The Myth of ETCS Level 3

Last week in London, I presented on the status of CBTC in Australia, and a few days later conducted my "CBTC Implementer" workshop on transport planning, both times including the discussion of technology selection for in-cab signalling between ETCS Level 2 (with or without Automatic Train Operation) and CBTC. Key questions I received at both events were in fact the same that I use to get every time I talk about in-cab signalling technologies these days:
"What about ETCS Level 3? Why didn't you include it in your analysis?"

I believe there is a big misconception about that almost mystical ETCS Level 3, and I have made it a habit to answer those questions with the following rather blunt statement:

ETCS Level 3 does not exist as a product.

That never fails to kick off intense and controversial discussion. Which is good as it produces a strong counterview to those "pied pipers" out there who (in my view) mislead the railway industry by inferring that ETCS Level 3 is already available or very close to it, ever since it was proposed as the best signalling option for the West Coast Main Line in the UK in the late 1990's (!).

For those unfamiliar with the topic it makes sense to briefly explain the desire for ETCS Level 3. ETCS (European Train Control System) was specified and designed to harmonise Automatic Train Protection (ATP) across numerous European legacy ATP systems, to support simplified and more economic cross-border train traffic within Europe. The work of the supplier cooperation UNISIG, several European railway user associations and the European Railway Agency ERA, resulted in the detailed specification of ETCS and respective products from various signalling suppliers which are interoperable with each other. This interoperable multi-vendor supply market has made ETCS hugely attractive for railways outside Europe and led to ETCS becoming a de facto standard for Automatic Train Protection on mainline railways around the world. The main issue with ETCS was that it was never thought for significant capacity increase as it overlays to conventional fixed block signalling.

Moving block principles that promise higher capacity are currently provided by CBTC (Communications Based Train Control) which is the by far most popular signalling technology for metro-style railways with high requirements for capacity and performance. CBTC however is a supplier-proprietary technology and does not provide interoperability between different suppliers' products. While that may be acceptable for metro railways with mostly segregated lines, it is a major problem for large scale mainline networks that need a multitude of interoperating products for their long-term sustainability.

ETCS Level 3 promises the best of both worlds - a CBTC-like capacity provision by utilisation of moving block, paired with the interoperability benefit of the ETCS standard. No wonder that mainline railway operators developed significant appetite for that miraculous solution.

And now there is me bursting that bubble by stating that ETCS Level 3 does not exist. How did I get to that view, and is it justified? Below is the answer I wrote to the enquiry of the participant of my "CBTC Implementer" workshop, which I think is elaborate enough to explain the issue.

ETCS is characterised by a very detailed SRS (System Requirements Specification) as well as several similarly detailed FFFIS documents (Form Fit Function Interface Specification). In contrast, all that exists for ETCS Level 3 is a high level FRS (Functional Requirements Specification) which roughly explains how ETCS Level 3 is supposed to function.

On that basis I believe it is justified to conclude that:
ETCS Level 3 does not yet exist
(as a commercially available product that satisfies ETCS purposes and standards).

If any supplier claims these days that they "have" ETCS Level 3, it is nothing more than a misleading marketing pitch. What they may have is a system that implements that supplier's interpretation of how ETCS Level 3 may look like in the future, but obviously without the ability to ensure that their interpretation will fit to and interoperate with other suppliers' variants of ETCS Level 3 (thus lacking the main characteristic of ETCS systems i.e. interoperability between suppliers).

I am aware that my above view is quite blunt, but I consider that necessary to counter the unjustified claims that fuel the myth of ETCS Level 3 and misleads large parts of the industry in what I think a very irresponsible manner. I would guess that at current development speeds it will take at least another ten years until ETCS Level 3 is specified to the extent of Levels 1 and 2 and ETCS suppliers can readily offer interoperable products.

Do you agree with the above view, sad as it may be? Or have I just missed some very recent development? If the latter, I'd be more than happy to be educated and revise my "negative" attitude towards the bright, shiny future of ETCS Level 3.

Doc Frank is an Australia-based leading expert on high performance railway signalling, with 16 years of exposure to ETCS in various functions. His current focus has shifted towards CBTC as it massively outperforms ETCS for suitable railway systems, at least as long as ETCS Level 3 is still a "pie in the sky"
(picture from wholeearthnatureschool.com).

For information on Frank's CBTC Implementer workshop or other training courses, visit https://cbtckickstarter.com.