the CBTC subsystem for Automatic Train Supervision. ATR can, for example, even out the intervals between trains if some are delayed and others on time. This is a useful feature because for a metro railway that relies more on “turn up and go” services than a rigid timetable the consistency of intervals between following trains is often more important than varying intervals to keep the undelayed trains on schedule. Another benefit of ATR can be to regulate out conflicts of train movements, by influencing both trains in a way that separates their arrival at the conflict point and avoids one train being stopped to let the other train pass.

But the real kicker is that the regulating commands that ATR sends to the trains are automatically, precisely and instantly executed by the Automatic Train Operation function onboard those trains. So it is not only ATR, it is the collaboration of ATR and ATO that really hits it off for CBTC.

There are a number of other benefits that could be listed for Automatic Train Operation alone, from energy-saving driving patterns to the removal of driving variability between human drivers. More driving consistency means less need for buffers in the timetable, which can be used for higher capacity, and also improved adherence to the timetable.

So there you have my CARA framework and an explanation what I mean by High Performance Signalling and how it provides more than just high capacity.

A word on ETCS regarding High Performance Signalling. I have written before in this column that the way to improve the performance of ETCS applications is to make them “more like CBTC”. That could mean the addition of ATO, an integrated traffic management system alike the ATS in CBTC systems, and methods for increasing capacity such as cab signalling without signals but with capacity-optimised fixed block sections, or a mix of fixed and virtual blocks to achieve shorter block lengths that support higher capacity. I tend call those enhanced varieties of ETCS “Level 2 plus”, and of course the CARA framework of benefits is applicable to those ETCS applications as well, even if maybe not always to the same extent as for CBTC.

When I introduced my idea of High Performance Signalling on LinkedIn back in 2016 I used the then world’s best sprinter Usain Bolt as an example for the difference between high capacity and high performance. If you fancy reading my thoughts from that time (AFTER you are good with this newsletter of course), check out this article: check out this article on Linkedin

So much for today. If you have any questions on this topic, contact us on the email office@newinsignal.com and you may get them answered in another column of mine that is planned on this website called “Ask the Doc”.

To read more and to have full access to premium content SUBSCRIBE to Basic Subscription Plan or Premium Subscription Plan.

[/restricted_content]

[membership_content level=”basic”]

How many suppliers (here: of ETCS trackside subsystems) are desired for the network-wide rollout?

The next important point is that the railway needs to look beyond the first project introducing ETCS, across the entire network-wide rollout, even though that rollout may take decades to complete.

And lastly, the word “desired” in the question above indicates a preference of the railway.

So much for today. If you have any questions on this topic, contact the editor and you may get this answered in another column of mine, “Ask the Doc”. Thank you for reading.

[/membership_content]

[membership_content level=”basic”]

How many suppliers (here: of ETCS trackside subsystems) are desired for the network-wide rollout?

The next important point is that the railway needs to look beyond the first project introducing ETCS, across the entire network-wide rollout, even though that rollout may take decades to complete.

And lastly, the word “desired” in the question above indicates a preference of the railway.

So much for today. If you have any questions on this topic, contact the editor and you may get this answered in another column of mine, “Ask the Doc”. Thank you for reading.

[/membership_content]

HIGH PERFORMANCE SIGNALLING | Doc Frank Heibel

Managing a Supplier Pool for ETCS

[restricted_content]

In my last column, I wrote that the ideal way of introducing the European Train Control System (ETCS) for a real interoperable outcome is to start with two trackside suppliers, give each of them one pilot project, and have them jointly develop the ETCS specification for your particular application.

Starting with more than two suppliers will increase the complexity exponentially without much improvement in the outcome, so the additional effort is not really worth it.

To read more about this article and to have full access to premium content SUBSCRIBE to Basic Subscription Plan or Premium Subscription Plan.

[/restricted_content]

[membership_content level=”basic”]

Starting with only one supplier is “normal,” but it can lead to problems in that the specification will be inevitably biased towards that supplier’s ETCS solution.

The problems will not necessarily show during the first project, but they will most certainly pop up during the procurement process for the second project.

Remember, if the railway’s objective is to have more than one supplier and interoperability between the products of all suppliers for that railway, obviously there is a need to bring a second supplier on board sooner or later. And what better time to do that than the second ETCS project, right?

But here is what will happen: The supplier who did the first ETCS project will bid for the second project as well. Of course, the objective of every supplier is to gain as much market share as possible, ideally 100%. And since there has been only one ETCS project thus far, that first supplier does have a 100% market share and will be keen to maintain it.

Here is the kicker which will bring the railway into a real dilemma. There is a fair chance that the bid of the first established supplier for the second project will be much more attractive than all the other bids from suppliers that are “new” to this market for ETCS. For one, the incumbent supplier will understand the railway’s requirements much better than all other suppliers. No wonder, since that supplier wrote the specification during the first project. The incumbent supplier may also offer a lower price because all of the initial effort to adopt that railway’s specification has already been done (and paid off) in the first project.

Now the railway organisation will find it very hard not to award the second project to the already established supplier. Better compliance and better price, what else could they decide without violating their own procurement principles?

Yet if that second project gets awarded to the first supplier again, what happens is that the head start of that supplier on the learning curve will increase so that the difference to any other supplier will get even bigger when procuring a third and fourth project in the future. At the same time, the competing suppliers will get more and more discouraged. You can now clearly see the pathway to a single-supplier situation, which, remember, the railway never wanted.

So what can be done? Once the railway started its first project with only one supplier, that decision cannot be made undone. But what they can do is force the introduction of a second supplier in the second project, regardless of how much better the bid of the first supplier may be. How? By dividing the scope of the second project into two parts and awarding those parts to two different suppliers. If these two parts have a direct interface, for example by splitting one line in the middle, there is even a forcing point for interoperability between the two suppliers. Of course, there will be some pain in implementing these two parts of the second project. But at least the threat of a monopoly for the first supplier gets mitigated.

Once again, in comparison to that scenario, it would have been much smoother to kick off the ETCS introduction with two suppliers in the first place. But if that is no longer possible, this is the best option remaining to ensure an interoperable multi-supplier environment.

So now that the railway has established a pool of two ETCS trackside suppliers with their first two ETCS projects, what is the way forward for the remaining rollout of ETCS across that railway’s network?

When the ETCS specification for this network is agreed between two suppliers, the risk of any bespoke peculiarities that are hard to replicate by other suppliers is significantly reduced. Therefore, the future introduction of a third supplier will be much easier than it was to get from one to two suppliers. Onboarding supplier number four would still be easier, but it is important that the number of suppliers is contained at a practicable minimum, greater than one, of course. This is because of the following considerations.

You might have heard the saying “use it or lose it,” often in regard to the muscular capability of aging humans. Guess what? The same applies to your established pool of ETCS suppliers. See, there is a fierce and global competition for ETCS resources, so the ones that have worked on your railway are very likely wanted and needed in other places, too. So if you cannot keep them busy, they will disappear into the sunset faster than you can say “ETCS”.

How do you keep them busy? Now that’s simple, give them ongoing project work. While they are still working on one of your ETCS projects, negotiate the next project after that, including the ideal timing for transitioning resources from one project to the next without issues for either project.

This is called a “pipeline” of work. It requires planning, timely funding allocation for subsequent projects, and, of course, awarding “shovel-ready” projects.

And here is where an all-too-generous size of the supplier pool may come back to bite the railway. A pipeline is needed for EVERY supplier in your pool that you want to retain. So, the very minimum for a multi-supplier setup is having two pipelines for two suppliers. That is obviously more complicated and requires more planning than having just one pipeline, but it is critical to maintain two active suppliers in your pool.

A third supplier requires, you guessed it, a third pipeline of work, and now it gets really complex. Now it gets clearer why, even for conventional signalling in the past, most national railways had only two “home suppliers” that shared the majority of the projects between them.

If you read my previous article in this column carefully or read it again, or just have a very good memory, you probably found that I repeated a few things here that I already emphasized last time. Why? Not because I have nothing new to say but because it is so important, so often overlooked or neglected by railway organizations who think ETCS interoperability “comes naturally,” and because the consequences of getting it wrong are so grave. Having two or more ETCS projects on your network without interoperability is a really big problem. Losing an established supplier due to lack of ongoing work is another really big problem because the railway spent years and lots of money to get them established, and then it’s all lost.

Sorry if you expected the technical “nuts and bolts” of ETCS; it’s a signalling newsletter, after all. However, from experience, the single most important thing about ETCS that anyone has to understand is that it only provides the POTENTIAL for interoperability, but achieving ACTUAL interoperability requires additional planning and effort during the railway’s ETCS rollout.

If you wonder about the ETCS onboard suppliers after I have written about trackside ETCS the whole time, let me remind you that the specific ETCS application of a railway is determined by the trackside specification. That specification should be the benchmark requirement for any supplier of ETCS onboard equipment on your network. In theory, that should be no problem because, in theory, the ETCS onboard equipment should be able to handle everything that the trackside subsystem could require. But in practice you better double-check, especially if you have any “special cases” in your ETCS trackside specifications. (The varying use of “Packet 44” comes to mind!)

So much for today. If you have any questions on this topic, contact the editor, and you may get an answer in another column of mine, “Ask the Doc.” Thank you for reading.

[/membership_content]

ASK THE DOC
High Performance Signalling

Monthly updates and articles by Doc Frank Heibel

HIGH PERFORMANCE SIGNALLING | Monthly Articles

Ask Doc Frank Heibel

Ask directly Doc Frank Heibel for professional inquiery related to railway engineering

Articles by Doc Frank Heibel

What is High Performance Signalling?

So what do I mean by High Performance Signalling? And why do I think that CBTC provides that more than ETCS or any other mainstream signalling technology?

I coined that term High Performance Signalling around 2016 because I thought, and still think, that the popular (at least here in Australia) term High Capacity Signalling…

ask the doc | doc frank heibel

Welcome to High Performance Signalling

Ask directly Doc Frank Heibel for professional inquiery related to railway engineering and read more..